© S Q O a © O Q S Q 0 ® §§§
Classic Reprints in Anthropology Editors Bruce Kapferer and Clyde Mitchell This series reprints classic texts in ethnography and methodology which have exerted a major influence on anthropological thought. The series w ill span a variety o f traditions and w ill also include translations. Each volume w ill be introduced by a preface placing it in the context o f current anthropological debate.. The overall aim is to re-sensitize students and researchers to key issues which have shaped the discipline and continue to have relevance today. ISSN: 1354-3601
Schism and Continuity in an African Society A Stud y o f N dem bu V illage Life
V . W . T am er1,
BERG Oxford • W ashington, D .C .
First published in 1957 by the University o f M anchester on behalf o f T he Institute for African Studies, University o f Zambia. Reprinted in 1964, 1968 and 1972 by the University o f M anchester and in 1996 by: B erg Editorial offices: 150 Cowley Road, Oxford, 0 X 4 1JJ, U K 13950 Park Center Road, Herndon, VA 22071, U S A
© Edith Turner, executor for Victor Turner, 1996
. A ll rights reserved. N o part o f this publication may be reproduced in any form or by any m eans without the written perm ission o f Rerg. Berg is an imprint o f Oxford International Publishers Ltd.
lib ra ry o f C ongress Cataloging-in-Publication D ata A catalogue record for this book is available from the Library o f Congress.
B ritish lib r a r y Cataloguing-in-PubUcation D ata A catalogue record for this book is available from the British Library.
ISB N 1 85973 110 4 (Cloth) 0 85496 282 4 (Paper)
Printed in the U nited Kingdom by W B C Book M anufacturers, Bridgend, M id Glamorgan.
' CO N TE N TS PAGE
P reface to
the
1996 Edition
by
B ruce Ka p f e r e r .......... v ii
P reface to
the
1972 E dition
by
M ax G l u c k m a n ..........
Preface t o
the
1968 E dition ..................................................xxi
Preface to
the
1957 Edition .............................................
xv
xxv
A cknow ledgements ............................................................ xxxiii I II III IV V VI V II V III IX X XI XII
H istorical and Ecological Ba c k g r o u n d ........................ 1 T he V illage : T opography and D emography ................. 34 T he S o c ial C omposition o f the V il l a g e ......................... 61 M atrilineal D escent : T he B asic P rinciple of V illage O rganization ....................................... 82 M atrilineal S uccession and the D ynamics of V illage Intrigue .................................. 131 V illage Fission, S lavery an d S ocial C hange ............... 169 V arieties o f V illage Fis s io n ............................................. 204 T he S tr u ctu r al I mplications of V irilocal M arriage within the V il l a g e ..................................... 234 Political A spects of K inship an d A f f in it y .................. 258 T he Politically Integrative Function of R itual . . . 288 T he C hieftainship............................................................... 3 18 Po s t s c r ip t ........................................................................... 328 B ib lio g r a ph y ....................................................................... 332 I n d e x ...................................................................................... 337 A ppen d ices ...........................................................(end of book )
Map 3 and the Appendices have been omitted from this reprint.
r
S'
C Ç
i
e
PREFACE TO TH E 1996 EDITION IC T O R Turner’s Schism and Continuity is among the more outstanding monographs that came out o f the work o f M ax Gluckm an’s ‘Manchester School’ o f anthropology. T his group developed from a research tradition established by Gluckm an at the then Rhodes-Livingstone Institute for Social Research (now Institute for African Research and part o f the University o f Zambia). Gluckman succeeded the former Director (Godfrey W ilson) in the closing stages o f W orld W ar II and set up a programme for the thorough ethnographic study o f the region which included what are now Malawi, Zambia, and Zimbabwe. W hen Gluckman was appointed to the Chair o f Social Anthropology at the University o f Manchester in 1947, the Rhodes-Livingstone Institute continued as the research base for the school o f anthropology that Gluckman was to develop. Those who gathered at the RLI and who largely built the fieldwork and theoretical perspectives that made Gluckm an’s M anchester School an important development within British social anthropology included Elizabeth Colson, Clyde Mitchell, B ill Epstein, John Barnes, M ax M arwick, and Victor Turner. These...scholarsJucLtheir. own distinctive ways pursued Gluckman’s_insistence_that.any„theQretical and!cQnceptuaL.undetstanding.Qf social forms and their definingideas must be. grounded.in the studyof social. practice. This .approachwhich Gluckman him self elaborated from the example, of. Evans-Pritchard became knownJbrQadly as ‘sit.uationai._analys.is’_.an.d..the.‘extendedr ca.se method’. The line that.Gluckman.encour.aged was also-informed by a m ixture o f Durkheimianism (largely through the interpretations elaborated by Radcliffe-Brown) and the ideas o f M arx and Engels. O verall,.. the theoretical and m ethodological halim.ark o f .the Manchester traditional concentrated on practices which were viewed as revealing the inner ‘logic’ o f what.Gluckman.re£erredto.as.custom or more generally, these days, as culture. It must be stressed that the Manchester position was innovative. Many in British anthropology at the time regarded the Manchester anthropologists to be m arginal to the mainstream and the vii
V
vili
Preface to the 1996 Edition
Manchester/RLI group valued themselves in this way. They were leftdrientedjmcLopposed.tQXQnservatiYe.,teiidencie5L,iixanthrQpQlQgy. In Central A frica they became sharp critics,.Qf..CnlQniaLilule^and werecfiticized in the press. Gluckman him selfwas officially refused entry into the region (and other areas o f colonial rule such as Australia-controlled New Guinea). Others in the Manchester group were similarly prevented from continuing their research. Many o f the Manchester/RLI group were directly active in-radical-politics..in England and elsewhere (although; Gluckman was .more suhduedLin / such affiliation, a fact that exposed him, to ,friendly jibes£rom Jiis colleagues and students). The Manchester/RLI group, in th e fifties and^early sixties carried their generally-radical attitudes-into their anthropological, practice. It was a distinctive feat.ur.e_ o f ..the Manchester. School. They..piQneered perspectives that-only became widely accepted much later in the seventies and eighties. The-.ev.ents o f 1968 in Paris were arr -important dntellectuaL-watershed for anthropology as. for o.thet_academic. disciplines. The. Manchester emphasis on practice was sometimes compared with .Firth,.s. contrast between..organization and-structure. They were very_different. Firth’s observation largely. boiled_dawn_.to.,th.e fact (injmy view fairly trivial) that the representations that anthropologists ancLthe subjects o f their observation make o f their forms, o f life diverge from, what they actually do...He-stresaed-the.-impos'tan.ce o f exploring the.actual- £luid-.oxganizations_o£activit.y...Leacb_pmbably carried this notion further.(a notiorLalready-implicit.in Malinowski) as. did Barth. The developments within.the. Manches.tet.Scho.ol.were more profound. They were not interested.in.the contradictions_within ‘cultural, systems’ (Leach)_ca^in the-dynamic&-Q£individuaLra£ional choice (Barth)—aspects that_they.did„notdgnoresrhut.were,jdirected more fundamentally to the.confcmdictionsjvithinhistcadcaL.,political and economic forces o f a-globaI.nature .which_were^part-o£the=wider circumstance within which particular^humariupQpulations. struggled to recreate or reproduce their social and. cultural ways o f life. (Jonathan Friedman’s critique o f Leach for overlooking the historical forces should be noted.) The^Manchester. orientation,.was more developed than many later and.more fashionable posfc-1968 ,attempts at, a Marxist perspective in other major centres..of.anthropology in
Preface to the 1 996 Edition
ix
England (for example, that at LSE where first Fir.th.and then. Bloch announced.acommLtmetitto.a_Matxism_thatmanifested.a-far_cmder functionalist economism that.in fact had n o n eed o fM arxo f Engels). The originality and possibility o f the Manchester anthropology perspective reached a milestone in Turner’s Schism and Continuity. The argument extends from Gluckm an’s lead, focussing on the dynamics o f social conflict and its foundation within contradictions at the heart o f systems o f relatedness through kinship and marriage. This is the central problem atic that guides the analysis (the contradiction between virilocality at marriage and a pattern o f matrilineal descent and inheritance). The everyday conflict and the course o f this conf lict was rooted in this contradiction which was further complicated by the fact that the Ndembu o f the villages studied by Turner were embroiled in larger historical changes effected through Colonial Rule. Turner shaws.hQw..the-Contradictions.a£ the heart o f Ndembu life became, in effect, more and more.irresolvable in the co.ntext._of. larger, economic and. political developments. He centres his account around the ambitions o f a key and in many ways a tragic figure, Sandombu. Through the struggles ofSandom bu the reader is lead into a remarkably vivid account o f the everyday life o f Ndembu villagers..and,. in.Marxistvein,_how_they_come_tQ_partLcipate in,the transformations o f their own world even as they_are caught in structural processes that are ultimately beyond their..control. Turher*s_metho.dological. innovations in „Schism .and._Qo.ntijmity overcame ..contradictions. ,in. social anthropology...betw een ..actororiented and.. structure-oriented .. perspectives.,... whereby, a concentration on one aspect obv_iated_the other._The overcaming o f this contradiction was. one .concern.o£Gluckman’s-methodological development o f ‘situational analysis’ which Turner elaborated.and transform ed. Broadly,, situational analysis expanded, beyond the sociological use o f cases or instances from life to illustrateJarger systems o f structure orinstitutions ofth e social order.that theanalyst discerned—the case as illustrative. T he aim was-to..demonstrate the intricacies o f larger processes oper.ating._within_the. dy.namics o f particular events. Thus Gluckmart explored how a bridge-opening ceremony in Zululand in Natal revealed the complexities o f a then emergent system o f apartheid. He applied the method to the process
X
Preface to the 1996 Edition
of judicial decision in Lozi courts in Barotseland in Zambia. However, it was Turner in Schism and Continuity who realized the full possibilities o f the approach. N ot cjnly.dicl he demonstratexhe ‘logic* in the event, Turner (by taking a number o f events in_serlesJnvoiv ing the same actors) was able to dem onstratehowparticipants.changed and transformed the very structural. circumstance5_oLtheir_own action. The general significance here o f Turners innovation should not be missed. GLuckman’s perspective opened out to...the.importance o f considering how people themselves constructed their contexts o f actionp.that. is,, they.had a ..role, to..pky=,in the makingjo£-their.so.ciopolitical and cultural realities.Tt_wasL^jmove.away-from.the,objectivist position, .of .th e .anthropological, .observer, a ‘thick.de.scripti.onl in Geettz’jsense.well before the latter had formalized-the. idea. However, Gluckman^and .his colleagues». tegardleiiSm.o£.the.ir.dnsisten.ce~on praetiee»^c.auld.,not,dR fact,, escape .the-case,_ox .the^event„.as an illustration o f .sociaLpxinciples. that they devised.independently o f participants.. Furthermore, .although,they stressed,an„.attention to the_change O F . systems of-social action, and-not their-tim eless repetition, they could not break out o f a form o f analysis that was more about how. systems. remained:the.same,rather_than. how they changed or. transform ed. I note, that, this was also the .marked difficulty o f other perspectives developing . in...British .social anthropology at the time—those o f Leach and. o f Barth. Turner’s analysis in Schism and Continuity broke out o f them ould. He did so in a marked Hegelian manner. Thus Turner analysed each crisis in the cQRstant.ly:.changing flux o f Ndembu village life as a dynamic o f rupture-teachingJtowards its overcoming in a new synthesis {resoiutiQn)_Qrorgani?ation-afjs_aciai relations. T he wider .env ironment or political economic field o f actixdtieSL.at.once.cexternal. an
Preface to the 1 9 9 6 Edition
xi
cultural conception andpractice. Coining, back tp the Hegelianism in Turner’s approach, he does have a notion o f an ultimate ‘higher unify. T his .is. estiablishejd in ritual and the growth in importance .of rites,. such_as_the.Chihamba curing rite, that cross-cut the social boundaries ofkin.gtQ up.an d village. The Chiham ba and its. political., importanceJs. driven..in the unresolvable divisions and ruptures emergent n ot .jus.tL-fmm_.the . contradictions underlying the ‘traditional’ order. (thejcontradictipn o f the m atrilineal andviriloeal principles) .butLinjemhracing.glob.aliy connected political and economic changes. In an important.way the Chihamba rite is generated-.and..reinvented-in_sudxjcircums£ances. It does not maintain the system, as a more. functianalistanaLysi&.might have'it (e.g. Gluckm an’s analysis o f the rites-.of-SwazLand-Zulu kingship that were influential on Turner but from whose implications Turner broke). Rather it is integral w ithin a. dynam ic.of Ndembu social and political reconstitution. The discussion o f the Chiham ba and other rites in Schism and Continuity is preliminary to what became Turner’s key focus—the symbolic processes o f rite. His later work, Chiham ba the "White Spirit, is a wonderful elaboration on themes indicated in Schism and Continuity. In.ihisiaterjgtudy flight tP his. own creative spirit,_engaging.avariety.Q£.pers.p.ectives..fromjotber_£ieldsjo£enquiry (literary, studies, Jungian and Freudian psychQ-analysis)_tQ_an understanding, o f Mdembu symbolic processes anxLhow.they^extend a comprehension o f the dynamics o f human _heingJthatLgo.es far beyond a particular cultural/histQrical cQntext...B.ritish anthropology up to and beyond the work o f Turner is often characterized as static British structural functionalism . T his common enough stereotype can only be so if his work is ignored and attention is focussed on what turned out to be the far more conservative and stultifying efforts o f those centres o f anthropology located in London and Oxbridge. I emphasize that it is by reading Schism and Continuity that a full grasping can be achieved o f the radical direction that Turner was to lead the anthropology o f ritual and symbolism. Turner (and his_Manchester colleagues) are what might be generally refexre.d, tO-as.dp.ractice theorists’. Turner should be compared w ith more recent developments in anthropology, for example, those o f
xii
Preface to the 1 9 9 6 Edition
Sahlins and especially Bourdieu. T h e latter explicitly sets his approach to practice in contrast to that developed from Manchester. He chooses to distinguish his line from that o f Van Velsen (The Politics of Kinship). Van Velsen’s study was conceived o f as a development from that o f Turner. It took an individualistic strategic choice direction. This was a possibility o f the Manchester orientation but probably a retrograde step (I state this seif-critically because it is also a direction in my own early work written in the Manchester context). Bourdieu’s line on practice is also potentially individualistic as evidenced in his free borrowing from the Am erican pragmatist traditions o f symbolic interactionism and ethnom ethodology. Turner’s analyses never hinges on a simple politics o f self-interest or o f strategy and, in my view, extends beyond many o f the otherwise positive and fruitful aspects o f Bourdieu’s work. Bourdieu attempts a synthesis o f Husserl’s (also Heidegger’s) phenomenology w ith a Levi-Straussian structuralism. His approach is extraordinarily illuminating but his synthesis is not thoroughly successful and the way human beings can shift their doxa or radically alter the circumstances o f their habitus or habituated activities is never clear. Bourdieu is committed to dynamics o f the reproduction o f the same and is not oriented to the production o f difference and originality. In this, I consider, Turner in Schism and Continuity and in later work is probably more successful. He does not attempt to force together tw o forms o f analysis (structuralism and phenomenology) that in Bourdieu’s treatment at least appear to be incompatible. Turner’s is an approach to practice—an approach that never relents on the density o f the cultural processes ingrained and developing out o f practice—that indicates how new modes o f symbolic comprehension and structures o f daily activity can be generated. There is much else I could say about Schism and Continuity by way o f introduction. For instance, here is an orientation that does not seecultural/socialfarm sas^coherem ly.boundecL.systerns^Turner developed-on-discussio.n-at_Manchester_concerning~.the_use.-.o.£_the concept oTsocial field^horrowecLfrQro the ^ c M pjsychologist. Kurt Lewin (see Turner’s .later.The.Drjuxos_Q/_Aj^cti.Qn)==whic.h-CQncentrated on the dynamics o f social structuration w ith an open field o f forces. Turner in his development was him self a generative centre
O o Preface to the 1996 Edition
xiii
for perspectives developed by others within the Manchester tradition (e.g. Epstein, Bailey)* A lthough intellectually unrelated, Turner’s unbounded field notion was ‘reinvented* w ithin the contemporary suggestions o f a postmodern anthropology. In Turner too we have a sense o f the many voices o f Ndembu villagers as they participate in the fashioning and refashioning o f their own existential realities. Schism and Continuity is a central work in the history o f anthropology. It is critical for a thorough understanding o f Turner’s later writing. But it has far more than a historical significance. It is exem plary o f the creative possibilities o f anthropological ethnography, the centrality o f rigorous ethnography in the anthropological contribution towards more general human understanding. T he work should be read and reread as one example o f how anthropologists might go about their research and develop analytical understanding. T his is particularly so at this historical moment in a general discourse w ithin anthropology concerning new analytical directions and the role o f ethnography in demonstrating them. B R U C E KAPFERER
o
o o
© o o o o
o o
o o o O
© © © a cs O
© © o
© © ©
© © ©
© ©
©
PREFACE TO TH E 1972 EDITION O C IA L anthropology in its m odem form is o n ly about forty years old, and during its short history it has changed ana developed fast. For fo>m.theJdbme„mhen..,anthrQpologists.hegan to carry out intensive ileldrstudies,..they, have experim ented w ith manylmethods: o f presenting their increasingly jrich .data. about socisd life. D r. Turner attempts a n ew m ode o f presentation in this book, but it is a m ode w h ich grow s out o f the history o f anthropology since the W ar. H e has h e re co m h in ed .general analysis w ith the individual, case-rstudy.in whatJLconsider_a..most fruitful and iHuminating w ay. W hen social anthropologists began to cope.w ith the. problem o T handling their very, detailed data on the tribal societies o f Africa, Oceania, Asia and the. Am ericas, they _seem_ first. tnJbaye sought to establish that some kind o f system existed. in_these societies. M alinow ski found the system in the interdependence o f culture,.RadclifFe-Brow n found it in social structure. T he
S
ne^t_generariono£anthropokQgistsintheBririshÇQmmo.nweaîth, among w hom I my se lf fell, continued this search for system. A s I see it, our analyses abstracted a set o f regular interconnections between various social relationships, ecological relationships, modes o f belief, etc. From the tim e w hen Evans-Pritchard analysed the Azande beliefs and practices relating to witchcraft, oracles and m agic, to show that these constituted a rational philosophy o f causation and philosophy o f morals, w hich w ere related to a particular m ode o f social organization, we.Jhave e3cbihi.ted_ th at, there, is ,a , system atic.structure in.-one. Jield o f tribal life after another. T he evidence for the existence o f these structures was m ainly qualitative, and consisted largely in illus trations from different situations o f action in a num ber o f families, villages, or political groups. These iUustrative,data-were-rarely related„tO-one_another. _In_r.esult,,when one,readsjeven the best books o f this period, it is. n ot easy .to. put th.e. systeminto.. w orking operation ,in actual life . In .order to. demonstrate—the.„existence o f system at all, w e . discarded m uch o f ¡the Hying, reality „about I consider, that this _was,a .necessary. stage_itrJhe_jdeyelcpment xv
XVI
Preface to the 1972 Edition
of_ our discipline. .. Faced. and form s o f social rektionsMps JvJiich_occ.urJii_t3ie,_tdbaL,sp.cieliesf w eJ ia d first to produce a m orphology o f tKeir,.stmctur.es,_.and general analyses o f h ow their, systems- w orked,, to . exhibit, .some stability w ithin constant , change .-of—personneL and—relations between personnel. B ut I think all o f us have felt that m ore could be done w ith the detail o f our data. It seems to me that our systematic know ledge has now developed sufficiently fo r us to achieve the next step. African Political Systems (1940), a col lection o f eight studies, marked an im portant step in the systematic study o f one field, and it was follow ed b y a series o f comparable, but fuller, monographs. Y e t it is significant that political studies since the W ar have, so to speak, put their systems back into a particular history, w hile still aim ing at the demonstration o f systematic interconnections : this is m arked in, fo r example, Evans-Pritchard’s The Sanusi o f Cyrenaica and J. A . Barnes’s Politics in a Changing Society, an historical analysis o f the Fort Jameson N goni. Some anthropologists, notably Schapera, have even m oved fu lly into a m ethod o f w o rk m ore like orthodox history, but still seeking for a systematic structure. T he same change is likely to occur in the fields w hich social anthropology has made peculiarly its ow n : for the problems w e m eet in the field o f tribal politics and law , w e share w ith political science, history, and jurisprudence. T he other fields are the study o f domestic and village life, o f fam ilial and kinship systems, and o f ritual— religion, m agic, w itchcraft, etc. And here Turner has made a significant contribution, follow ing leads set, fo r Central Africa, b y M itchell, Colson and Barnes. Great classics on domestic and village life, like Fortes’s books on the Tallensi and Firth’s on the Tikopia, analyse fo r us the regularities they find in a variety o f actual situations and relationships, and the interdependence between those regularities. Theirs are mag nificent analyses. B.UJL.they„.use._ihe._m.ethod_o£-appEopriate, illum inating illustration, and hence, in a w a y , o f chance illus tration. W e cannot from their analyses,w o rk o u th aw iaiiy..o n e group o f people lived, at. a.-particular. time-.and in a_particular place, w ithin this social structure and using_thes.eL_xusioms. Turner has used a different (node o f iilustration, w hich I consider deepens the understanding w hich he givesL_us .o f Mdembu. tribal life. Indeed, it w ould be more_acc.uratej:o.sayLthathe^abandons
Preface to the 1972 Edition
I
V / >
O' x v ii
illustration. altogether,, to d evelo p w hat .is,. for . African„stadies, a new .m ode ofan alysis. H e first gives us a system atic outline o f the principles on w hich N dem bu villages are constructed, and measures their' relative im portance w ith unusually adequate numerical data. Then he takes the history o f one village through tw enty years to show how these abstract principles have operated through that history, w ithin the chance occurrences o f illness, death and other misfortune, o f go od luck, o f individual tem perament and am bition, and finally o f the m ajor changes w hich have resulted from British overlordship. H e thus shows us h o w certain prindples_.a£_orgamzation_and.xem im.daminmt~values operate through.both.schism s andjceconciliations^-and.how the individuals ...and. gronps ..concerned_try_.to..-.exploit-.-the -varied princ^es-andj«alues_tCL-their.own eiids. Turner him self w ould not claim any particular points he makes as original, for m ost o f them he has taken individually from the w orks o f his predeces sors. B ut this use o f the detailed case-study, in the background o f general system atic analysis, com bines w hat he has taken over into a pioneer study. The late B uell Q uain in his Fijian Village attem pted this task, but his death perhaps prevented the final pulling together o f his analysis, and it did not quite com e off. O ne thing I can do, in this forew ord, is to answer a possible objection to this m eth o d : h o w is the reader to assess that Turner has selected a * typical * village ? This question is partly answered in the careful num erical analyses o f m any villages w hich precede the 4village-study \ Secondly, I am in the privileged position o f having heard Turner present analyses o f other village histories; and hence I can vouch that fo r all its uniqueness, the village o f 4M ukanza ’ is * typical * o f N dem bu villages. Turner hopes to publish separately studies o f some o f these other villages ; and collateral validation fo r this present analysis w ill also com e out o f the study o f N dem bu ritual w hich he is n ow w riting. For he plans to use the same method o f analysis in studying ritu a l: and again I consider this w ill m ark a significant advance in this field o f research. Ttimeiübiasü>.uütLJhás^.tody„„oTJMLukanza-ViUagfi„.aro.und.Jwhat heicalls 4the social, .drama %-r-one o£_a series jofocriseswoccurring in (the history o f the village, when, either a quarreLhetween some o f ih e inhabitants, or a m isfortune ascribed by._foe_pejoplje.and b y divination to ancestral spirits or.iorcery,.precipitates..threats to
o
© 0
0
© © ©
© ©
© © © ■©
© 0 €>
© €1
© 0
© © C
V
xv iii | | | I ; | I ! ! I
I | ! [
Preface to the 1972 Edition
the unity o f the village. T he village._as„ajvvhole, and itsjie ig h hours, as w ell as com|?Qnent groups^_witHn.Jthe-j^Ulage,^tEy. to useLdifferent_forms..ofLredxess^to_meet_these_j;hreats. Turner argues that when the confoc£-emexges._frani...the .opposed-interests and claims o f protagonists acting under a single social-principle, jud icial institutions can.be invoked to meet the crisis, fo x ajratiDnal attem pt can be made to adjust, claims, w hich are similarly-.basedB u t w hen claims are advanced under different..sociaLpm iciples, w hich "are inconsistent w ith one .another..even—to—the-poin t o f contradicting one another., there can be no rational settlement. Here recourse is had to divination of.sorcery or ancestral wrath. causing...misfortune..; and. ..ultimately—to...a-.ritual reconciliation whi£h can reassert a ll.the values-heldJby. decent..Ndemb.u,.Junder the pretence that harm ony is restored withm . .those, yalues. H e shows that after .this situation_has..occurredr there~is-a..teraporary resp ite; but the ...deep, conflicts-between-groups-and-m dividuals in- the. village continue_Thro.ugh...the..intervening,—com parad ve ly peaceful .period, struggles-.continue-tilI-th.ey—precipitate a new crisis. Each crisis..m arks.jhe culrrdna.riojX-o£.anerioxd o f altering, alignments o f pow er and shifts ,of. aUegiance-withinjthe village— though matrmneal attachment, is alw ays-dom inant in the end. . . The crisis itself, through four stages which, hexarefiilly delineates,. is... tem porarily . solved ; .. but. again .. this—involves a definite shift in the village’s internal balance-of pow er. I have sketched som ething o f Turner’s m ethod : he can speak better for him self.. B u t I m ake a few points to support him . First, I hope no one w ill turn aw ay from his analysis in dislike o f the phrase * social drama *. Several o f us have tried, w ith Turner, to find another phrase w hich is less likely to meet objec tions : w e have failed to, and he w ould be grateful for sugges tions. Secondly, I w arn readers that it is not easy to fo llo w the com plicated story o f this village, w ith its com plicated internal genealogy and its im portant links in other villages. Turner has done w hat he can to help the reader, w ith a main genealogy and periodically w ith subsidiary genealogies o f those involved in each drama. B ut the labour o f learning w ho the characters are, and h o w they are related to one another, is w ell w orth w hile. The reward is equal to the labour. I m yself found that I soon got the main characters clear, and the rest follow ed easily. Then I became absorbed in the story itself, and particularly in the tragic
Preface to the 1972 Edition
xix
story o f Sandombu— m oved n ot o n ly b y his personal am bition, strong though that was, but also m oved by the pride o f his lineage, to struggle for the village headmanship w hen everything was against him . W ith his generosity and capacity for affection, he fought w ith his quick tem per and the curse o f his sterility, to achieve a headmanship ; and Turner shows clearly out o f his general analysis, that Sandombu was doom ed to lose, and on the w ay to defeat to incur a reputation fo r sorcery. H e is a tragic figure, indeed. B ut w e learn too that his rivals could not wholeheartedly take advantage o f his weaknesses ; fo r when they had overcom e him , they in turn w ere m oved by guilt, for, under the dom inant value o f N dem bu, was he not bom from the same w om b as they ? Into this story,Turner..has_woven. a.striking generaLanalysis, o f jschism o f groxms . and .relationships,^.and...o.£...c.onrinmty o f society .md_principle-and_Js^ue_ahov^e..jthe.-Schism=^a.xoniinuity
o f a * com m unity o f suffering ’, fo r it. is..misfox.tuneLJwliicli assembles the cult-groups w hose rituals dram atize_foatxontinuity. Thu^incidentally. as the. story .unfolds _we leam wiiaLareJ^Ideinbu beliefe and customs. W£__s.ee-Jh£se_j^c^kmg.~i^ of social „life,.jäh ere, they both control people, and are exploited b y „people. .„Custom s, and.. heKefe„ate.»jreal,„system atizedjthxough social relationships : they are n ot mere adjuncts._Incw orking all this, „out, Turner, also-t-and this is an achievem ent— brings in the changes o f values m d. prindiples ,iQ fJ.Q rganim ttonjdbich are em erging out o f British overlordsnip^the. developm ent.o f w ageearnjngand cash-cropping,Jthe estabUshmento£peace,_the^3kiIlin^ out of-gam e w h ich has struck at the dom inant m ale v alue o f hunting. In all this I see a contribution to our know ledge o f A frica, and to anthropological theory and m ethod. It is achieved in a book w hich w ill fascinate anyone. M a x G luckman U niversity of M anchester, June 1956
pp
\
o o
0 P R E F A C E T O T H E 1968 E D I T I O N T is n ow m ore than a decade since Schism and Continuity was first published. Seen from this peopectiye it^emeEgesjclearly as ja transitional book between the prevalent structural-function alism o f British anthropology in the 1940s— the period in.w hich I received m y training— and the processual analysis ofltheuisxSos. W hatever influence it. may. have had seems to have lainJnu the rapidly advancing sub-disciplines o f diachronic m icrorsociology and . .m icrorpolitics, Y e t it has-som etim es..heen._fbrgo.tten b y those., caught , up in the first...enthusiasm, for ‘ processualism * that.* process -is.in tim ately bound up _with A structured ancLthat ati^de_q.uate_.analysiso£saci_aLli6b.nece^sitates-a-ngQrousL-consideration o f the relation betw een them ._H istorical-hindsight reveals a diachronic profile, . a.. * tem poral structure.’ -in. events, and this_structure_ cannot-.be.. .understood in isolation. ftom -th c series o f synchronic profiles w h ich make up the . structure o f a social field at every significant point o f arrest in the tim e flow . PrOcessuaLstudies-can... nexerJbe.. the.jnegatiotL-of. structuralism; rather do they put the discoveries of. structuralism, to-new.use, in the„peope_c.tives^ In^HsL.intEoduction .to„the Jihst„edition, o f this . hook..P.rofessor G luckm anJiascalledaltentiontothe^ w ayT n-w hichxasem aterial is(used to. facilitate, w hat he h aslater described asd_the-intensive study-.o£-the-processes-.o£.CQntrol in..a.limited_.are.a..o£jocial life view edlover_ a. periQd„.of_tim e’ (Introduction to The Craft o f Social Anthropology, 1967 b y A . L . Epstein, London: Tavistock Publications, x vi). H e contrasts this ‘ extended case m ethod * w ith ‘ the m ethod o f apt illustration.’ . (of. abstract.. structural principles), and suggests that its systematic.use. woulcLdeepemour understanding o f law an d m orality. B ut it was not solely w ith the^oilecdQn-cdLa_difleren.t...ktnd„.Qf-data.„thatJ_w:as„concerned, eyen .at _the_time_of w ritin g the book,, but .w ith. a_difEerent kind o f analysis. In form ulating, the notion o f
I
! J
xxi
?
JL_
0
0 © © 0 0
0 o © ©
© © ■©
©
©
0 0
© ©
©
© ©
€1
■
IT in T:
O c. •
9 O £•N \
c
O o c
o o C) .o 0 v..7
0 i'"'. \.•• €)
C
O
u
xxu
Preface to the 1968 Edition
towards the nQdQn..Q£stiidyiiig.JdbL&..s0:ucturs pC.successive events im social processes o f varying scope and depth. Rut. while .the synchronic-structures. investigated by .many^anthropologists-could be shown to rest upon custom and.habit,.aud. very -ofteu-toxemain beneath .the level o f conscious awareness,.jhe. .d^chronic_ structure o r .‘ processional form s’ that I was interested ...in.,exposing-.and analysing developed out o f clashes and.. alignments__of .human volitions and purposes, inspired by private and public interests and__ideals. , .Yet the social drama itse lf represented a. com plex interaction bet ween .normative ,pat.terns.-laidsrx3o3ivn_in-.the,caurse o f deep regularities, o f candid.Qnmg,and.^
immediate aspirations, ambitions and other conscious goals and strivings o f individuals, and groups in. the here. and..now.-.. A t the tim e, how ever, I was only able to raise some o f the problems o f processual analysis, not to provide any o f the answers. Thus, I was able to suggest that to. the.differentphasesofthe-saciaLcham a, breach,...crisis,^_redressive_ ..action.,...and... immediate-...result,....there corresponded .a particular .style .of. social in teraction particu lar patterns o f norms and values, specific .types. ofLgoals-^ndLgaah orientated. behaviour and specific form s o f conjunctive^ancLdisjun ctive behaviour. B ut I was unable to probe these differences m inutely or especially cogently. For one thing, there was little at that tim e in the w ay o f com parative data o f this type. Since then there has been a considerable accum ulation o f extended case material, for exam ple, in the works o f van Velsen (The Politics o f Kinship, 1964, Manchester U niversity Press); G ulliver (Social Control in an African Society, 1963, London: Routledge and Kegan Paul); and Abner Cohen (Arab Border Villages in Israel, 1965, Manchester U niversity Press) to name but a few , w hile new theoretical perspectives on the study o f social processes (particu larly political processes) have been opened up by such scholars as Gluckm an (Politics, Law and Ritual, 1965, Chicago: Aldine); B ailey (Politics andSocial Change, 1963, B erkeley: U niversity o f California Press); Adrian M ayer (The Significance o f Q uasi-groups in the Study o f C om plex Societies, 1966, in the Social Anthropology o f Complex Societies, A .S .A . M onograph N o. 4, London: T avistock Publica tions) notably his concept o f the ‘ action set \ a concept further elaborated b y G ulliver (Dispute Settlem ent w ithout Courts : the N dendeuli o f Southern Tanzania, paper given in W ennerGren Sym posium N o. 34, 1966, Ethnography o f Law ) ; Sw artz,
Preface to the 1 9 6 8 Edition
xxiii
Turner and Tuden (Introduction to Political Anthropology, 1966, Chicago : Aldihe) ; Swartz (Introduction to Local Level Politics, 1968, Chicago : Aldine) ; Nicholas (Factions : a Com parative Analysis in Political Systems and the Distribution o f Power, 1965, A .S .A . M onograph N o. 2, London : T avistock Publications) and the articles on case m ethod by A . L. Epstein and van Velsen in The Craft o f Social Anthropology (op. cit.). O ther scholars are making contributions to a steadily broadening stream o f studies in processual analysis, but clearly a great deal remains to be done i f serious and rigorous comparisons are to be made in crosscultural terms between diachronic structures. T o end on an ethnographic note : I have but recently learnt that headman M ukanza died at a ripe old age in 1967 and was succeeded b y Kasonda. I predicted in this book that Kasonda w ould found a new farm and that Sakazao w ould remain in Mukanza V illage w ith most o f his lineage kin. This is one illustration o f the difficulties besetting prediction in anthropology, for in 1954 when I left the field t did not anticipate that Sakazao w ould die w ithin a few months and that M ukanza w ould live for a further thirteen years. B ut since m y m ain doubt as to Kasonda’s succeeding to the M ukanza headmanship was based upon his relative youth in 1954, and not upon his capability o r the extent o f his potential political support, m y error was not perhaps so grave. In 1967 Kasonda was sufficiently mature to meet the tacit age requirem ent for a headman— and a headman he became !
I
PREFACE T O TH E 1957 ED ITIO N ‘ General Forms have their vitality in Particulars, Sc every Particular is a Man.’ W iix iam B lake ,
Jerusalem.
N this book I attem pt to isolate the cardinal factors under lying N dem bu residential structure. I^ ^ u sjd ieJ n ^ sttg atio n ujxon-tibte village, a significant locaLunit,-an fsm d yititw o ways. Firs&,_! have, com pared, a number. oE.villages, w ithjceference to such^measurable .criteria a?..sl^e...md_geneafo^£alj:omposition. Secondly,.! have analysed.a. sequence. o£sociaLdram as.involvm g the membership o f a single village,- and-the m em bers..of other
I
»
xxv
h
o o
o Çi
O O c o
o o o o o o e
©
o O
o o Ci
c ü
©
o G O G © Ci
© O
X XVI
Preface to the 1957 Edition
viU ag _ esJjiik ed „to It^ iajthe first instance I have examined j:eg]dadries o£^rm,ox:curring throughout many v illa g e s . in . the,secQnd,.,Ldiscuss_regularities o f process in the social maturation, o f. a. single,.,village. The tw o approaches complement one another.
B y num erical analysis o f genealogical and census data I was able to infer the effective principles determining village structure. These on the w hole w ere in conform ity w ith the ideal pattern o f residential relations presented by informants. W ithin villages the dominant principles influencing residence w ere maternal descent and virilocaH ty.1 U nder-hldem hu ~conditions, „conflict between them was xneradicable_and accomited..iEL_considerable measure for. the unstable and.„fissile~characteno£~villageL-organization and fo r ..the ...high, degree o f .individualjonobility. It is possible that hunting, a purely masculine pursuit, and virilocal marriage, w hich binds together male kin in local descent groups, are parallel expressions o f structural opposition between m en and wom en in this m atrilineal society. Hunting, the m en’s sphere in the basic econom y, is socially valued above its objective con tribution to the food supply, and is highly ritualized. Cassava cultivation, the w om en’s sphere, is correspondingly undervalued, and is ritualized to a lesser extent. Y e t w om en’s w ork ensures the physical survival o f the group, since hunting is fitfu lly pur sued and success in it is uncertain. H unting im plies a rather greater stress on econom ic co-operation than cassava grow ing, although in all sectors production is m ainly individualistic. Throughout the W est-Central Bantu, hunting is linked w ith eminence or aristocratic ranking. Am ong Ndem bu, professional hunters are highly honoured, and all men hunt to some extent. Hunting is equated w ith virility and reinforces in some ways the structural opposition between m en and w om en. For mar riage, as noted above, is v irilo ca l; wom en, on w hom the social continuity o f villages depends, reside at their husbands’ villages after m arriage. Nevertheless, maternal descent governs prior rights to residence, succession to office and inheritance o f property, 1 * Virilocal * in this book refers to die post-marital residence o f a woman in the village to which her husband takes her. ‘ Uxorilocal * refers to the postmarital residence o f a man in his wife’s village. ‘ Patrilocal ’ refers to residence in one’s father’s village. * Avunculocal ’ refers to residence with one’s mother’s brother. ‘ M atriJocal’ refers to residence with one’s mother.
Preface to the 1957 Edition
XXVii
even o f guns, the professional hunters’ m ost cherished items o f equipment. T h e nuclear residential group consists o f m ale m atrilineally related kin. T o rem ain together this set o f kinsmen must im port their w ives from other village lineages and export their sisters. B ut w ith maternal descent as the basis o f village continuity' a contradiction arises between the role o f men. as fathers w h o w ish to retain their w ives and children w ith them , and their role as uterine brothers and uncles w h o w ish to recover the allegiances o f their sisters and sisters’ children. W ith ou t that allegiance m en cannot found enduring villages n or can they effectively press their claims fo r headmanship w ithin their villages. Thus both marriages and villages are inherently unstable and in-laws struggle continually fo r control over w om en and their children. Another consequence o f virilocal m arriage m aking for in stability in residential structure is the great measure o f auto nom y it confers on the m atricentric fam ily. Frequent divorce weakens the link between father and children but strengthens the tie between m other and children. B u t virilocal m arriage separ ates a m atricentric fam ily from its maternal kin-group during the m inority o f the children w h o are reared in their paternal village. I f they return to their maternal kin after the divorce or w idow hood o f the m other, die narrow er lo yalty o f the members o f the m atricentric fam ily to one another tends to com e into conflict w ith their w ider allegiance to the village as a whole. A matricentric fam ily matures into a uterine sibling group w hich is the m ost frequent unit o f secession to form n ew villages. Radical incom patibility, then, betw een m aternal descent and virilocality gives a keen edge to conflicts between uterine kinsmen and husbands o f w om en, w h ich result in quarrels between vil lages ; to m arital conflicts, producing a h igh divorce rate ; and to conflicts o f loyalties between narrow - and wide-span maternal descent groups, w h ich inhibit (he developm ent o f deep localized lineages. Another effect o f virilocal m arriage is a high rate o f patrilocality for children o f m ale village kin. Children adhere closely to mothers, and w h ile a m arriage lasts the children stay at their father’s village. Since villages are sm all (mean size— 10*6 huts), they tend to be at once shallow lineages w ith a fringe o f seminal children and cognatic kin, and bilateral extended families, i.e.
x x v iii
Preface to the 1957 Edition
often the headman’s uterine sibling group, brothers outnum bering sisters, and their children. There is a tendency to m erge kin on the m other’s and on the father’s side as members o f a genea logical generation. W ithin the village, membership o f a genea logical generation cuts across affiliation b y m atricentric fam ily and unites cross-cousins w ith parallel cousins. T h e senior o f tw o adjacent genealogical generations exerts authority over and levies respect from the jun ior. O n the other hand, alternate generations jo k e w ith one another and behave m ore or less as equals. Adjacent generations tend to build huts in separate arcs or the village circle, alternate generations in the same arc. Generation oppositions and alliances tend to reduce tensions in the relationsmps between matricentric families and lineage seg ments, and between children and sisters’ children o f headmen. M arriages between classificatory grandparents and grand children w ho frequently belong to the same maternal descent group, and marriages between cross-cousins, also tend to bind the various com ponents o f the village together. Nevertheless, aespite-these-centripetaLand..accrerive-tendencies N dem bu_villages remain, inherently unstable.— N dem bu- have the ideal aim o f building up large villages.although.this..is..:un ceasingly _rebutted by. -rea lity.-. For the.-XaliensL described - b y Professor Fortes the deep localized lineage is the-skeleton o f the social structure. For N dem bu, on the other hand, a deep, lineage is a seldom-realized goal, the end-product o f the sagacious manip ulation b y headmen o f a number, o f ..organizational, principles w hich confiict w ith-one -another-in-various_situations. N dem bu villages have a w ide range o f spatial m obility and each maternal descent-group in the course o f tim e is scattered throughout different vicinages (discrete clusters o f adjoining villages) over the w hole region. Thus vicinages are hetero geneous in com position, for adjacent villages are seldom linked b y maternal descent. T h e m obility and instability o f villages partly determines and partly is determined b y the absence o f political centralization. T he Lunda ancestors o f the Ndem bu came from the great pyram idal state o f M w antiyanvw a’s empire in the C on go, but in course o f tim e succumbed to the decentraliz ing influences o f their w a y o f life, accelerated b y the slave trading and -raiding o f the nineteenth century. B u t a vestige o f the Lunda state pow er remains in the ritual role o f the Ndem bu
KJp B
Preface to the 1957 Edition
xxix
chief, Kanongesha. T h e chieftainship sym bolizes the unity o f Ndem bu and their ownership o f a com m on territory. T h e m stabiiity o f the secular social. structure can be .palliated but.-not controlled b y secular means. B itu a l...^ otâated -w ith corporate groups...such _as lineages ancLjülage& is^meagre—and canuonly act as_^.Jém poraiy.b ra k e, against, fission^_ButL.ritual periform edby cu lt associations, that cut. across^.inUages^jyidnages and even adjacent .chiefdoin&.of Lunda. o rigin , acts tokeep L th e com m on valu esof.N d em bu socd ety.con stan tlybeforeth erovin g individualists o f w h ich itL.is com posed— These,. values include historical renow n, hunting, and virinty,. fertility .and m otherhood, and health and strength. T he m obility o f the-society. isreflected in the-contingent and occasional character o f _the~rituaL T h e m k fa ttu n e s.o flifo .in clu d in ^ .b ^ wox3ien’sreproductive.disorders,andsevereillness jfor.both.sexes, are.attributed to the punitive action o f ancestor spirits, w ho are exorcised and placated in rituals specific to each m ode o f affiiction. /."The ..curt association is made up_,o£ .doctors aixd_adepts who w ere themselves once patients m d can(hdates ,in.that cult. The w idest com m unity o f N dem bu is therefore a com m unity o f suffering. iu c th e c o n te x to f the com m on .values o f the w hole society are stressed in „symbol,..mime... and .precept. T he associations are transient-grQUps.caUed.into-existericeJby the unpredUctablernisfortuiies o f m obile .individuals. H ence the cult groups are too fleeting and shifting in com position to develop internal stresses and divisions. „It.is,,. I suggest, because the organisational principîçs. w h içh govern, the seadar .stm cture are contradictoqLand_produce percnniaI_conflicts--between-persons an d groups, that rituals, are constm dy being_perform ed.. b y unitary .though transitoty—associarions^ a n d thaL-these-rituals stress com m on values_.over..and above. th e.d ash „of..section al interests. N dem bu ^ritual does n ot reflect©.or express,_as.does Tallensi_ox-S-wazi.„ ritual, the structure o f „a stable,,society,, w ith ritual „role. _corresponding, as it w ere, to secuiarrole-i.-rather, it compensates fo r the integradonal deficiencies.„ofla.poHticaUy-unstable-society. Toverty_Qf_seculaxstams is confrQrLted_with rich developm ent o f ritual roles in m any c ult associations. T h e rang e o f effective political or econom ic co-operation is .small _;.„aUome perform ances o f ritual m ore than a thousand people m ay attend. In Chapter O ne I briefly describe the historical and ecological
o o
0 © O
O
o
0 o 0 Q
© ©
© © ©
© ©
© ©
O
G ©
©
© ô
© ©
Q
C
©
©
XXX
| ]
| 1 |
Preface to the 1957 Edition
background to this study. Chapter T w o presents the relevant dem ographic information» w hile in Chapter Three I attem pt to isolate structural principles governing the social com position o f villages on the basis o f genealogical and census inform ation. T h e next four chapters are centred on the them e o f maternal descent as the main principle underlying village continuity. In Chapter Four the concept o f social drama is first introduced, and it is further refined in the course o f this and the follow in g chapter w hich deal m ainly w ith struggles to Succeed to headmanship between the matrmneal kinsmen in a single village. In these social dramas the pattern o f factional intrigue is revealed and its consistency w ith, o r degree o f departure from , the basic social structure under varying circumstances is discussed. Chap ters Six and Seven are concerned w ith village fission and w ith the pattern o f secession. Chapters Eight and N ine deal w ith the second m ajor determinant o f N dem bu village structure, virilocal marriage, and h o w it operates w ithin, and then between, villages. In. Chapter T en the integrative role o f the cult associations is analysed. In Chapter E leven the chieftainship, in its ritual importance and secular weakness, is briefly discussed. T h at the pervasive^them eof the book is conflict^andjhe_resolu ria m o f conflict .arises..-from, m y preddection_fon_the_yiews, fastJbecoming a theory, o f .that school o f British sociaLanthropologists w ho are com ing to regard a- social. system,as -‘ a^field of.tension, fu ll o f ambivalence,^ofxcteQperatiomand..contra&ring straggle ’-1 F or _ihese„..antbrQpQlogb^ a static< m odel, a harm onious p a ttern ,.n o r.fo e.^
!l
o f a. monistic outlook... A,„sociaL.system. is ~a-Jield~~ofLforces in which, to quote TorteSj^ ^centrifogal tendencies^and centripetal tendencies pull against one anotherJ, and^whose™power to persist is generated -by its ow n socially transmuted,.conflicts.
| '
U nderlying-the. whole-_.smdy-is,,jlie jconcept^^most jrecently reform ulated .by .G lu clq a ^ and Cpkon,® .that ,g£QUpsJiave.J an
l
1 Giuckman, M ., R ituals o f R ebellion in South-E ast A frica ( 1954)» p- 2 1 . Giuckman has recently developed and expanded this theory in his Custom and C o n flict in A frica ( 1955)* 8 Fortes, M ., T he Dynam ics o f Clanship among the T allensi ( 1945)» P* 244. 8 Giuckman, ‘ Political Institutions in T he Institutions o f Prim itive Sodety (1954), pp. <56-8o; Colson, E., * Social C o n tr o l and Vengeance in Plateau Tonga Society ’, A frica , xxiii, 3 (July 1953).
Preface to the 1957 Edition
xxxi
inherenttendency'to segment and then to beemne .bound-together by.erossrcutting.alliances........conOictsJn.one.s.etLofj!ela£ionsJhips are absorbed and-redresse.dinthe.co.unJ3ansailmg.relatio.ns’. My analysis o f village structure is influenced by this conception. But it must be pointed out that among the Ndembu, conflicts in secular, non-ritual, relations speedily sharpen to the point o f irreconcilability in terms o f the maintenance o f local cohesion. The high rates both o f divorce and o f village fission attest to this. But conflicts which split sub systems tend to be absorbed by the widest social system and even to assist its cohesion by a wide geographical spreading o f ties o f kinship and affinity. CenmfugaL.te.nde.ncies... prevaiL.on the. whole...over centripetal, tendencies .at the .level. o fc o r poratekinship and ..local groupings, but.centrifugality is confined .within. the Jb.Q.unds..o,£.the totabsQcio-geagraphicaLsy&tem-.o.Tthe.Ndembumation. A n.overall ritual unity is.co.ntrap.Qsed.to .the.fissile.nature.,ofseculanlife^ecular life shows unceasing attempts to build. up_coherentgrQUps,.hutthese attempts are as unendingiy-.fr.ustrat.ed.by-,centrifugaUancL.fissile tendencies. T h e concept o f Ndembu unity, transcending .all the divisions o f the secular system, is the productofcinnumerable^fitfuliy performed occasions o f ritual, .each couched in the idiom- o f unity through common misfortune.
-Jï
o
© © ACKNOW LEDGEM ENTS HIS account o f the Lunda-N dem bu is based on tw o periods o f field research carried out betw een D ecem ber 1950 and February 1952, and between M ay 1953 and June 1954, after m y appointm ent as a Research O fficer o f the Rhodes-Livingstone Institute. I have no skill adequately to express m y debt to the Ndem bu people o f M w inilunga D istrict fo r their contribution to this book. M any m ore than I can name instructed m e patiently and pains takingly in their w ay o f life. Villagers w ere never offended b y m y presence at m any o f their sacred or intim ate occasions. From them I learnt not only some fascinating facts, but also that human frailty m ust be forgiven— in oneself as in others— i f human social life is to be em iching. M y especial thanks are due to Musona, m y shrewd assistant, to Samutamba, to Sakazao, and to Headman Kajim a. Others w ho gave m e unstinting help were W indson Kashinakaji, M uchona and C h ie f Ikelenge. I am grateful to M r. R . C . D ening, the D istrict Com m issioner, for access to his excellent maps and to the D istrict N otebook. Indeed, m any persons have contributed, directly and in directly, to the shaping o f this book. M any separate skills and much collective know ledge and w isdom w ere put at m y dis posal as field-report became thesis and thesis was fashioned into book. From Professor Forde I received m y first training in social anthropology. It was in his lectures and seminars that m y enthusiasm for the subject received form and direction. D uring m y field-w ork and w hile w riting up, tw o sets o f colleagues, several o f w hom belonged to both sets, gave m e invaluable assistance. I refer to the Research Staff o f the R hodes-Livingstone Institute and to the Departm ent o f Social A nthropology at the U niversity o f Manchester. Professor M ax Gluckm an, a form er D irector o f the Institute and present Head o f the D epartm ent, has given m e training, guidance and criticism o f the highest order throughout this period. It is w idely acknowledged, am ong those w ho have w orked under his direction, that he has the g ift o f arousing in his students a zest
T
xxxiii
© © O 0 o
© © 0 © © © ©
©
© G €> ©
© © ©
© © G C O
O O
c
r iv ^ ^ o O H ?*)
O (T' O O O O ô ô 'H C) o
€ ©
§
xxxiv
Acknowledgements
for anthropological theory w hich enables them to keep on w orking w ith a w ill in spite o f the m any and often form idable difficulties and discomforts o f field-w ork. I am deeply grateful to Professor Gluckman for his inspiring teaching, keen criticism and generous friendship. Several members o f the Departm ent and the Institute have greatly helped me by reading the manuscript o f this book w ith critical attention. Professor Elizabeth Colson, a form er D irector o f the Institute, made m any constructive comments during the crucial period between spells o f field-w ork. Professor John Barnes read the manuscript at a late stage o f preparation w ith incisive thoroughness. His advice on the layout o f tables, dia grams and genealogies was o f exceptional assistance. I w ould also like to thank D r. A . L . Epstein for m any valuable observa tions on m y general argument. I ow e a special debt o f gratitude to Professor C lyd e M itchell, w ho was D irector o f the Institute w hile I was in the field, fo r his unfailing help and encouragement. H e gave me m y first practical training in field-w ork methods during a happy fort night o f research in Lambaland. His studies o f Y ao village life opened up several fruitful lines o f thought to me. M r. C . M . N . W hite, M .B .E ., for a tim e A cting D irector o f the Rhodes-Livjmgstone Institute, gave me the benefit o f his w ide know ledge o f the western tribes o f N orthern Rhodesia. The influence o f Professor Fortes is w rit large in this book. M any o f his ideas, first encountered during m y student days, had already become part and parcel o f m y thinking w hen I entered the field. M y w ife collaborated actively in all aspects o f m y field-w ork. Her contribution included most o f the photography and much o f the measuring o f gardens and m apping o f villages. In addi tion, she drew the maps for the book. Her comments on the argum ent have helped to shape its form . In short, this book is in a very real sense the product o f collective authorship. V. W . T.
i MAPI NDEMBUAREA
NORTHERN
RHODESIA
MAP 2 DISTRIBUTION OP NDEMBU VILLAGE
CH APTER I H IST O R IC A L A N D E C O L O G IC A L B A C K G R O U N D HIS book is prim arily w hat Professor Fortes w ould call a * histological * study, an attem pt to analyse in close detail the form and functioning o f a sub-system, the village, w ithin a w ider system, the totality o f N dem bu society. B ut som ething must be said, how ever compressed, about the w ider historical and socio-geographical background o f the village. O ur village m icrocosm is a com ponent in a system o f social relations existing at a specific tim e and place. B u t it also has a specific history. B y European notions, this history is b rie f and blended w ith m ythology. Nevertheless, fo r N dem bu it is a history as a record o f successive, interdependent events, m aking an irrever sible process, w hich in their view explains w hat they are and where they are. The M w inilunga Lunda-N dem bu, hereinafter called N dem bu , 1 num bering about seventeen thousand, today inhabit the western portion o f M w inilunga D istrict in the N orth-W estern Province o f N orthern Rhodesia, approxim ately between n ° and i z ° south latitude. B efore European occupation the N dem bu chiefdom o f Kanongesha com prised m ost o f die territory between the W est Lunga, Zam bezi and L u fw iji rivers. N o w the tribal land is crossed b y die international boundaries that demarcate N orthern Rhodesia, A ngola and the B elgian C on go. Kanongesha’s cbiefdom is virtually bisected b y the Angola-R hodesia boundary, and on either side o f the boundary there is a ch ief w h o calls m m self
T
1 A num ber o f Southern Lunda groups call themselves Ndem bu , including the inhabitants o f the chiefHoms o f Kanongesha, Ishinde and Kazem be Mutanda (or Xzaizai). M y inform ants say that N dem bu was originally the name o f a river at which these chiefs and their follow ers stayed together for several years after their departure from M wantiyanvwa’s capital (see p. z). I em ploy the term Ndem bu throughout this book to distinguish Kanonge$ha*s Lunda from the Lunda-K osa (or A kosa) o f C h ief M usokantanda. This ch ief lives in the Belgian C ongo near M usonoi but has nom inated Sailunga as his representative in Northern Rhodesia. The Kosa occupy M winilunga D is trict to the east on the W est Lunga R iver. Government refer to C h ief Kanongesha*s Lunda as * N dem bu * and to the Kosa group as ‘ Lunda * in official publications.
TTrK.s
O s'"-'. ■(
€) ( ) f. ; C; C'"; O G, O O O O c
G .O O
C G
C.)
£
>.
U C) O G
2
Schism and Continuity in an A frican Society
Kanongesha and is recognized as such b y the colonial pow er. This book is based on research am ong the Rhodesian Ndem bu and m ay not apply to Ndem bu living in Angola or the C ongo. History and Traditions1 T he Ndem bu, like their Kosa neighbours to the east, claim to have com e as invaders from the N orthern Lunda 2 empire o f M w antiyanvw a between the Kasai and upper Bushimaie rivers, and to have conquered or received the submission o f small scattered groups o f indigenous M bw ela or Lukolw e. The inva sion appears to have taken place before the beginning o f the eighteenth century. According to Dias de Carvalho 3 the de parture o f Kanongesha and his follow ers from Luunda took place during the reign o f the fifth M w antiyanvw a, Ianvo N oeji, whose reign Verhulpen places between 1640 and 1660.4 Accord in g to the traditions o f both N dem bu and Kosa, Kanongesha*« m igration occurred at about the same tim e as the migrations o f other Lunda leaders such as : (1) Kazem be Mutanda, w ho established a chiefdom ju st to the north o f Kanongesha*s; (2) Musokantanda, w ho founded a chiefHom to the east o f the N dem bu in w hat is now M usonoi D istrict in the Katanga, and in eastern M w inilunga ; and (3) Ishinde, whose chiefdom , like that o f Kanongesha’s, was later divided between Angola and N orthern Rhodesia where it occupies the eastern part o f Balovale D istrict. For m any years these chiefdoms, like that o f Kazem be on the Luapula, described b y Lacerda, Gam itto and others, w ere tributary to M w antiyanvw a, and even today visits are made by 1 A considerable literature exists on Lunda history, much o f it represented in the bibliography attached to Miss McCulloch’s T he Southern Lunda and R elated Peoples, Ethnographic Survey o f Africa, ed. Daryll Forde (1951). In addition to works cited below in the text, Lewis Gann {‘ The End o f the Slave Tirade in British Central Africa : 1889—1912. T he Rhades-Livingstone Journal, xvi (1954), p. 36) makes some useful comments on the Angolan slave trade. See also my annotated translation o f selected passages from Dias de Carvalho, €A Lunda Love Story and its Consequences in T he Rhodes-Livingstone Journal, xix (1955), for the Northern Lunda traditions. 8 The land o f Mwantiyanvwa is called Luunda or Luwunda , and his people A luunda , by Ndembu. 3 Dias de Carvalho, H. A., ExpediçSo Portuguesa ao Muatianvua (1890), p- 5 4 *. 4 Verhulpen, E., Baluha et Balubaisés de Katanga (1936).
H istorical and Ecological Background
3
Southern Lunda chiefs or their representatives to the potentate’s capital on the Lulua river w hen a new M w antiyanvw a succeeds.1 W hen a Southern Lunda ch ief succeeds to office, confirm ation is sought from M w antiyanvw a. "When the British Govern ment sought to abolish the N dem bu chieftainship o f Nyakaseya in 1947, die incum bent hastened to M w antiyanvw a w ho w rote a letter to the W estern Provincial Com m issioner on his behalf. As a result he was reinstated. According to W h ite,2 the Southern Lunda largely retained their own language and to a greater degree their social system. One reason for this seems to have been that they migrated into a broken thinly-populated country where they had less chance o f being absorbed by the people whom they found already there. The latter were o f Lukolwe stock as evidenced by the names o f rivers which they have le ft.. . . Indeed the Lukolwe, whose remnants today are a very primitive people, may have been one factor which was responsible for the failure o f the Lunda to show the same degree o f enterprise as their neighbours. . . . Lunda also suffered continuously from the incursions o f their Lwena and Chokwe neighbours which may indicate that they were a later migration, at a disadvantage with their earlier established neighbours, which prevented their expansion and development. Hence the Chokwe and the Lwena were able to live in more open, flatter country* whilst the Lunda took refuge in broken hilly country between, the Zambezi and Kabompo headwaters where they lived in small scattered communities. There are today in the M w inilunga N dem bu region some tw enty-six villages, the residential cores o f w hich claim to be o f ‘ M bw ela * origin. T he term ‘ M bw ela * appears to have been w idely applied b y in-com ing groups o f Lunda stock to the m ore prim itive Bantu peoples, such as L ukolw e, w hom they encoun tered to the south and w est o f their hom eland. Ndem bu today use the term kabeta Kambwela to denote * the south \ Sometimes they group M bw ela and K aonde together as having a cultural affinity. T h e M w inilunga group o f putative M bw ela origin are * As in 1951, when Mbaka succeeded. Both Kanongeshas sent represen tatives, Musokantanda and Kazembe Mutanda went in person, and several Native Authority Sub-Chiefs such as Ikdtenge, Nyakaseya and Mwininyilamba visited the new Mwantiyanvwa, each with his entourage who were given food by the potentate. 2 "White,C. M . N., ‘ The Balovale Peoples and their Historical Background*, T he Rhodes-Livingstone Journal, viii (1949), pp. 35-6.
4
Schism and Continuity in an A frican Society
generally know n as *K aw iku *, after the name o f a plain to the north-east o f M w inilunga Bom a, w here they are first said to have been encountered by Kanongesha’s invaders. According to Ndem bu and K aw iku traditions they resisted the invaders for m any years before they submitted. In A ngola, the Ndem bu apply the name * H um bu * to a population o f M bw ela origin in the south, and the senior Hum bu headman Kafwana plays an im portant role in the Angolan Kanongesha’s installation ritual. T oday K aw iku and Hum bu are culturally and linguistically almost indistinguishable from N dem bu, but on occasion stm maintain that they are * the owners o f the country \ Ndem bu did not establish themselves as a higher caste or class above the K aw iku. B oth groups interm arry freely and live at the same econom ic level. Ndem bu, how ever, consider K aw iku to be turbulent, clownish and discourteous people and jib e at their selfish and quarrelsome behaviour. K aw iku take delight in fostering these beliefs b y jo k in g about their ow n uncouthness. It is doubtful whether Ndem bu have retained the social system practised in their hom eland to the extent indicated by W hite. Several w ell-inform ed accounts o f M w antiyanvw a’s state o f Luunda exist w hich give the picture o f a h ighly centralized political system.1 In the royal capital a host o f nobles and officials w ere perm anently stationed, supported b y the tribute that Sow ed in, at the height o f Lunda pow er, from chiefdoms as w idely separated from one another as Kapenda-ka-M ulem ba to the west and Kazem be on the Luapula to the east, a distance o f m ore than seven hundred miles. N o t only tribute but trade contributed to the surplus w hich maintained the Lunda nobility. M w antiyanvw a had trade agreements w ith a num ber o f Ovim bundu chiefs w ho despatched caravans into the interior bearing trade goods in exchange for slaves, ivo ry and, later, rubber and bees w ax.2 Capello and Ivens 3 w rote in 1882, when the empire was already on the wane, ‘ at a short distance from the mu-sumba [Musumba, “ state capital ” ] are established vast markets, true 1 Dias de Carvalho ; Buchner, M., ‘ Das Reich des Mwata Yamvo und seine Nachbarländer ’, Deutsche Geographische Blätter, 1 (1883) ; Pogge, P., Im R eiche des M wata Yamvo (1880). 2 Childs, G. M., Umbundu K inship and Character (1949), p. 205. 3 Capello, H., and Ivens, R ., From Benguella to the ’T erritory fo Yacca, trans lated by A. Elwes, (1882), vol. 1, p. 3 8 9 .
H istorical and Ecological Background
5
bazaars containing straight lanes or streets w here flour o f various kinds, ginguba, palm -oil, fresh and dried m eat, massambaia, salt, tobacco, m aluvo (palm w ine), mabellas, and other articles are displayed, and are bartered for merchandise, such as blue and red baize, cottons, printed calico, large w h ite and sm all red beads, pow der, arms ana bracelets \ Verhulpen m entions h o w the razzias o f the Lunda and Luba states sw ept die C entral African area fo r slaves to sell to the Bangala and O vim bundu m iddle-m en o f the Portuguese in exchange for guns and trade goods. T o maintain this system o f tribute, trade and pillage, M w antiyanvw a had at his disposal a pow erful arm y quartered in the capital and divided into advance, flank and rear guards, each arm y corps having a reserve o f liaison personnel and scouts. Capello and Ivens 12 m ention 4 wars o f exterm ination * w aged by M w antiyanvw a against * recalcitrant tributaries * and dynastic rivals. T h e N dem bu, on the other hand, seem never to have had any thing like the same degree o f political centralization as their Lunda ancestors.® It is said that Kanongesha, ‘ a son o f M w an tiyanvw a *, was accompanied, o r follow ed shortly afterwards, by tw elve senior headmen drawn from various Lunda noble families, w hen he first entered the present M w inilunga area.3 Each o f these was accom panied b y relatives, m ale and fem ale, and a retinue o f warriors. It took at least a generation o f inter m ittent raiding to subdue the indigenous peoples, and the invaders w ere scattered far and w ide over the country in pursuit o f them . Eventually, so the tradition goes, the Southern Lunda chiefs, together w ith some neighbouring Lw ena chiefs such as 1 Capello and. Ivens, vol. i, pp. 390-2. 2 For instance, Livingstone writes in his journal for 1854 that in the Lunda area o f Kazem.be (immediately north o f Kanongesha’s Ndembu), * the frequent instances which occur o f people changing from one part o f the country to another show that the great chiefs possess only a lim ited pow er * (my italics)— M issionary Travels and Researches in Southern jdfrica (1857),
P* 305.
s Campbell, D. {W andering in C entral A frica (1929), p. 52 et seq.) is, I think, substantially correct when he attributes the * colonial expansion * o f the Lunda— o f which the Ndembu migration was a part— to * wars o f jealousy and internecine strifes that scattered leaders, who, with ambitious followings went farther and farther afield . . . nibbling into the weak frontiers o f other tribes \
p
o o o o e?
o o o c
o o o o o
e 0
o o o o. o o O o C) cC)
o u Q © €)
6
Schism and Continuity in an A frican Society
Chinyam a and N yakatolu, whose forbears had com e from M w atiriyanvw a in an earlier w ave o f m igration, m et together and divided up the conquered M bw ela and Lukolw e land between them , using the larger rivers as boundaries. Kanongesha re ceived as his share the land between the Zam bezi, W est Lunga and L ufw iji, w ith Ishinde as his southern neighbour, M usokantanda to the east, Kazem be Mutanda to the north, and the Lwena dhieftainess, N yakatolu, to the west. Kanongesha recognized the claims o f his tw elve headmen to certain tracts o f land w hich they had conquered, and later allocated further areas to certain o f his sons and dose matrilineal kin. H e retained the largest share, including m ost o f the M w inilunga N dem bu region south o f w hat is n ow the north-western pedicle o f M w inilunga Dis trict, and the Angolan region around the L ovw a river, to be under his direct control. T h e headmen and relatives to w hom he had given areas w ere supposed to send him tribute o f various kinds ; unrelated headmen had to send him tributary w ives from their ow n matriUneages. H e him self had to send a portion o f the tribute thus obtained to M w antiyanvw a, a custom w hich was discontinued several years before die C h okw e invasion o f that potentate's territories in 1885. A fter the division o f the land, N dem bu tradition reports a long period o f peace, during which there w ere no wars against neighbouring tribes. In this period, how ever, it w ould appear that the m inor and m inim al political units, the senior headmen's areas and the villages w ithin them, became increasingly autono mous and independent o f Kanongesha’s control. Traditions exist o f feuds during this period between senior headmen and even between individual vmages and vicinages. Several factors are responsible for this regression towards local autonom y. A m ong these m ay be reckoned isolation from Luunda w ith its relatively high level o f culture and social organization ; die small num ber o f N dem bu invaders and M bw ela invaded alike, scattered over a w ide region o f forested terrain ; and the lo w level and rudim entary character o f econom ic production associated w ith small shifting settlements. Villages were not perm anendy anchored to specific tracts o f land but m oved across the face o f the Ndem bu region w ith relative freedom . H igh spatial m obility contributed pow erfully to the considerable political autonom y o f the village.
H istorical and E cological Background
7
T he decentralization and m obility o f Kanongesha’s N dem bu were accelerated b y the slave-trading and slave-raiding w hich reached their height in this region in the second h a lf o f the nine teenth century. O vim bundu slave-traders m ade trade pacts w ith Kanongesha and im portant senior headm en such as Ikelenge, in north-w est M w inilunga,1 and these pow erful men raided their ow n tribesmen and sold them into slavery in return for guns and d oth . Later, C h okw e and Lw ena, heavily armed w ith Portuguese muskets, attacked the N dem bu, w h o at first presented no united front to the slave-raiders and w ere defeated village b y village, the unfortunate inhabitants being marched o ff to jo in the long lines o f slave caravans that converged on N yakatolu’s Lwena capital from die interior. T h e N dem bu w ere rallied, not b y a ch ie f but b y a com m oner, Chipeng’e, w ho went round the country urging diem to sink their private feuds and jo in d y hurl out the invaders. H e inflicted several defeats on a large w ar-party o f C h okw e and Lw ena w h o w ere under the leadership o f the Lwena ch ief K angom be (mentioned b y Dan C raw ford).2 A fter this, the raiders w ith drew into Angola. B ut the N dem bu w ere n ot left lo n g in peace, for in 1906, follow ing die announcement o f the K in g o f Italy’s arbitration o f 1905, w hich settled a long-standing dispute between Great Britain and Portugal over the exact position o f the western boundary o f N . Rhodesia, officers o f the British South Africa Com pany, accompanied by native police, commenced the administration o f the new M w inilunga D istrict. As a result o f the harshness o f an early administrator, there was a mass m igration o f Ndem bu out o f the Com pany’s territory into the Belgian Congo and A ngola. T he officer in question was discharged, but hardly had the people returned than there was a new exodus — in 1913, w hen taxation was first introduced. The senior head man Ikelenge, later to becom e a S u b -C h ief under the N ative Authority, was the o n ly im portant N dem bu to remain, arguing that die taxes w ould produce benefits fo r the people in the form o f roads, services and peaceful governm ent. Chipeng’e, now an 1 Gann, L. (p. 33), points out that the Angolan slave dealers, unlike the Arabs, never established political rule o f their own, but preferred to work through native chiefs. 2 Crawford, E>., Thinking B lack (1912), p. 116; Tilsley, G. E., D an Craw fo rd , M issionary and Pioneer in Central A frica (1929), pp. 104-5.
8
Schism and Continuity in an A frican Society
old man, encouraged the flight, realizing that the Europeans w ere too strong to eject by force. Gradually the population trickled back and w ere regrouped into villages b y die N ative Com m issioner, the leaders o f the malcontents being com pelled to build near the Bom a (a w idely applied term for the head quarters o f the British D istrict Adm inistration) where an eye could be kept on them. In 1906 Kalene M ission was founded b y D r. W alter Fisher o f the Christian M ission in M any Lands at Kalene H ill in the northern pedicle, and a hospital and school for Africans w ere built soon afterwards. This event was to have im portant effects on the distribution o f N dem bu villages and ultim ately on their internal structure. For a time N dem bu regarded Kalene H ill as an alternative administrative centre and headmen took tribute to D r. Fisher as a mark o f vassalage. A fter each exodus from the Bom a area, m any headmen returned to M w inilunga to build near the M ission, and others came from A ngola w here illicit slave-trading continued into the tw entieth century. A trading store was early established at Kalene ; and in 1921, D r. Fisher’s son, fFolliot Fisher, started H iliw ood Farm, near Ikelenge’s v il lage, tw elve miles from Kalene, and a store was eventually added to this. Som e years later Sakeji School, a boarding school for the children o f European missionaries o f various Protestant denom inations, was founded close to H iliw ood Farm. In 1952 nearly sixty children w ere in attendance at the school. Altogether, in 1952, there w ere about a hundred Europeans, including teachers o f European and African school children, doctors, state registered nurses, missionaries, and farm managers at H iliw ood, in the >edicle area. Each o f the three European centres em ployed a arge A frican staff, including hospital attendants, farm hands, road labourers, school teachers, clerks and dom estic servants. In addition to these, others w ere em ployed as store capitaos (foremen) and maintenance men by European com m ercial firm s, whose headquarters w ere on the N orthern Rhodesian Copperbelt. Each African in regular w age em ploym ent, whether he lived w ithin the precincts o f the European settlements or in an adjacent N dem bu village, tended to becom e surrounded by kinsmen w h o sought his patronage. Three senior headmen in the pedicle became N ative Authority
{
'£»
H istorical and Ecological Background
9
Sub-Chiefs under Governm ent, and each o f them is surrounded by an entourage o f paid petty officials : clerks, kapasus (N ative A uthority Messengers), assessors, and the like. A t each SubC h ief *s capital village there is a L ow er School under the super vision o f the Christian M ission in M any Lands at Kalene H ill. A t one, that o f C h ie f Ikelenge, there is a dispensary, a PostOffice, and a N ative A uthority M arket. T he tendency to cluster round a European settlem ent in v il lages w hich began in D r. Fisher’s tim e has continued, and today each European centre, and indeed each N ative A uth ority head quarters, has becom e the focus o f a large A frican population. N o t only those already em ployed, but those w h o seek w age em ploym ent, com e to settle in villages and other local units in the pedicle, either b y attaching themselves to already existing villages w here they nave kin, or b y com ing in kinship groups and founding n ew residential units. T h e fact that the soil in the pedicle is considered by N dem bu to be especially suitable fo r cassava cultivation is a further strong inducem ent to settle there. Paym ent o f domestic servants and labourers and the creation o f opportunities for buying trade goods and selling native produce in the northern pedicle led to further increases in its population. Game was virtually exterm inated in the pedicle b y Africans w ith m uzzle-loaders and Europeans w ith rifles.' Local trade in dried meat and fish developed w ith Angolan and C on go Africans w h o exchanged these com m odities fo r cash and store goods. This area was the first in M w inilunga to grow crops fo r cash as w ell as for subsistence. T oday this priority in developm ent is re flected in the existence o f a num ber o f A frican traders, especially in Ikelenge’s area near M illw ood Farm, w h o bu y cassava m eal from the villagers fo r re-sale to the B om a and to B elgian traders in the C on go. T h e pedicle area exhibits the highest degree o f disruption o f the traditional social system seen in the Ndem bu region. M any im m igrants, including some Lw ena and O vim bundu, have entered the area since 1920, often to seek asylum from the Portuguese w ho. had put dow n several rebellions o f the interior tribes. T h e social com position o f the pedicle has thus become extrem ely heterogeneous and its position astride labour routes from the adjacent colonial territories to the Rhodesian lineof-rail have m ade it subject to continual stim uli from alien
O 0 0 © Q O
o
© © © 0
©
© ©
© © o ©
©
© © © ©
c? ©
© © o c ©
C
©
io f
f"\
r;-;. r- \
:
Schism and Continuity in an A frican Society
cultures. In die last few years profound changes have occurred in die residential structure in this area : the m ost noteworthy has been the breakdown o f traditional villages into small units headed by younger men w ho participate in the encroaching cash econom y. These units know n as mafwami (from the English * farms ’) increased enorm ously in num ber between m y first and second periods o f field-w ork. B u t since change is not proceeding evenly either in tempo o r direction in all areas alike, I found it possible to observe m any villages w hich retained the indigenous structure, even in the pedicle, and w hose inhabitants still tended to conform to traditional norm s o f behaviour. T h e traditional village was a circle o f pole-and-m ud huts typically containing a core o f m atrilineally related kin under the leadership o f a member o f the senior genealogical generation chosen b y the villagers. The ifwami or * farm * consists o f one or m ore Kim berley-brick houses bordered b y a few mud huts and it is occupied b y the form head, his elementary fam ily and a small fringe o f kin and unrelated persons. In the pedicle near European settlements and N ative A uthority Sub-Chiefs’ capital villages a ribbon-developm ent o f forms has begun along the Governm ent m otor roads. In this book I lay m ajor stress on the problems posed b y the structure o f traditional villages and leave over for a further study the social organization o f the form. B u t this sudden m ultiplication o f forms in a few years was only the out w ard manifestation o f changes in values and o f conflicts between old and new principles o f social organization, w hich had been goin g on w ith gathering intensity w ithin the fram ew ork o f the traditional village for some time. I have tried to indicate some o f the effects o f social and cultural change on the rate and type o f village fission at different periods in this century in Chapters Four to Seven. B ut until about 1950 the m ajor effect o f economic and political changes was to increase the rate and not to alter significantly the type o f fission. The basic unit o f fission was a maternal descent group. T oday there are signs that the elem entary fam ily is becoming the typical unit o f fission and that the links binding together uterine siblings and matrilineal parallel cousins are wearing thin. I m ention these tendencies but do not discuss them in detail. In the group o f villages I selected fo r particular study (most o f them in the area o f the deposed N ative Authority Sub-C hief M ukang’ala), the effects
H istorical and Ecological Background
II
o f change on the social structure w ere visible, b u t they w ere far less pronounced than in die pedicle. W hen Indirect R u le was introduced in 1930, the N orthern Rhodesian claim ant to the Kanongesha chieftainship was given the status o f Senior C hief, and six senior headmen w ere made Sub-Chiefs.1 These w ere : C hibw ika, Ikelenge (or Ikelengi), Nyakaseya, M w ininyilam ba, M ukang’ala, and N tam bu Lukonkesha. C hibw ika— like Kanongesha reduplicated in A ngola— is the title given to the nom inated successor o f Kanongesha. His area lies to the south o f Kanongesha^ w ith w h ich it is contiguous. Ikelenge is descended from an official called Kalula w ho accom panied the first Kanongesha. A t the installation ritual o f a Kanongesha his duty is to rem ove the lion and leopard skins from the chiefly chair. T oday his area lies in the south-east o f the pedicle.2 T o his west is the area o f M w ininyilam ba whose ancestor was the Ifwota or * scout \ the leader o f a w ar-party w hich went in advance o f Kanongesha’s forces during the invasion. Nyakaseya C she w ho pours out i) is a tide m ade for the first Kanongesha’s senior w ife (mwadyi) whose duty was to pour out beer when the ch ie f offered it to his guests. T o d ay Nyakaseya is a male Su b -C h ief in the north o f the pedicle. Mukang*ala*s tide was created b y the second Kanongesha, Kabanda, for his son w ho subdued the K aw iku. H is village was situated fairly near the K aw iku Plain in order that he m ight prevent rebellion. The senior K aw iku headman N sang’anyi paid him tribute and 1 Actually, Government Had given Je fa cto recognition to these headmen as * Sub-Chiefs * long before 1930. C . M . N . "White (personal communica tion) prefers to call mem ' Sub-Chiefs * rather than * senior headmen ’ in the pre-European organization. White writes that * they have always had royal insignia, their titles go back to the early days o f the Lunda chiefs' coming, and the Lunda themselves in any interpretation, o f what the vernacular means, regard them as Sub-Chief; *. I use different terms for their traditional and modem positions because I believe, as I bave argued in Chapter Eleven, pp. 323-7, that their status rested on different criteria before British rule. Jh the past they enjoyed moral and ritual pre-eminence, but did not neces sarily possess effective political authority. Today, since they are supported by the Administration, they have greater powers o f coercion and are organized into a hierarchy o f authority, but some or them, whose status rests on insecure or inadequate traditional foundations, do not command the respect senior headmen formerly possessed. 2 See Map 2.
12
Schism and Continuity in an A frican Society
acted, as his Chivwikankanu, a ritual office w hich gave him an im portant role at the installation ritual o f a M ukang’ala. Kafwana, the senior headman o f the Hirnibu in A ngola, was Kanongesha’s Chivwikankanu. Kafwana’s office, w hich means literally * he w ho puts on the lukanu *, involved the investing o f a new Kanongesha w ith the lukanu bracelet o f human sinews and portions o f genitalia, the supreme sym bol ofLrnida chieftainship, and thereafter the periodical renewal o f its ritual pow ers by washing it w ith special medicines. The other Sub-C h ief recog nized b y Governm ent in 1930 was N tam bu Lukonkesha. N tam bu was not tributary to Kanongesha in the past but paid tribute directly to M w antiyanvw a. Although his territory was m uch smaller than Kanongesha’s his status was ju st as high. The father o f the present Angolan Kanongesha was C h ie f N tam bu and the father o f N tam bu was a form er Kanongesha. Tradition ally, N tam bu had the right to sit on a chair in the presence o f Kanongesha, but Kanongesha’s senior headmen, som e o f w hom are n o w Sub-Chiefs, had to sit on the ground. In 1936 N tam bu, and in 1947 M ukang’ala, had recognition w ithdraw n from them . M cC ulloch, on the basis o f inform ation from W hite, w rites : 1 * Form erly these Sub-Chiefdom s seem to have been quite firm ly established, but today political bound aries have been greatly altered and in the interests o f centralization recognition has been w ithdraw n from “ redundant ” Chiefs and several areas have been amalgamated.* T he areas o f both 4redundant * chiefs have been incorporated in Kanongesha’s ow n area. Few o f the m odem Sub-Chief! w ere in fa c t4firm ly established O f the tw elve headmen w ho came originally w ith Kanongesha, o r follow ed him shortly afterwards, some are n o w ordinary headmen. These include Chibw akata, the present incum bent o f the office o f Kamhanji (* war-leader ’ and * executioner ’), whose village is a few m iles from M w inilunga Bom a, Kabtm g’u and N yachilesya in A ngola, and M wanta waLuunda, Nyachikanda, K abuya and C hota in M w inilunga District. B y an odd quirk o f fortune, K akom a, the 4war-leader * o f M usokantanda (one o f the senior m igrant c h ie f from Luunda), has been appointed a S u b -C h ie f; and Musokantanda’s * em blem -purifier (Chiu1 McCulloch, p. 26.
H istorical and Ecological Background
;
13
wikankanu) Sailim ga, lias been appointed Senior C h ie f o f the Akosa. Su b -C h ief M w ininyilam ba’s original area was on the east o f the Lunga R ive r near the Kasanjiku Stream, in what is now * Sailunga’s A rea*. Tow ards the end o f the nineteenth century the incum bent o f the tim e fled after a quarrel from this area and settled in w hat is n o w the pedicle. T h e senior headman there was a perpetual * son * 1 o f Kanongesha, called K afw eku, w ho collected tribute from village headmen fo r Kanongesha. W hen Authorities w ere appointed in 1930, M w ininyilam ba, although a refugee from another area, was appointed Sub-C h ief o f K afw eku’s area. N dem bu say this was done because the M wm invilam ba o f the tim e was a forceful character w ith a large village, whereas the K afw eku was a tim id m an w ho avoided meeting Europeans. It is doubtful in fact whether the headmen recognized as SubChiefs b y Governm ent w ere ever regarded as ‘ ch iefs*2 b y N dem bu w ith the exception o f N tam bu Lukoukesha. T h ey were sim ply village headmen w h o possessed im portant historical titles. Certain o f them claim to have received their insignia from M w antiyanvw a. I f they w ere strong m en w ith large villages Kanongesha m ight have entrusted them w ith certain tasks. Finally, i f he w ere n ot related to them b y ‘ perpetual kinship *, 1 The son o f a Lunda chief who has been granted his own area and at the same time excluded from the succession to his father’s title is called M wana awuta or M wanawuta. Mufcmg’ala is also a M wanawuta o f Kanongesha. I. G. Cunnison (A Socia l Study o f a Bantu People , unpublished thesis, Oxford University, p. 116), defines perpetual relationship as a relationship between the titles o f headmen and chiefs winch always remains constant irrespective o f the actual genealogical relationship o f the individuals holding titles, and which expresses die political relationships between the different chiefs and headmen. Cunnison has recently published a paper on ‘ Perpetual Kinship: a political insti tution o f the Luapula peoples ’ in T h e Rhodes-LivingstoneJournal, xx, pp. 28-48. 2 All were entitled to the respect due to their historical status as founders o f the Ndembu people and all had certain insignia such as wooden slit-gongs, xylophones, bangles, short swords, etc. Some utilized the initial advantage given them by this status to dominate small areas. Harding (In Rem otest Barotseland (1904), p. 146) describes Ikelenge (* jEakaling *) as a * very important chief under Kanongesha \ His capital village was then at Kalene Hill, which was an important centre in the slave-trade with the Ovimbundu. It is likely that the Ikelenge headmen built up wealth and influence by trading in slaves. Harding published his book in 1895. He visited Ikelenge in the x88o*s when Ovimbundu slave-traders still dealt with Ndembu headmen.
14
Schism and Continuity in an A frican Society
he m ight m arry into their families. B ut although the boundaries o f their areas have been rigid ly demarcated b y the Adm inistration it appears as i f m any o f them ranged w id ely over the N dem bu region in the past. M ukang’ala, for exam ple, during part o f the nineteenth century made his capital village w ell to die east o f the Lunga. I m ention below the spatial m obility o f other im portant villages, including Chibwakata, and in the last chapter discuss in some detail the role o f the senior headmen in the indigenous social system. T he Kanongesha in A ngola is regarded b y m ost N dem bu in M w inilunga D istrict as their authentic chief, since he possesses the lukanu given b y M w antiyanvw a to Kanongesha Kabanda (see p. 320). T he first Kanongesha to enter M w inilunga D istrict, M ulum bi, took this lukanu w ith him w hen he left A ngola in the early i92o’s. M eanw hile a new Kanongesha was elected b y the senior headman o f the Angolan Ndem bu w ho blamed M ulum bi for leaving the Isenga R iver Island area w here the graves o f form er Kanongeshas w ere located, ha 1927 the A ngolan Kanongesha, Sakayoli C hifiiw u, conspired w ith Kafwana, the Angolan H um bu senior headman, w ho then obtained the lukanu from M ulum bi on the pretext that it had been defiled and had lost its virtue. Kafwana told M ulum bi that he w ould wash it w ith m edicine at the cross-roads to the west o f die capital village at m idnight to cleanse it. H e took the lukanu and w ent straight to A ngola w here he gave it to Sakayoli. M ulum bi tried to obtain a new lukanu from M w antiyanvw a, but according to m ost N dem bu was not given an audience b y him . H e returned indeed w ith a new lukanu, but this is said to have been made b y a Luunda court official,1 n ot b y M wantiyanvwa. T he form al structure o f the Ndem bu political system under the N ative Authorities Ordinance and the N ative Courts Ordinance presents a rigid hierarchical appearance quite foreign to the indigenous structure. In the past there was o n ly a single N dem bu 4ch ief ’ (mwanta wampata) confirm ed in his office and tributary to M w antiyanvw a, and bis effective authority was confined to his ow n im m ediate area. The first senior headmen (ayilolu) 2 w ere initially allocated tracts o f bush in w hich they had rights 1 Chipawu. 2 For a discussion o f the meaning o f this term see p. 323.
H istorical and Ecological Background
15
to hunt and cultivate and to exact tribute from incom ers w h o wished to hunt and cultivate there. In return for these rights they had to g ive Kanongesha tribute w hen he asked fo r it and tributary w ives from their o w n matrilmeages. B u t w ith the continual m ovem ent o f villages from site to site, and w ith the frequency o f village fission, it became extrem ely difficult fo r senior headmen to exert political authority over the inhabitants o f their areas. T h e basic econom y o f the N dem bu, cassavagrowing and hunting, did not com pel any individual or group to remain in one specific area. In the period o f slave-trading and -raiding, ties between groups and localities must have been further loosened both b y internecine raiding and b y the instant necessity to be ready to flee before an external enem y. In such an unstable and highly m obile society Kanongesha could not enforce his rule on senior headmen and the latter could not exact obedience from the villagers in their areas. Villages resorted directly to force in the form o f self-help in defence o f the rights o f individuals and groups. I f a ch ie f or a senior headman managed to establish some degree o f centralization in his area this was due m ore to personal qualities than to his social position. U nlike such chiefs as Kazembe o f the Luapula V alley or Chitim ukulu o f the Bem ba, the Kanongeshas had no m ilitary forces perm anently stationed at their capitals. T h ey had to rely on possessing rather larger followings o f kin and slaves than other headmen, even to enjoy local political pre-em inence. In Chapter Eleven I argue that the position o f the Kanongesha was ritually rather than politically important and that he was respected rather than obeyed. The position o f the C h ie f and o f those senior headmen w h o have been appointed as Sub-Chiefs has been considerably strengthened b y Governm ent. M cC ulloch, follow ing W hite, writes th a t1 today all civil cases and lesser criminal cases can be tried by the court o f the Native Authority (the Chief's court) as constituted by the Native Courts Ordinance. The C h ief is officially supposed to preside at trials, but in practice cases are often heard only by assessors. From this court, appeal lies in the first instance to a special Appeal Court o f the Native Authority, then to the District Commissioner, and 1 McCulloch, pp. 2 6 -7 , see also p. 71, based on MSS. and personal com munications.
16
Schism and Continuity in an African Society
finally to the High Court o f Northern Rhodesia. More serious crimes are taken directly to the District Commissioner, who may remit to the Native Authority any cases which he considers to lie under his jurisdiction. Imprisonment can be given for certain criminal offences by courts o f the Native Authorities, but this happens comparatively rarely, and the usual penal sanction, is a fine. Both fines and court fees are paid to the Native Treasury. In fact many minor cases tend to be setded out o f court, but the Lunda Chiefs discourage this. . . . Cases concerning land rarely come into the Native Authority courts, but are usually settled by village heads. In the past there do not appear to have been such centralized judicial institutions. Kanongesha and his im portant headmen— and their im portance it m ust be remem bered varied w ith personal and other adventitious factors— m ight on application test a diviner’s attribution o f w itchcraft or sorcery to a person b y recourse to their ow n poison oracles. O ccasionally they m ight officiate in the ritual o f Musolu to bring on belated rains and confer fertility on crops, animals and men. B ut in general m ost disputes w ere settled at the level o f the vicinage (chitungili). M en skilled in the advocacy o f cases, know n as mahaku, o r in judgm ent, know n as akwakusompesha, w ere jo in tly invited by a plaintiff or his representative and a defendant or his represent ative to discuss the case and give their verdict. Such m en were often, but need not necessarily have been, headmen. I f there w ere a senior headman (an im portant chilolu) in the vicinage most cases in. that vicinage w ould be settled in his village. Cases w ere seldom brought to Kanongesha’s court on appeal. Kanon gesha in fact was politically hardly m ore than a senior headman w rit large and his effective authority was confined to the few vicinages o f his im m ediate area. T he N ative A uth ority and Sub-Chiefs are allow ed to legislate under the guidance o f the Adm inistration. B ut there is m uch passive resistance against the implementation o f N ative A uthority Rules and Orders by villagers w ho resent the political powers vested in the N ative Authority. Apart from those areas in the pedicle nearest to European setdements and Sub-Chiefs’ capital villages, m any features o f the indigenous social system have tended to persist in much o f western M w inilunga, especially at the village and vicinage levels. This is probably due to a series o f factors, including the thinness o f
H istorical and Ecological Background
17
European settlem ent in the D istrict, the distance from urban centres, the inadequacy o f roads until recently, and the high cost o f transport. little stimulus has been, given to agriculture and consequently the traditional system o f subsistence has not been greatly m odified. Since this m ode o f subsistence was congruent with the pre-European social organization the effects o f m odem changes on the M w inilunga N dem bu have, until the post-w ar period, been m uch less radical and far-reaching than in other parts o f N orthern Rhodesia, such as the N orthern Province. In the last few years, how ever, changes brought about b y the grow in g participation o f N dem bu in the Rhodesian cash econom y and an increased rate o f labour m igration, have in som e areas, notably the pedicle, drastically reshaped some institutions and destroyed others. In this book I shall consider in greatest detail the m ore conservative area o f ex-S u b-C h ief M ukang’ala w here the effects o f change have been hitherto less prom inent. Social change o f the type w e have been considering is to some extent reflected in T able I w here population densities in the pedicle chiefdoms are compared w ith those in other N dem bu chiefdom s. Population figures w ere kindly made available to m e b y the TABLE I T otai . P opuxation in N dembu A rea N ative authority
Area (sq. iw.)
Population
D ensity (per sq. m .)
6,927 3,<514
3 0 12 *2*55
Pedicle chiefdoms : Ik e le n g e ............................... N y a k a s e y a .......................... Mwminyilamba.....................
230 288 162
Total pedicle chiefdoms .
680
* 1,5 * 9
J6*9l
Other Ndembu chiefdoms : Kanongesha.......................... Chibwika................................
1,098 1,216
3,476 3,3 5 *
3 *r7 2*76
Total other chiefdoms
2,314
6,827
2*92
Total Ndembu _ .....................
2 ,9 9 4
18,346
613
978
6*04
o ó: ©
o o e: o o o
o o o D O €? C) Ó '■ O '
ó fc C í ) 1
t j ' O
18
Schism and Continuity in an African Society
D istrict Com m issioner, M r. R.. C . D ening, and. are based on taxregisters. The areas o f chiefdoms were plotted on a map o f M w inilunga D istrict made b y M r. D ening. Table I shows the m uch higher density o f population in the pedicle, especially in C h ie f Ikelenge’s area,1 some reasons for w hich have been previously considered. M ukang’ala’s area, where m ost o f the field, research for this study was conducted, is included in K anongesha’s area b y Governm ent and must have roughly the same population density as the latter. It is an area o f lo w density, and in M w inilunga D istrict by and large lo w density is associated w ith cultural persistence and higher density w ith culture change. In M ukang ala’s area, and particularly am ong the K aw iku, the traditional econom y is still dominant, including a certain am ount o f hunting and collecting. In the next section o f this chapter m y account o f the N dem bu econom ic system is founded on observa tions I made in that area. Ecology M cC ulloch 2 has summarized the available literature on the physical environm ent and the main features o f the econom y o f the M w inilunga N dem bu. I give below an outline o f her account o f the physical environm ent. The Ndem bu region is part o f the Plateau region o f the north western area o f N orthern Rhodesia,3 a very ancient land surface form ed upon the basement com plex o f Archaean rocks (in tensely folded and metamorphosed gneisses or schists, quartzites, limestones and shales, w ith intrusive granites). A ltitude at M w inilunga station is 4,500 to 5,000 feet. M ost o f the D istrict is fíat w oodland w ith outcrops o f hills at Kalene. The rains usually begin in O ctober but sometimes in September. T h e heaviest rain falls during N ovem ber and Decem ber, the main planting months, and the w et season ends in A pril. D uring the period between 1926 and 1937 annual rainfall at M w inilunga ranged from 47*88 in. to 74*83 in. w ith a mean o f 54*75 inT he mean m axim um temperature is 84o, and the m axim um m onthly temperature remains very constant except in September 1 See also Map 2 , showing high concentration o f villages in the pedicle. 2 McCulloch, pp. 2-4, 14-17. 3 Trapnell, C. G . and Clothier, j . . Ecological Survey o f N orth-W est Rhodesia (i9 3 7 )* •
H istorical and Ecological Background
19
and O ctober, w hen it rises to 89*6°. T he m onthly minima show a greater range o f fluctuation, from 44*6° in June and July, to 6o*8° in January and February. T he year thus falls Into three regimes o f temperature : N ovem ber to A pril, an equable period, during w hich the m axim a and m inim a show a small range ; M ay to August, characterized b y lo w m inim a ; and September and O ctober, w hen high m axim a are registered. In the cold season night frosts sometimes occur. The soils o f M w in ilu n ga1 are m ostly o f the Kalahari contact type, ranging from pale-coloured sandy loams and clay-sand soils to better-class fight-reddish and brow n soils o f similar texture, and representing denuded o r m odified remains o f a greater extension o f the sands. W ith the exception o f redder samples in M w inilunga D istrict they are o f lo w productivity. M ost o f the D istrict is covered w ith Brachystegia woodlands. Thick Cryptosepalum forest occurs in the south o f the Ndem bu region between the K abom po, Lunga and M anyinga rivers. D ue to its situation close to the Z am beri-C ongo D ivid e, M w inilunga District is w ell w atered and in fact is a source-area o f m any rivers, most o f them perennial. T h e Lunga is the o n ly fa irly large river in the region o f the M w inilunga N dem bu. Grassy plains some times surround rivers for some distance from their sources. D ry patches o f grassland are not infrequent. Game was form erly abundant, especially on river plains, and in grassy tracts w here w oodland is stunted : but the gam e has been almost exterm inated in the pedicle o f land in the north-w est o f the D istrict that projects betw een A n gola and the B elgian C ongo. In the rest o f the D istrict w oodland, duiker and bush-buck, and grassland dik-dik, are still quite plentiful. B u t hunting, form erly the dominant m asculine activity, has been considerably restricted by Governm ent and N ative A u th ority rules and orders. M en must obtain a licence to purchase a gun and the num ber o f licences issued every year is small. Perm its m ust be obtained to buy gunpow der or cartridges, and these are granted on ly for lim ited quantities. C ertain types o f traps such as pitfalls and snares are forbidden. There is a system o f Controlled Areas b y w hich the D istrict is divided into First and Second-Class Controlled Areas 1 Trapnell and. Clothier. May be consulted for details and for an excellent map showing the distribution o f soil and vegetation types.
20
Schism and Continuity in an A frican Society
and U ncontrolled Areas. In a First-Class Controlled Area, where gam e is plentiful, even residents need a perm it to hunt. These are given o n ly to know n hunters in the area, and the D istrict Com m issioner has arranged w ith the Gam e Departm ent for the Governm ent C h ie f to have a m ajor hand in control. In Second-Class Controlled Areas, those w ho do not reside there must have perm its to hunt. T he N dem bu region is classified as N ative Trust Land w ith the exception o f some C row n Land near the jun ction o f the Lw akela and W est Lunga rivers, set aside as Forest R eserve in 1952, and the territory occupied b y M ission Stations and b y H illw ood and Mafconchi Farms. Agriculture
N dem bu are shifting hoe cultivators and hunters.1 T heir staple crop is cassava, o f both the sweet (manihot palmata) and the bitter (imanihot utilissima) varieties. Finger-m illet is grow n, m ostly for the m aking o f beer. M aize is cultivated in streamside gardens (matempa) in black alluvial soil and also around villages, in addition a w ide variety o f subsidiary crops are cultivated, including beans, groundnuts, sweet potatoes, yam s, cucurbits, tomatoes, onions, potatoes, cabbages and a num ber o f small relish plants. B u t w ith the exception o f beans and sweet potatoes these are grow n on ly in small quantities. T h e m ain gardens (maha) are those containing cassava and, in the first year o f cultivation, some finger-m illet. In the area w ith w hich I am m ost fam iliar, that o f the deposed Su b-C h ief M ukang’ala im m ediately to the north o f the D istrict Head quarters, the usual practice is to clear high Marquesia or Brachystegia woodland grow ing on red o r orange loams and to heap up the felled branches in irregular piles, occupying anything from an. eleventh to an eighth o f the cleared area. These are then burnt and the resulting ash-patches (mashita) are broadcast w ith finger-m illet. Som e maize, pumpkins and other cucurbits are planted on the ash-patches, but m any mashita m ay not be planted at all i f seed is not available. T he finger-m illet is generally sown in D ecem ber and harvested in M ay and June, w hen the rains are 1 McCulloch, p. 14, writes that * the Southern Lunda were in the past primarily hunters and subsistence cultivators *.
H istorical and Ecological Background
2,1
over. A t the onset o f the next rains, in O ctober, the garden is hoed uj> into mounds and planted w ith cassava cuttings. A t the end o f the follow in g rains, i.e. about eighteen months after planting, some cassava is taken from the garden and thenceforth continuously fo r about tw o-and-a-half years until the crop has been used up. T h e land is abandoned after one o r sometimes tw o plantings o f cassava on it, and the bush is then allow ed to regenerate com pletely, resulting in a fallow period o f about thirty years. W e studied the pattern o f land-holding o f a vicinage in Septem ber 1953, in the neighbourhood o f M w inilongs airfield between four and seven miles from the Bom a 1 along the Bom a-K alene m otor road. T he gardens w ere plotted b y compass and tape b y m y w ife and an African assistant trained in surveying m ethods b y the A gricultural Departm ent. It was found that fo r a total population o f 461 the area under fingerm illet and cassava cultivation totalled 441*1 acres, or *96 o f an acre per person. Streamside gardens under m aize cultivation totalled 52*5 acres, or *11 o f an acre per person. T aking both bush and streamside gardens into account the total land-holding o f the vicinage am ounted to 493*6 acres, or 1*07 acres per person. I showed these figures to M r. Allan, form er D eputy D irector o f Agriculture in N orthern Rhodesia. H e considered it reason able to assume that cassava gardens are cultivated continuously for either four o r five years and then left fo r thirty years to re generate, and that the cultivable percentage o f land suitable fo r the fertility requirements o f cassava under the vegetation-soil conditions prevailing in M w inilunga lay approxim ately between 20-40 per cent o f the land surface. O n this basis he calculated that the critical population o r Im d-carrying capacity for the Ndem bu system o f land Usage probably lay betw een 17-38 persons per square m ile. These figures represent the approxim ate population lim it w hich cannot be exceeded w ithout setting in m otion the process o f land degradation. T h ey w ill be exam ined later in relation to village m agnitude and distribution. The Social Organization o f Gardening In general it m ay be said that tree-foiling and clearing and burning the undergrow th are the w o rk o f men, w hile planting, 1 Headquarters o f the District Administration.
22 22
n
n
O' o f"' % .■
o
' ' t.; X Lfi X.
;X"
€} e .
Schism and Continuity in an A frican Society
weeding and barvesting, save for cassava w hich is planted by both sexes, are wom en’s w ork. Gardens are owned and w orked m dividually and m ay be made wherever suitable land is available. U sually the individual members o f a fam ily m ake adjacent gardens. A polygynist husband’s cassava garden is in the centre o f a clearing and those o f his w ives at either side. A com m on pattern is for a man, his w ife or wives and a favourite sister to w ork contiguous gardens. Brothers and male inatxilineal parallel cousins tend to cultivate in separate blocks o f gardens. A s m ay be seen in M ap 3, the holdings o f a single village m ay be scattered in discrete blocks over a w ide area. T he cutting and clearing o f bush, and later the initial hoeing up o f the cleared ground into mounds, m ay involve a collective w ork-party (chenda) o f kin and neighbours. W hen a man wishes to fell trees and dear bush fo r a fm ger-m illet garden he lets it be know n that he has brewed beer or killed gam e for those par ticipating in a chenda to be held in a few days. Such a chenda is restricted to men and older boys. T he beer is displayed at the beginning o f the w ork w hich commences at dawn and continues until the sun becomes very hot. I f m uch beer has been brewed m uch bush w ill be cleared ; i f little, a m uch smaller area. The productive individualism o f N dem bu finds expression in the grum bling and mutual recriminations over the am ount o f w ork to be done, w hich take place between the sponsor o f the chenda and the volunteers w ho attend. Since m ost adult m en sponsor w orking parties, it means in effect that the m ajority o f men in a vicinage w ork collectively for each other in turn. Each man sponsors a single chenda only, so that when it is over he has to com plete the clearing b y himself. Hoeing parties for first-year cassava gardens m ay be sponsored b y men or wom en and contain members o f both sexes. T he m en proceed in advance o f the w om en hoeing up the rough ground, and the wom en follow , breaking up the soil m ore finely and m aking neat mounds. M en and w om en usually w ork as married couples, although a brother and sister or father and daughter m ay w ork together. W h en the w orking party is over the sponsor subsequently w orks alone or in partnership w ith his or her spouse. Streamside gardens are ow ned and w orked individually b y
H istorical and E cological Background
23
wom en only. Gardens around the village peripheries or in old village sites are ow ned and w orked individually b y members o f both sexes. V illage gardens, in contrast to bush and streamside gardens, m ay be ow ned b y youn g unmarried individuals o f either sex. In the past m arriage was uxorilocal for the first year, in the course o f w hich the husband had to build a hut for his m other-inlaw and clear and hoe up a garden fo r her w ith his w ife's help. Then he took his w ife to his ow n village and cut, cleared and hoed up the rough ground into a garden for her. In subsequent years a m uch larger share o f cultivation fell to the w ife’s lo t and she was also expected to help her husband in his ow n gardens. Nowadays, the w ork for the m other-in-law is often com m uted into a cash paym ent, ranging from 10s. to £ 1 or m ore, depending on the wealth and status o f the parties involved. Agricultural production, then, is pronouncedly individualistic in character. C ollective w orking-parties do take place but they are brief and sporadic and perform ed for individuals. There is no concept o f a jo in t estate, w orked collectively b y a village or lineage-segment, the produce o f w hich is ow ned in com m on. Marital and fam ily teams frequently w o rk together but b y no means necessarily or invariably. M en and w om en ow n their own gardens and spouses have no rights in one another’s gardens. O n divorce a w om an m ay continue to w o rk in, and feed her children from , her ow n garden at her ex-husband’s village i f she lives nearby. I f a man dies his w id o w ’s standing crop is sold on her behalf b y her nearest m atrilineal kinsman when he comes to take her back to his village after the period o f m ourning is over. Today, when the surpluses o f subsistence crops are sold for cash, and when in some areas new ly introduced crops such as rice are grow n solely for the cash they bring in, husband and w ife retain for their separate use the m oney obtained from the sale o f all their respective produce. W hen m y w ife and I attempted to collect ‘ fam ily budgets ’ w e found that spouses never pooled their incomes and that few o f them knew h ow much cash the partners had obtained. Consumption A lthough agricultural production is dom inantly individualistic, consumption is m ainly com m unal. In villages the men eat in
24
Schism and Continuity in an African Society
the central thatched shelter (chota or njangu) and the w om en take turns in cooking fo r the w hole group. In die village I knew best there was a clockw ise rotation in w h ich each adult wom an in the circle o f huts cooked successively fo r all the men. Cassava mush, the invariable staple o f every meal, was prepared b y a wom an and her daughters. T he dried cassava roots had been previously pounded into flour b y one wom an, or m ore rarely a team o f w om en, ranging from tw o to four. Such pounding teams con sisted o f a pair o f co-w ives, or a group o f sisters, or a m other and her daughters. W om en, girls and wicircum cised boys ate their food, cooked on a separate fire from that on w hich the m en’s food had been prepared, in the kitchens, either in fam ily groups o r w ith friends. Sometimes an elem entary fam ily, including the husband, w ould eat together in the w ife’s kitchen. In the northern pedicle this arrangement is becom ing m ore com m on today than the collective meal o f the m en’s group. This is yet another aspect o f the general breakdown o f the traditional social organization in this area w here subsistence cultivation is steadily givin g w ay to petty com m odity cultivation. It w ill give w ay m ore rapidly w hen access to urban markets becomes easier and when new local markets com e into existence. Generally speaking, the bulk o f the cassava required fo r the m en’s food is taken from the men’s gardens and the m ajor quantity for the fam ily meals from the w om en’s. V isiting kin and friends are fed from the gardens o f their hosts, not those o f the spouses o f their hosts. B ut w ives are required to process the food and cook for their husbands* guests, and failure to do so is placed high am ong the causes o f early divorce. In the total agricultural system, from the clearing o f gardens to the cooking o f food, w om en play a m uch m ore important role than men. M en w o rk in short spectacular bouts o f energy ; the regular patient labour o f the w om en in hoeing up mounds, w eeding, digging up roots, soaking, carrying, drying, pounding, and sifting, and finally in cooking the cassava mush provides the steady sustenance o f the group. N dem bu consider a balanced m eal to consist o f cassava mush and relish. E ven when meat or fish is unavailable as relish wom en can substitute beans, cucurbits and small relish plants from their streamside gardens. M ale activity, generally in the co-operative form o f the chenda, sets in train, as it w ere, an agricultural process which is then
H istorical and Ecological Background
25
\
maintained throughout the year b y w om en, each w orking alone in her garden o r fitfu lly assisted b y her husband. In agriculture, as in other aspects o f N dem bu life, fem ininity m ay be equated w ith continuity, m asculinity w ith discontinuity. Nevertheless, the overall productive role o f m en is not regarded by Ndem bu as being m erely ancillary or peripheral to that o f the wom en. T he men have a role in the productive system w hich ;V is autonomous and from w hich w om en are debarred not m erely 4 by its inherent dangers and difficulties but in addition by a number o f ritual taboos. This role is that o f hunter. •
Hunting: A Masculine Occupation It m ay almost be said that the N dem bu social system is pivoted > on the im portance o f hunting.1*3 This im portance does n ot derive 4 from the objective contribution to the food supply made b y the chase. H unting owes its high valuation, on the one hand, to an association consistently made am ong m any Central and W estern Bantu between hunting and high social status, and on the other, to an identification made am ong these peoples o f hunting w ith masculinity. There are num erous traditions ascribing the foundation o f Central and W est-C entral A frican states to hunters. In other societies in this area hunters appear as an aristocratic class. T h e Lunda em pire in the C o n go w as said to have been founded b y a Luban hunter called Chibinda (m eaning * hunter *) Ilunga, whose son became the first M w antiyanvw a. T h e Luba state itself w as established b y * a tribe o f hunters com ing from the great lakes to the north*.8 T h e O vim brnidu chiefdom o f V iye (Bihe) was founded b y a hunter o f Lunda origin. A ccording to Capello and Ivens * the * real aristocracy am ong the Jinga o f western Angola was com posed o f hunters and warriors \ Baumann points out that the muri class o f nobles am ong the M bala w ere hunters. T orday 4* writes that * the ch ie f character in the _. _ ai;-:■
£
-—
____
1 In this they appear to resemble the N orthern Chokw e o f whom M cCulloch, p. 35» writes : * N orthern Cholcwe society is oriented around the importance o f hunting in its econom y.* * Dias de Carvalho, p. 54. 3 Capello and Ivens, vol. ir, p. 69.
4 Torday, E ., * O n the Ethnology o f the South-W est Congo Free State*, p . 152.
Journal o f the R o y a l A nthropological Institute (19 0 7 ),
26
Schism and Continuity in an African Society
peculiar politico-religious revolution which resulted in the institu tion o f the riamba cult am ong the Bashi-lange was a hunter, and the pakassero revolutionary society [sic.] o f Ladislau M agyar (p. 2 6 6 ) was a society o f hunters \ Baumann * writes o f the C hokw e that there is a ‘ hunting class ’ w ith special privileges and a high social status, and Chokw e religious and m agical rites emphasize the importance o f hunting.2 T he Bangala tribe o f W estern A ngola w ere founded b y a Lunda hunter, Chinguli. These examples could be greatly m ultiplied from the literature, and are cited here m erely to demonstrate the high value set on hunting throughout the Central Bantu culture area, especially am ong groups o f Lunda and Luba origin or affinities. M any o f these tribes have traditions o f protracted m igration from one region to another, during w hich they raided settled peoples for their crops and killed game for their meat supplies. E ven when, like the Bangala and Shinje o f western A ngola founded b y Lunda leaders, they finally established themselves in a single area, their villages still practised a semi-nomadic w ay o f life, livin g in m ore or less tem porary villages and m oving to new sites w hen game was scarce. It has been pointed out b y de Préville s that m any o f these hunting tribes inhabiting the forest country to the south o f the C ongo Bend grow cassava. He endorses D r. Livingstone’s observation * that cassava-growing involves very little w ork ’ and that it is m ainly carried on b y the hunters’ w ives. This is certainly true w hen one considers the extent o f masculine partici pation in the agricultural w ork o f a m illet or sorghum econom y, carried on b y chitemene ash-planting techniques. I do not mean to suggest that grain cultivation is incom patible w ith an emphasis on hunting but m erely wish to indicate that the extrem ely im portant status o f hunting in the N dem bu hierarchy o f values is congruent w ith a type o f cultivation w hich does not make heavy demands on men’s time and energy.4 Since cassava was intro1 Baumann, H. von, Lunda : bei Bauern und Jägern in Inner-Angola, Ergeb nisse der Angola-Expedition des Museums für Völkerkunde (1935), p. 11. 2 Torday writes (Cam p and Tramp in African W ilds (1913), p. 208) : * the most striking feature o f the Kioko in later times is that ¿hey were essentially a nation o f hunters and. ironworkers, but principally the former.* 3 Préville, A. de, L es Sociétés africaines (1894), p. 195 et seq. 4 Cassava certainly anchois a group to the ground for at least four years after planting, but during this period men are less heavily involved in agri culture than in a grain economy.
H istorical and Ecological Background
27
duced into Central A frica on ly in the sixteenth century and most authorities;, are in agreement that the first m igrations from the Lacustrine R egio n took place at an earlier date, it m ust be assumed that the original hunting invaders did n ot possess cassava. B ut its rapid and ready acceptance b y tribes w ith a hunting tradition attests to its com patibility w ith hunting in the total system o f subsistence. The im portance o f hunting, then, am ong the N dem bu, a migrant group from Luunda, has its roots in tradition. B ut its value is also accentuated b y its role in the conflict between the sexes.1 In the idiom o f N dem bu ritual, hunting and m asculinity or virility are sym bolically equivalent, and the sym bols and gear ofhuntsmanship are reckoned to be m ystically dangerous to female fertility and reproductive processes. Thus, when a w om an is in labour, her husband must rem ove his hunting equipm ent and medicines from her hut and kitchen, otherw ise his lu c k and skill at the chase w ill fail and she m ay die in labour o r have a mis carriage. Conversely, a w om an m ust n ot approach too closely a hunter’s village shrine or she w ill have a protracted and painful menstruation. The same supernatural penalties w ill also be im posed on her i f she approaches too near to the circum cision medicine o f a senior circum ciser (tnbimbi), w hich, like m any hunting-medicines, is placed in the cleft o f a forked stick (chishitiga or muchanka). This m edicine (njutzda) is the supreme sym bol o f masculinity as opposed to fem ininity, and the close parallel between the effects o f its supernatural sanctions against w om en and those o f hunting m edicine and apparatus supports the association w ith huntsmanship. * For the man, huntsmanship ; fo r the wom an, procreation * (neyala wubinda, namumbanda tusemu), is the Ndem bu saying. This is illustrated b y a folk-tale in w hich a man, w hen his village is raided, snatches up his gun and hunting gear, w hile his w ife’s first thought is for her baby. She runs into the bush w ith her child, w hile he flees w ith w hat he m ost values, his hunting things. Later she goes to her relatives. W hen her husband follow s afterwards she refuses to return w ith him . H e 1 Conflict between the sexes is ritually expressed among a number o f related tribes in western Northern Rhodesia and eastern Angola, including Iamda, Lwena, Chokwe, Mbunda and jLuchazL Cf. M . Gluckman's ‘ The Role o f the Sexes in W iko Circumcision Ceremonies *, in Social Structure : Essays Presented to A . R. Radcliffe-Brow n , ed. Fortes (1949).
28
Schism and Continuity in an A frican Society
demands his child. H er relatives say, * N o , the child is hers, for she has saved its life. The gun and axe are yours, for your only w ish was to save them / Ndem hu, both men and wom en, put forw ard this story as an argument in favour o f matrilineal descent. Cassava, maize, sweet potatoes and other crops cultivated and tended b y w om en frequently appear in rituals as emblems o f fem ale fertility. T he conflict between m ale and fem ale is ritually expressed in terms o f each sex’s dom inant contribution to the m ode o f subsistence. H unting and cultivation are com plem en tary activities but they are also opposed and com petitive. In the past, w hen hunters w ent into the bush they left their w om enfolk undefended. I f a hunter was killed his w ife and children w ere fo r a tim e left unprotected. A t certain times o f the year, a man’s econom ic roles as hunter and gardener m ight com e into com peti tion. T h e first rains w hich softened the earth and made it pliable fo r the hoe also stimulated the flush o f fresh leaves and grass w hich attracted the antelope and made them easy prey to the gun and bow . H unting brought m en together and divided them from w om en. A s w e shall see, this probably had im portant effects on the social com position o f residential groups, and the form o f post-m arital residence. The Social Organization o f Hunting H unting, like agriculture, m ay be carried on either collectively o r individually. N dem bu classify hunting techniques into tw o m ajor categories and to some extent these determine the social organization o f the hunt. Hunters are divided into hunters w ith gu n s1 (ayiyanga ; sing, chiyangd) and hunters w ith bows and traps {ayibinda ; sing, chihinda). T he chiyang*a tends to make a profession o f hunting, devoting as litde tim e as he can to gardening activities and as m uch as he can to roam ing about in the bush1 1 Gann, p. 36, states that4in. 1897 the Western Lunda had still been reported to be without guns, but by the time the officials o f the British South Africa Company began to enter their country [1904], they were well supplied*. The Ndembu themselves say that they first obtained gum from Ovimbundu traders and already possessed a limited number o f muzzle-loaders before the Chokwe raid (c. 1885). One old man o f about eighty years o f age told me that when he was a boy (c. 1880), the cult for hunters with, forearms, called W uyang'a (see Chapter Ten) had only recently been introduced from tribes to die west (Lwena and Luchazi). It is likely that Ndembu have had guns since about 2870.
1 3
<3 O O O
0
o © © © © © ©
© © ©
© ©
© ©
©
©
o ©
©
®
i
PLATE I T h e V ailue
set o n
H u n t sm a n sh ip
Four hunter-adepts play wooden stridulators before a shrine for hunterspirits during a performance of the W uyang'a ritual. The winnowing basket on the antelope horns contains herbal medicines to make the hunters invisible to their prey and to attract animals within range of their ancient and unreliable firearms.
H istorical and Ecological Background
29
in search o f gam e. Frequently, he obtains vegetable produce in exchange fo r m eat. Ayiyang*a participate in a cu lt in w hich there are four degrees o f initiation. M em bership in this cu lt enables them to range freely over Nclem bu ana indeed over Lunda territory, for w herever they g o they are sure to fold a cult-fellow w h o w ill give them tem porary hospitality in return for a share in their kills. T he chiyang*a, the * big hunter *, w ho is also a * gun-hunter tends to seek out the bigger game, such as buffalo, eland, gnu, and roan and sable antelope, although most If! o f his kills consist o f w oodland buck such as duiker and yello w backed duiker and small plains animals such as dik-dik. He is ; not infrequently a solitary, going w hither he w ill, and return% ing to his ow n village w hen he chooses. H ow ever, ayiyang*a occasionally form tem porary partnerships w ith other hunters, especially i f dangerous animals such as buffalo are being pursued. Sometimes, too, they take w ith them young m en w ho have shown an early flair fo r hunting and w h o have been apprenticed to them b y their fathers or mothers* brothers. M ost ayiyang a endeavour to train their sons in hunting and w oodcraft, and the father-son link am ong hunters is ritualized in a num ber o f w ays. A hunter father is also entitled, to bestow one o f a lim ited num ber o f names norm ally reserved fo r hunters on his son, even although the latter m ay n ot be a hunter. In his life-tim e a hunter m ay train a num ber o f apprentices : sons, m atrilineal kin, and even the jun ior relatives o f his friends and neighbours. A chiyang a m ay lead a sm all band consisting o f men and boys o f his ow n village o r o f contiguous villages into the bush on quite long expeditions lasting a w eek or m ore. O n such an expedition the party build grass or le a f huts near an area know n to be fre quented w ith gam e and continue to hunt until they have killed several animals. T he m eat is cut up in to joints w h ich are sm oked on a rough fram e o f branches to preserve them . T h e size o f the band varies from about three to ten. T he boys, often in their early teens, are taken into the bush fo r training and to carry the meat back. I f the expedition is a lo n g one the boys carry supplies o f vegetable food such as cassava roots and. green maize-cobs for the party. In the past there w ere annual organized gam e drives in the socalled ikuna areas, w here the trees are short and stunted and the grass long and rank. These drives no longer occur am ong the
30
Schism and Continuity in an A frican Society
M w inilunga Ndem bu, but Singleton-Fisher1 has described a gam e drive am ong the related Lunda o f south-western Katanga. T he ikuna areas provide shelter fo r gam e when, the surrounding bush has already been burnt, and the rem aining ikuna is burnt in an organized w ay. T he operation is directed b y the ‘ village headman \ T he ikuna is fired, and about thirty m en and boys armed w ith guns and bows and arrows encircle it and k ill the animals as they run from the flames. According to SingletonFisher, the village headman, w ith the headman from a neigh bouring village, presides over the distribution o f the gam e obtained. Each hunter w ho has killed an animal is given a special ritual portion, including head, neck, heart, lungs and intestines. This portion, called the * head * [mutu) or yinjila (literally ‘ forbidden ’ or * sacred * things), m ay be eaten only by the hunter and is thought to maintain his m ystical powers o f huntsmanship. H e is also given a leg. The remainder is shared out am ong all those w ho have taken part in the burning. Such drives are apparently organized b y any headman w ho takes the initiative, and not necessarily b y a ch ief or senior headman. Hunters w ith bows and spears possessed much the same type and range o f organization as hunters w ith guns. B ut whereas almost every m an hunted w ith bow and trap, hunting w ith guns was confined to a select few , and o f those few o n ly hunters w ith great skill and ritual prestige belonged to the higher grades w ithin the Wuyanga cult. N early all m en made traps and snares, o f a w ide variety o f kinds, and each trap and the animals caught by it belonged to the individual w ho made the trap. I have re corded several disputes in the village courts between trap-owners and men detected in the act o f stealing game from their traps. It m ay be said that although in m any respects hunting conform s to the pattern o f productive individualism 2 found in agricul ture, nevertheless, it involves on the w hole a greater degree and intensity o f co-operation than does crop-grow ing. M en hunt in bands, men and w om en perform m uch o f their gardening x Singleton-Fisher, W ., * Burning the Bush for Game ”, African Studies, vü, i ( 1948), pp. 3<5-8. 2 C f. Torday, op. cit., p. 15a : * the profession o f hunting naturally induces an adventurous and self-reliant character, and encourages a roving disposition/
15 O H istorical and Ecological Background
31
alone. It should be noted that the m ajority o f productive activities in volvin g collective w o rk are carried on b y men— clearing the bush, hunting, and also hut-building, w hich w ill be referred to below . Ndem bu m ight hunt anyw here w ithin Lunda territory and were not confined w ithin a restricted area.1 It was custom ary for a hunter to give the chest or a thigh o f any large animal he had killed to the senior headman o f the area in w hich he hap pened to be hunting. He could then carry the rest o f his lo ll back to his village fo r distribution. Certain plains and thicket forests o f Cryptosepalum trees {mavunda) had a w ide reputation for containing m uch gam e, and hunters w ould com e to them from far and w ide. T he hunters’ cult and the great lateral range o f remembered kinship ties w ere utilized b y hunters to obtain • hospitality and shelter in areas other than their ow n . Meat Distribution and Consumption The pattern o f distribution o f m eat varies w ith the size and com position o f the residential unit. I f this unit is small, contain ing a sm all bilateral extended fam ily (ntang*a), in addition to the ; portions reserved for the hunter and those given to the senior headman or chief, a back leg w ill be given to the hunter’s brother or m other’s brother or is divided betw een several brothers ; a back leg or a front leg w ill be given to his m other ; a front leg w ill be divided am ong his sisters ; the saddle w ill g o to his w iv e s; the breast w ill go to his father ; and any small pieces that remain w ill be distributed am ong boys too old to live in their parents’ huts. In the larger villages containing tw o or m ore minimal matrilineages (see page 8on.), a leg m ay be allocated to the senior man o f each lineage, and it w ill be further sub divided by the latter am ong the m arried m en o f his lineage. Consum ption o f gam e, like consum ption o f cassava mush, tends to be collective. It is considered to be good manners for each man o r w om an w ho has received m eat to have a portion o f it cooked for the men’s eating-group in the village shelter, each 1 M cCulloch, p . 16, writes that * m ost hunters confine their activities to within twenty miles o f their village, and to their tribal areas \ B u t I have known hunters who have gone forty to sixty miles from their village to hunt. This m ay have been due to the disappearance o f gam e from much o f north western Mwxnilunga, com pelling hunters to rove further afield.
©
© @ ©
© 0 o
o
0 ©
© ©
© ©
©
e
©
© ©
P : D
:-o ■ D -o
lo ;c :.Q ¿O
:o o o c
¡G O
P o
o
c
D (.i ;o
■V. ic
ID f)
32
Schism and Continuity in an A frican Society
retaining a share for his or her own. elem entary fam ily. A good deal o f grum bling goes on both over the precise division o f the m eat and over the amounts cooked for the men’s group. Again, hunters are frequently accused o f having made a kill and eaten the meat themselves in the bush, feigning on their return that they w ere unlucky in the chase. Throughout the econom ic system a tension is set up between the individual producer or killer o f food and the group w ho b y custom have claims in it. A ‘ greedy person *, one w ho persistently retains w hat he pro duces for himself, ultim ately m ay be expelled from the village. Conversely, a professional hunter w ho resents the claims o f rem ote classificatory kin on the product o f his hunting m ay lead aw ay from a village a small group consisting o f his w ives, and children, and his sister’s children, to found a new village. Successful gun-hunters are regarded as sorcerers,1 w ho acquire their pow er in hunting from killing people b y means o f their familiars. That is w h y great hunters seldom becom e successful headmen, in the opinion o f Ndem bu. T heir nom adic inclination, their tendency to favour prim ary rather than classificatory kin in their ow n villages, and their association w ith sorcery, disqualify them from perform ing a role w hich requires tact, generosity to classificatory kin and strangers, and constant participation in the group life, for its successful functioning. A bow hunter m ay often make a good headman, a gun-hunter but seldom does so. Summary o f the Main Features o f the Mode o f Subsistence N dem bu practise a form o f subsistence cultivation in w hich cassava-growing is associated w ith hunting. In addition to 1 The rancour and jealousy aroused in the situation o f meat distribution are not only expressed in a number o f folk-tales (tuheka or ayiskitnu) but also appear to influence strongly many beliefs about witchcraft and sorcery. A witch or sorcerer kills his victim * for meat \ whatever his other motives for killing are thought to be. Witches gather in covens to participate in necro phagous feasts on the bodies o f their victims. In Social Drama I (p. 95), and IV (p. 148), quarrels over meat distribution provide grounds for subse quent accusations o f sorcery and witchcraft. Professor Monica Wilson discusses a similar connection between witchcraft and the lust for meat among the Nyakyusa o f Tanganyika (‘ Witch Beliefs and Social Structure \ T he Am erican Journal o f Sociology , Ivi, 4, Jan. 1951). She writes that * witches lust for meat and milk— the prized foods o f the group — and it is this which drives them to commit witchcraft. They delight in eating human flesh and gnaw men inside, causing death to their victims’ .
H istorical and Ecological Background
33
cassava, finger-m illet is grow n b y small circle ash-planting (chitemene) methods m ainly fo r beer-m aking, and maize is cultivated in streamside gardens fo r food and beer. O ther crops are grow n in bush, village and streamside gardens m ostly to provide relishes. In agricultural production and in the pre paration o f food w om en have the dom inant role. H unting is a male m onopoly. A lthough in both agriculture and hunting production tends to be individualistic in character, where co•n-K operation is w ider than that between members o f a fam ily it rends to be betw een men, w hile w om en w o rk on their ow n. J =v|4 ; But members o f an elem entary fam ily often w ork together, i though co-w ives seldom assist one another in gardening. In -.the sphere o f consum ption, on the other hand, the emphasis is I V V ' on sharing the product o f gardening o r hunting am ong kinsfolk j , living in jo in t residence. Here again the larger collective units 'M ’ 0f consumption tend to be m ale village kin, w hile the wom en eat separately w ith their unmarried daughters and uncircumcised sons.
". '•8
15
C H A P T E R II T H E V IL L A G E : T O P O G R A P H Y A N D D E M O G R A P H Y Introduction H E N dem bu are settled in discrete circles o f huts w hich are n ot scattered evenly over the w hole region but tend to be loosely grouped into spatially distinct vicinages. These circles o f huts are traditionally know n as nyikala (sing, mukala). This term m ay be translated as * village \ T oday, N dem bu distinguish villages entered on the Governm ent T a x R egister (and still called nyikala)t from other residential units w hich have not yet been so entered. The latter are know n as * farms * (majwami), an English term said to have been borrow ed from African tribes livin g near the Copperbelt w h o possess similar units. Few farms are older than eight years. T he term * farm * conceals tw o clearly distinguishable types o f residential grouping. T he first o f these represents the first stage in the life-cycle o f the traditional N dem bu village ; form erly such incipient villages w ere know n as nyana ya nyikala, * the children o f villages V the original village being regarded as the parent o f the new one. The second is a new kind o f settlement, possessing certain special features o f hut arrangement and social com position unknown to the pre-European N dem bu social organization. These features w ill be discussed in subsequent sections. I propose to use the term ‘ registered villages * for the villages appearing in the tax register, ‘ unregistered villages * for new ly established villages organized on the basis o f traditional principles o f social structure, and ‘ farms * for the new type o f settlement. It m ust be noted, how ever, that at the present tim e there exist a num ber o f transitional types between these kinds o f residential unit, and that it is often difficult to distinguish ‘ unregistered villages ’ from * farms * unless one has personal know ledge o f the econom ic interests and social ambitions o f the head men. 1 Sometimes the new village was known as mukala watwansi, * village o f the juniors \
34
0
0
©
0 ©
0 @
0 © Q © © © Q
© © Q
o
® © O © © ©
P in b
P LAN
^ cP manwsa
O
OF
M UKANZA a
•iO
d
j
a
c
e
n
V IL laJ « r s
a n d
t
p
a
r
^
|
g iÇM IKA3A 1
:o
\
MARIO T f t - J J
ï)
i \
D a u a V U t-} « f
U U C IA N A ELeW lfc»**".
O
« H ülfHI - j|p
\ I i MM vp QIH liflmwm'H «tA A 'O rA
h ^ ssss? *«
CHAW UTONO’lSJ^SÆT"' UXiktidaMJJ.
'O
x
7fcAS0Nl>A « a d yJ^ HAW»»ro^6*l
/S
./ÎF
ÎY A W A .J -, (SAKOW AM A
g S G /^y^Skiêiw < /
mukaa^ ta / ü r e z Z Ï M jF AWKAAG&i i*: S i * G y/UAAGÆ / __ /%•"—-------- moH»ma^KS“lo.»\ ' ' M A N C A U S H J ^ M g g e R , . _ y f ' H A H A L > S A U > Y ,,r * t U C îA N A n Î r f tM H J T E lfe S H R 1 M 6 ? r-------/,'YKAM«OSJWA, LUtfANCehnL* / ' i{— ' ,ft II* M U K A N ÏA / l ^ f f î gm*1...... * ** .« & * n > * ), I I gÈ8S8»C [^ * ^ ' ■ / H Y A M W A H A .F T \ \ \ y / y ft-l* _ PfeSBî'® y ilIf I JctrA ^oMbA.cir ■ Z o L lfflN A 'S V » t « > . '"-. i 2?XS^li&OW&uji^MiiKRM^A’S trfSÎÿ ^ t y '/ (1|l| |k' 1 Il I I >&»A * X U r Z U U I M M / ' S A K H ÏflO —^x3T*ï=/ /{XSSSSPwl' * —' "yivqm&# u■.;:/*;/ IH •n niTHJÇI A | ‘ ■U A M B fiS ftR IN E ------X / «WIHAMSA / , ^ T .ÿ «oN 3r-MM MV AV MK OOLHAO'»tRM»2S ) „ 'SMaifijE— l t / , 'SAKAIZAO'S Sfetw»// r c —y w VüiUAK CM Aw s*— nr«6 ^|^H AV^VT0H G *l ^ Cf« rn ) _ >/ . . W A M VM NCW *'.- / •I y K £NSOM*S MOYOM8U «fotn'AMI r„ J W W V r t lV A M I’ » * S o i « ' ¡ (f a r K ïN S < M « «iS E I-tï/ l’S S o o j ___ .. SELtffj/* Î JJ '*•----------------------— ---------------- ----------------------.ftfH în w .M »(o'»w »3 iw ic w j
o
ja » m r o ;
k w k m
kabk o a
el
O
G O
u
fCM AV^CTbNC•*> Q :
G
i 1
,SAK*2AO>fr* jN VA «.vw e frtA ,G 4
/MA rtYALX iVW C
;o f'
o S e v^ V ' Wy »ij .|
C¡ S AJK A ^ A O ,« ? r f |im
p «=si
>
VAWA’ SJ
CH AW JTONCN
c
A A T iK I^ — ^
G
»rAUSW CMA
HYAKAlMSiA
O
,C) :0
"J^ X J
ÏK |A
>
C H IH A M B A
SH A I N E NYAMUKANC*A.L f a - *.V t_»VAMf*>
: (MyATIOUl .iiiiun»
/
Zà
C) •G
....... .1 AftAM 'llllllllfilil
A A A M (J7
m\fmiffi
/ " 7 «W * T IO ta , H l
/T^/eyvt
V -.j IKATAY«
jliiit'i!;;;,,i',i ‘IIIIJMARM
V
<1 m ............ fc V T ’r ' K lA 4fll9 t» 19
£ t o
_CHAMSALQT», ft 2.0
(âtb«.ndio'|s«^
CHAM W w t»
!.€) :C
b
MAP 4
kg»W»v
i
\ X 1 M Îîh in ^ A W V K G L . „„
... iiiillii!!!>!li!nih,.
i ^y a k a l u ^ a
NVALUW£MA «Y A K A IU S A NYATIOM YAH A KATAYJ M AR IA
V..
ï$é£3
C M w u to N e '« VANA
a
•O
C
i
n
'I
The V illage : Topography and D em ography
35
i Village Topography The registered village consists o f a roughly circular or oval arrangement o f huts around a cleared space in the centre o f w hich is a round unw alled shelter (chotaor njangu), used as a men's palaver hut and mess-room. M ap 4 shows a typical registered village, Mtxkanza.1 Conspicuous features include the con centration o f ancestor shrines before the huts o f the headman and the senior representative o f another village lineage, the maize patches around the village periphery manured by kitchen rubbish, the m eal m ortar sites shared b y specific groups o f wom en, the cassava-drying platform s, and the absence o f grain granaries. The ancient hut type was the round or square grass hut, and % ■'. it was replaced in the present century b y the pole-and-m ud hut, •' which was introduced, say inform ants, from A ngola b y Lwena ;V immigrants. T he grass huts survive as seclusion huts for girls undergoing the puberty ritual and as hunters* tem porary bush camps. I am told that villages in the rem ote past w ere m erely hunting camps o f grass huts, easily made and easily abandoned. H owever, during the slave-raiding epoch they w ere usually palisaded w ith strong stakes and sometimes w ere encircled w ith high earthworks. T he diameters o f village hut circles are fairly sm all and range between about thirty and seventy yards. T he most com m on type o f hut w hen I was first in the field was the rectangular or square pole-and-m ud hut. H ut measure ments lie betw een 9 ft. b y 9 ft. to about 15 ft. by 25 ft. for this type. W hen I left, the pole-and-m ud hut was everyw here giving w ay before the house o f sun-dried K im berley brick. Some o f these exceeded 60 ft. in length, although the average was about 30 ft. Pole-and-m ud huts usually contain o n ly a single room , although som e m ay have tw o. M ost K im berley-brick houses have at least tw o room s, and a few , the houses o f chiefs, have as m any as six. A t one tim e there was a certain correlation between size and type o f hut, and social status ; but the recent 1 As it was in 1953. Frequent reference will be made to this village in the text. I have concealed its identity under a pseudonym and altered the names o f its inhabitants. I have adopted this procedure elsewhere in the book whereever necessary.
36
Schism and Continuity in an A frican Society
Governm ent p olicy o f inducing all men to build brick bouses is obliterating this distinction. T he forest is cleared all round Ndem bu villages in a wide ring to prevent damage from trees blow n over b y the d ry season winds or felled b y die lightning strokes o f the rainy season. Since the first cassava-gardens are made close to a new village i f soil conditions perm it, m ost villages are clearly set o ff from the environing deciduous forest o f the D istrict. W hen a village moves to a new site— in the past this took place every four or five years at the end o f the rains— circular grass huts are built for the first few months or sometimes even fo r the first year o f residence. Then the pole-and-m ud huts o f the headman and senior m en o f his genealogical generation are built for them b y the younger men, in an approxim ately semi circular form ation, leaving fairly w ide intervals between huts. W hen these are com pleted the m en o f the jun ior, proxim ate generation build pole-and-m ud huts in a sem i-circle opposite their seniors, assisted b y their sons and sisters sons. M ature m en o f the ju n io r alternate generation to the headman m ay then build huts between those o f their grandparents. Unregistered villages are often built today, when village m obility is less, w ith w ide intervals between the huts to leave room for future members. In the past this was less com m on since the diameter o f the hut-circle could soon be altered at the next village m ove. Farms present a quite different spatial arrangement from villages. T h ey are no longer a circular grouping o f huts but typically consist o f one or more K im berley-brick buildings flanked on either side b y a few small pole-and-m ud huts and kitchens.1 Occasionally, a hedge o f flow ering shrubs m ay be planted in front o f a farm to screen it from the m otor road. M ost true farms are situated beside the m otor roads, fo r the typical farm -head is a m an w ho has earned m oney, often on the line-ofe rail, and w h o intends to earn more locally. H e m ay be a petty trader, a tailor w ith his ow n sewing-m achine, a 4tea-room * proprietor, a * beer-hall * owner, o r a peasant producer raising cash crops. For all these purposes easy access to m otor roads is necessary and propinquity to administrative and trading centres 1 See inset, Map 4, Sandombu Farm.
The V illage : Topography and D em ography
37
advantageous. For these reasons m ost farms are found in the northern pedicle, or near the B om a, or close to chiefs’ capitals, or in the vicin ity o f M ission Stations— i.e. w herever there are large concentrations o f population and good com m unications w ith urban areas. Village Magnitude Ndem bu villages have a sm aller average m agnitude than those o f m ost other Central A frican peoples about w hich data are available. I mapped, and collected genealogies in, fifty regis tered villages, eighteen unregistered villages, and nine firm s. In addition, I obtained census inform ation in sixteen o f the registered villages, eleven o f the unregistered villages, and tw o o f the farms, in w hich I collected genealogies. These genealogical and census studies w ere examples, rather than a random sample. T h ey include a high proportion o f settlements near to m y first and second tour camps in Ikelenge area and Kanongesha area and along or w ithin easy reach o f the m otor roads. T he proportion o f registered to unregistered villages and farms is definitely incorrect, since in Ikelenge A rea alone, in 1953, there w ere 15a unregistered villages and farms as against 101 registered villages. Nevertheless, I believe that the inform ation regarding the relative average m agnitude o f registered and unregistered villages and farms is n ot far from being correct. Estimates b y eye o f the num ber o f huts in m any settlements 1 w here I did not collect genealogies suggest that these settlements had closely similar means and ranges to those in m y sample. There was a total o f 657 huts in the 50 registered villages w here I collected genealogies, the smallest having four huts and the largest fifty-fou r. The mean num ber o f huts per village was 13«1 and the median was 12. For unregistered villages there w as a total o f 135 huts in 18 villages, the smallest having one hut and the largest 16. T he mean num ber o f huts per village was 7*5 and the median was 8. For farms there was a total o f 23 huts in 9 forms, the smallest having one hut and the largest 5. T he mean num ber o f huts per form was 2*6, and the median was 3. ; ‘
1 1 employ the term * settlement * as a generic name for registered and unregistered villages and forms.
38
■ o
Schism and C ontinuity in an A frican Society
The distribution o f huts over the three types o f settlement is shown in Table II. For all types o f settlement it was found that in 77 settlements there was a total o f 815 huts, or an average o f 10*6 huts per TABLE II
CN
um ber o f
H uts
in
S ettlem en ts
iar\ ■ jxy
•o ■ C 0 c c :;0 te
st: t )
settlement. T he median was 9 huts per settlement. B ut it must be remembered that m y sample was not stratified in accordance w ith the proportions existing between these three categories o f residential unit in the total Ndem bu area. In the pedicle, for example, farms and unregistered villages outnum ber registered villages. O n the other hand, in the rest o f the N dem bu area, registered villages probably outnumber the other tw o categories. From the census sample it was found that there was a total number o f 467 inhabitants living in 198 huts in the sixteen registered villages, o r an average o f 2*35 persons per hut. In the eleven unregistered villages 158 persons occupied 72 huts, or an average o f 2*2 persons per hut, w hile there w ere 15 persons livin g in the six huts o f the tw o farms, or an average o f 2*5 persons per hut. In the total sample 640 people occupied 276 huts, or an average o f 2*32 inhabitants per hut. Thus the average registered village o f about 13 huts contains about 30 persons, the average unregistered village o f 7-8 huts about 16 -17 persons, and the average farm o f 2-3 huts about 5—8 persons. In all villages w ith more than tw enty huts, it was found that the main village circle contained in no case m ore than tw enty
T h e V illage : Topography and D em ography
39
huts. W hen this num ber was reached, either a n ew hut circle was started a score or so yards aw ay, or new huts w ere built outside and along the periphery o f the main ring. I summarize below some inform ation on the size o f N dem bu villages as com pared w ith the size o f villages in a num ber o f Central A frican peoples.
People
Luchazi1
...............................
Ngoni of Fort Jameson 2 . Bemba : 3 Com moners.......................... Chiefs..................................... Nyakyusa4 .*.............................. Plateau Tonga 6 ..................... Unga of Lake Bangweulu * . Lwena1 ............................... Lamba, Mushiri reserve 7 . Lamba: 89 Old A r e a ................................ New A r e a .......................... Intermediate Area . . . . Kaonde8 ............................... Lunda N d e m b u .....................
N o . o f huts per village
N o. o f taxpayers per village
Mean population per -village
Up to 200-300 (divided into sections o f 20 huts each) 6-120 30-50 300-400 30-40 30
20 69 20-30 42*1
17 (mean)
7-27 50-1
400 38-1 n*r (mean) io*6 (mean)
22*1
1 McCulloch, M., pp. 66-7. 2 Barnes, J. A., * The Fort Jameson Ngoni Seven Tribes o f British Central Africa, ed. Colson and Gluckman (1951), p. 206. 3 Richards, A. I., * The Bemba o f North-Eastern Rhodesia \ Seven Tribes, P* 4 Wilson, Godfrey, * The Nyakyusa of South-Western Tanganyika *, Seven Tribes, p. 270. * Colson, E., ‘ The Plateau Tonga of Northern Rhodesia*, Seven Tribes, p. h i . * Brelsford, W . V., Fishermen o f the Bangweulu Swamps, Rhodes-Livingstone Paper No. 12 (1946). P- 3 *» 7 Allan, W., Studies erf African Land Usage in Northern Rhodesia, RhodesLivingstone Paper No. 15 (i9 4 9 ). P* 27. * Mitchell, J. C. and Bames, J. A., T he Lamba Village, Communications from the School o f African Studies, New Series No. 24 (1950)* pp- 21, 29. 9 Watson, W ., * The Kaonde Village \ T he Rhodes-Livingstone Journal, xv (July, 1954), PP* <5- 7 -
40
Schism and Continuity in an A frican Society
These figures are presented m erely to give a rough impression o f com parative magnitudes, and not to make precise distinctions. T he material available is in some cases not strictly comparable. For instance, the data cited from Richards and W ilson w ere collected before the Second W orld W ar, and it is highly probable that there has been a marked reduction in the average size o f villages in both areas since 1946. O n ly in the cases o f the Unga, Lamba, Kaonde and N dem bu were the total num ber o f villages and their range o f size stated. B ut in spite o f the obvious deficiencies in the data it w ould seem that N dem bu villages have a m arkedly smaller average size than those o f the other peoples shown, w ith the exception o f the Kaonde, eastern neighbours o f the M w inilunga Lunda. Today Ndem bu villages probably have a sm aller average size than the figure given (io*6 huts) since there has been a big increase in the num ber o f forms since I95°. In The Lamba Village (p. 2) the term * O ld Area ’ refers to Reserve N o. 1, N dola D istrict, w hich was considered to be overpopulated in relation to the requirements o f the agricultural system. * Intermediate Area ’ refers to án area o f C row n Land reopened to Lam ba settlement where the people w ere bound to observe certain agricultural rules, and where their villages were laid out for them. In the * N ew Area * the boundaries o f the area w ithin w hich a village could make gardens w ere mapped b y agricultural officers according to the estimate o f its needs o f arable land. For the M w inilunga Lunda, W hite 1 gives a historical reason for the small size o f villages. T he Lunda w ere never strong enough to repel C hokw e and Lwena invaders, even i f concen trated, and therefore preferred to seek safety in dispersion. W hite suggests that the continued preference o f the Lunda for living in sm all villages is due to their * general conservatism * : * kinship ties remain stronger, and the village headman is m ore respected, than am ong the Lwena or Luchazi.’ H ow ever, m y older informants have repeatedly declared to me that when they w ere you n g (some fifty years or so ago, w hen the slaveraids had hardly ceased), some villages were m uch larger than those o f today. Some o f them have even told m e, from m em ory, 1 Cited by McCulloch, p. 24.
T he V illage : Topography and D em ography
41
how m any huts there w ere in the village in w h ich they grew up, and some villages today w hich have o n ly a dozen huts once contained about thirty. E arly travellers am ong the Southern Lunda,1 adm ittedly for the m ost part am ong the Lunda o f C h ie f Ishinde to the south o f Kanongesha’s chiefdom , m ention * large, stockaded villages * in the Rabom po area. H arding, for instance, writes o f ‘ a Lunda induna w h o resided in a fortress o f no mean dimensions, enclosed w ith a high earthen w all \ O n the other hand, older N dem bu have told m e o f a tradition that ‘ long, long ago \ before the slave trade, their ancestors used to live in small encampments o f grass huts, o f similar shape to those described and illustrated b y sketches fo r the Lunda o f M w antiyanvwa b y Dias de C arvalho.2 These settlements w ere appar ently little m ore than hunting camps ; after destroying the game in a given area the village w ould m ove on. I f there is any historical truth in this; tradition then, the present sm all size o f villages represents partly a return to an ancient type, and partly, in the case o f * farms *, a response to changing socio-econom ic conditions in Central A frica. This response has been assisted, no doubt, b y the fact that Governm ent in spite o f its * ten tax payers ru le’ did not m ake determ ined attempts to im pose on the M w inilunga Lunda the rule that a village should contain a minimum o f ten taxpayers. Since 1950, in fact, the p olicy o f the D istrict Com m issioner has been to allow the unchecked grow th o f sm all forms, so that in accordance w ith central govern ment p o licy the * parish * rather than the village should in tim e become the basic unit o f N ative Adm inistration. It w ould seem, how ever, that some D istrict Officers made sporadic attempts to com bine several sm all villages into single residential units (see, fo r exam ple, p. 182). Further evidence against the view that dispersion in the foce o f slave-raiders was the on ly cause o f the small size o f villages is found! in. the foct that the m ajor slave-raiders o f the epoch in W est Central A frica, the C h okw e and Bangala, themselves lived in small villages. Capello and Ivens,3 when passing through 1 Livingstone, op. dt., p. 283 ; Harding, p. 137 ; Cameron, V. L., Across Africa (1885), p. 405 (of Lunda ‘ between sources o f Lulua and Zambezi ’). 2 Dias de Carvalho, p. 220. 3 Capello and Ivens, voi. i, pp. 236-8.
42
m
;o :0
& 1 ::
C P
A. O
:D :;0 e •C;
|L> I:.; :;:-V
if
-C X-
I)
Schism and C ontinuity in an A frican Society
C hokw e territory in 1877, m ention m any sm all * hamlets \ * miserable little hamlets ’ (vol. 1, p. 216), and they insert illustrations o f villages (‘ senzalas ’) o f eighteen huts and less. In Bangala territory (vol. 1, p. 362) they m ention * a tibata or village [which] is a favourable specimen o f its kind. The stockade by w hich it is surrounded encloses an ample space containing a couple o f dozen habitations/ Baumann’s descriptions o f C hokw e villages, summarized by M cC ulloch,1 show that they had a remarkable sim ilarity to N dem bu villages. In the northern savanna country, where hunting is the main activity, ‘ straw houses are m ore frequent, palisades are rare, and settlements are more or less tem porary \ W ithin the villages houses are arranged round a cleared space, in the centre o f w hich is the tsota or meeting-house. The two village plans w hich M cC ulloch reproduced from Baumann show that in one village, from the northern, m ainly hunting, C hokw e, there are 14 huts ; and in the other, from the southern, m ainly agricultural, C h okw e, there are 31 huts. Capello and Ivens (vol. 1, pp. 217-23) describe tw o hunting camps o f C h okw e th ey found in the bush, * o f grass-thatched huts in an irregular circle \ containing * a score or so o f men w hich resemble those described by older Ndem bu. It is indeed quite probable that an ecological factor, i.e. a com m on emphasis on hunting, played a prom inent role in restricting village size am ong both C h okw e and Ndem bu. Later die effects o f various sociological factors w ill be discussed for N dem bu in this connection, but these too w ill be shown to be directly and indirectly related to the high value attributed to hunting. O n the basis o f their agricultural system alone there is no reason w h y N dem bu should not live in larger villages.2 Indeed, today, in the north-western pedicle where hunting is obsolescent, high population densities actually do subsist on agriculture. B ut even in this area individual villages tend to be small and are in fact becom ing smaller w ith the rapid grow th o f * farms \ W hat one finds here is ribbon developm ent along the road o f small settlements, a few hundred yards aw ay from one another, each w ith its ow n headman and each containing a small 1 McCulloch, pp. 40-3, gives Baumann’s hut diagrams. 2 See p. 21 for Allan’s estimate o f carrying capacity o f land under Ndembu system o f cultivation.
The V illage : Topography and D em ography
43
corporate grouping o f kin. H ere the factor o f cultural persistence or inertia seems to be im portant. B ut even in areas where game is still plentiful, and where European contact and influence is less marked, villages tend to average about a dozen huts. It is possible that village size is here regulated rather b y the game resources o f the surrounding bush and neighbouring plains than by the fertility requirements o f cassava, the staple crop. In the case o f the few larger villages fo r w hich I have records I shall show that special factors are involved w hich maintain the integrity o f uie settlement under a single headman. It should, o f course, he m entioned here that the current grow th o f small villages is a phenom enon found throughout N orthern Rhodesia. It is partly due to G overnm ent p o licy w hich n ow encourages the parish system at the expense o f die traditional village ; to die pax britarmica w h ich has abolished slave-raiding, blood-feuds and inter-tribal w ar ; and to the rapid exterm ination o f lions. Everyw here the need to rely on one’s kin and neigh bours is being reduced b y these factors. Q uite as im portandy perhaps, labour m igration to the urban industrial areas is posi tively em ancipating the individual from his obligations to his kinship group. Again, i f a man wishes to accum ulate capital to set up as a petty trader or tailor, or to acquire a higher standard o f Hving for him self and his elem entary fam ily, he must break away from his circle o f village kin towards w hom he has tradi tional obligations. E veryw here, w e see the spectacle o f corpor ate groups o f kin disintegrating and the emergence o f smaller residential units based on the elem entary fam ily.1 In most cases w here fission has occurred I shall show that the parent village seldom contained at the tim e when it split m ore than tw enty huts. In ail extant villages aanuuning around tw enty huts pow erful social tensions exist and the lines or imminent fission are clearly marked. M y ow n hypothesis, necessarily tentative, about the historical developm ent o f N dem bu villages is that before the period o f the slave-raids m ost villages w ere about the same size as those found today in areas rem ote from European setdements, and that during those raids, some villages tended to unite into larger 1 For a more detailed examination o f changes brought about by the modem cash economy see my article (with E. L. B. Turner) on * Money Economy among the Mwinilunga Ndembu \ T h e R kodes-Livingstone Journal, xviii (1955).
44
Schism and Continuity in an African Society
concentrations w hile others dispersed into still smaller units.1 Several large com posite villages have been mentioned to m e by older men. O ne such village was under the leadership o f the * national * hero Chipeng’e, w ho organized his people against the C hokw e, and it contained several groups o f N dem bu and K aw iku w ho form erly lived in independent villages. Senior headmen such as M ukang’ala Ibala controlled sim ilar composite villages. O ther headmen, such as N yachiu, form ed com posite villages from w hich they raided the traditional sm all villages o f their fellow-tribesm en to sell the inhabitants as slaves to O vim bxrndu traders. Later, under the pax hritannica, these composite villages broke up again into their com ponent groups o f closely related matrilineal kin w ho founded separate villages. I have discussed this question o f village size at some length since, given the fact that N dem bu villages are inhabited by kinsmen, lim itations on their m agnitude impose lim its on the num ber o f kin w ho can d w ell together. I f maternal descent is an im portant principle governing residential affiliation, and i f in small villages lineages are stunted, it is probable that the social mechanisms fo r prom oting cohesion and reducing conflict in the residential unit w ill diner from those in societies practising maternal descent w here the large village is typical. I w ill return to this question later, m erely pointing out at this stage o f the enquiry that there is probably a functional relationship o f some im portance between the size and the structure o f the residential unit. Village M obility: The Changing Composition o f Vicinages In the past (and in some areas today) N dem bu villages changed their sites very frequently. O ne village, Nyaluhana, the core o f w hich is com posed o f matrilineal kin o f the Senior C h ief Kanongesha, is said to have m oved its site 17 times since it was founded about 75 years ago. Its average duration o f occupation o f a residential site is, therefore, about 4*4 years.2 M ost o f these 1 See p. 220 for a further discussion, on this point. Whether village members dug themselves in and built stockades or dispersed and hid in the bush depended on the fortunes o f war. 8 This agrees with Trapnell’s and Clothier’s estimate that Lunda villages generally move * after some four to six years * (Trapnell and Clothier, p. 3 8 ).
© m
0
The Village : Topography and Demography
45
moves have taken place in a narrow orbit w ithin a few miles from the present village, but its fifth m ove took it to the vicin ity o f the traditional capital o f Kanongesha in A ngola, and in 1908, after returning to its present area, it w ent to A ngola again for about ten years to escape from the British South A frica Com pany’s administration. Since its return it has rem ained m ore or less in its present vicinage. A nother village, M ukanza, o f K aw iku origin, is said to have m oved tw enty-six times since its foundation about a hundred years ago. This village again has tended to remain w ithin about seven miles o f its present site, although, when taxation was introduced in 1913, its members fled from the vicinity o f the B om a to Ikelenge’s area in the pedicle about 40 miles aw ay, returning in 1919. B u t other villages have ranged more w id ely across the face o f the country. T he village o f Chibw akata, fo r exam ple, founded b y the w ar-leader or Kambanji o f the first Kanongesha, at various times has occupied sites in the extrem e south-west o f M w inilunga and in the Kosa chiefdom o f C h ie f Saihmga, sixty m iles to die south-east o f its present location w hich is five miles from M w inilunga Bom a. O ther villages have m igrated into M w inilunga, before the Europeans came, from Lunda chiefdom s in the C on go and Angola. In general the K aw iku villages, containing descendants o f the auto chthonous population, have tended to remain m ore firm ly fixed in a particular locality than villages descended from the Lunda invaders. In the past tw o or three decades m any villages have m igrated to the pedicle area w here European settlem ent has held out hopes o f em ploym ent fo r the m en and markets fo r vegetable produce fo r the w om en. T od ay, a tendency towards permanent settlement in one site, noted above, is becom ing clearly discernible. A fter villages have m oved near a mission station, the Bom a, a N ative A uth ority centre or a European farm , they tend to becom e stabilized. A t the present m om ent a process o f fragm entation o f traditional villages into sm all forms, containing elem entary families w ith a sm all fringe o f m atrilineal kin o f the headmen, is going briskly ahead. Since at m ost forms K im berley-brick houses are being bu ilt it seems lik ely that these n ew settlements w ill rem ain in their present sites fo r a long tim e. I have made m any enquiries about w h y villages m oved in the jast and in no case w as it stated that they m oved because accessible and suitable for cassava cultivation had becom e exhausted. T he
i
© & o ©
© © © ©
0 © ©
© © © © ©
© ©
© ©
© © ©
© ©
!CT C
P
46
r?
r<.
:e--f:
P t )
■ b
:o
;C:
P
.;o 0
:-0 (5
sf"j. rt; ' ■O K ■; n t ;
•C I ,; ;c bX-.
:
Schism and C ontinuity in an African Society
most frequent explanation for change o f site given me was that in the past older people w ere buried under the floors o f their huts w hich w ere then burnt dow n, follow ed by the m ovem ent o f a village to a new site.1 Fear o f the spirits o f the dead led to change o f site. For m any years, how ever, it is said, Ndem bu have follow ed the European practice o f burying everyone in graveyards. These are situated in the bush a few hundred yards from the village. A s a result frequent change o f location has becom e unnecessary. Another reason for m ovem ent was alleged to be feuds between villages in the same or in adjacent vicinages. Feuds were sometimes prolonged and the weaker villages fled to new areas. Again, I have been told that when a village has remained on the same site fo r a num ber o f years, larger game and small mammals get destroyed in its immediate neighbour hood, and the village m oves nearer to new hunting resources. Thus the conscious m otives for change o f site are given as ritual, political and econom ic, in that order. The econom ic m otives are couched in terms o f the exigencies o f hunting although there is little doubt that the proxim ity o f suitable soil (Ndem bu are soil-selectors, choosing land fo r cassava cultivation b y the dominant species o f trees on it) helps to determine their choice o f one site rather than another. B ut it is unlikely that villages w ere com pelled to m ove every four or five years b y the ex haustion o f local soil. G iven a cultivable percentage o f about 20 per cent, a village o f thirty inhabitants, using the five-year garden system described above, w ould require about i,ooo acres to maintain it in perpetuity. Theoretically, therefore, i f the population remained stationary and the forest regularly regenerated, there is no reason w h y a village should m ove its site at all. Factors other than those inherent in the agricultural system were probably at w ork. A n attempt w ill be made later to isolate and analyse the most im portant o f these factors. M eanwhile it can be stated that N dem bu villages w ere small, changed their sites often, and frequently m oved considerable distances from one site to the next. W hen the new site was w ithin a mile or so o f the old one, the old cassava gardens con-
:( 1 It is said that after the death o f a headman, or after the consecutive deaths o f two mature villagers, a village always moved.
1 *
The V illage : Topography and D em ography
47
tinued to be utilized from the n ew site until they w ere used up. In the past, w hen a village m oved to a com pletely new area, its inhabitants are said to have made large finger-m illet gardens in the first year, for food as w ell as beer. Finger-m illet is a quickmatuiing crop (sown in D ecem ber, harvested in June), w hile cassava roots take eighteen months to mature sufficiently to be used as food. A gain, when a village m oved into a new area its members used to borrow portions o f the standing crop in cassava gardens from villages established there, cultivating extra land for the donors the follow ing year. O r they m ight exchange meat, honey or craft products for cassava until their first crops had matured. Villages arc rarely built in com plete isolation from neighbours. Map 2 shows h ow villages tend to be grouped in discrete clusters, o f varying numbers and form ation. Each such cluster is called a chitung ili (from ku-tung a, * to build *). This term m ay be translated as a * vicinage * or * neighbourhood cluster *. T he vicinage has certain jural, econom ic and ritual functions w hich w ill be discussed in subsequent chapters. Here I w ish to examine it in relation to village m obility. Vicinages are not spatially defined areas w ith permanent boundaries. The vicinage is not an objectively based enduring unit, in this respect it differs from a traditional senior headman’s area or a m odem subchiefHom. A vicinage is a cluster o f villages, o f changeable territorial span, and fluid and unstable in social com position. It has no recognized internal organization w hich endures beyond the changes in the identity o f the villages m aking it up. B ut it is not just a neighbourhood round any village. As I w ill show in Chapter Eleven, the vicinage becomes visible as a discrete social entity in several situations, and a particular headman within it usually exercises m oral and ritual leadership. Villages in a vicinage, how ever, do not m ove as a unit, but each village in its ow n tim e and to its o w n site either w ithin the same, or another, vicinage. The frequency w ith w hich villages change their sites means that the com position o f vicinages is constantly changing, new villages com ing in from other areas and old villages m oving out. It means that each vicinage is sociologically heterogeneous, few o f its villages having mutual ties o f lineal kinship or even originating in the same chiefHom. For example, in the small vicinage in w hich Nyaluhana is the most im portant
48
Schism and Continuity in an A frican Society
headman 1 there are eleven villages and farms w ithin an area o f about tw elve square miles. Nyaluhana was founded b y the sister o f a Kanongesha w ho reigned about 1880. O ne un registered village, W ukengi, has recently split from Nyaluhana and has built about a quarter o f a m ile aw ay. A m ile aw ay is W adyang’amafu, w hich cam e in the nineteenth century from the Lunda-Kosa chiefdom o f Nyam wana, in the Belgian C on go. T w o miles aw ay is the K aw iku village o f Kafum bu, w hich broke aw ay from M ukanza V illage in another vicinage. A m ile and a h a lf on the other side is Sampasa w hich came from C h ie f Chibw ika’s area fifty years ago. Four miles aw ay in another direction is Machamba, w hich recently broke aw ay from M w enilunga V illage, founded b y a son o f the first Kanongesha m ore than tw o centuries ago. N ear Kafum bu are Sawiyem bi and M ukom a Villages w hich cam e from senior headman Chikeza’s area in A ngola thirty-five years ago and separated from one another recently, the Sawiyem bi lineage having originally been slaves o f the M ukom a lineage. N ear them again is Nyampasa, recently founded b y a wom an-headm an m atrilineally related to the K aw iku senior headman N sang’anyi, whose village is in another vicinage. A few small forms have detached themselves from some o f these villages and built at some distance from their parent villages. Num erous ties o f affinity and kinship n ow inter link the earlier established o f these villages but few belong to the same maternal descent groups. Villages w hich form erly in habited this vicinage have now settled elsewhere. O ne village in the vicinage is already m aking plans to m ove for to the south o f the D istrict. In the rem oter, less stabilized areas I have visited, I have seen m arked changes in the com position o f vicinages between 1950 and 1954. Fission o f Villages Ndem bu thus live characteristically in loose and tem porary collocations o f small neighbouring villages. B oth size and m obility are influenced b y the fission o f villages w hich divides larger villages into small units. I f fission has been accompanied b y violent quarrels the seceding group often builds for from the 1 See also Chapter N ine fo r a detailed analysis o f the social com position o f another vicinage.
The V illage : Topography and D em ography
49
parent village. T he sociology o f the situation o f fission w ill be discussed at length below : here it is proposed to examine its demographic effects only, the w ay in w hich settlements w ith a com m on origin have been dispersed after fission across the Ndem bu region. Chibw akata V illage has already been m entioned in connection with village m obility. This village in the course o f its history o f perhaps tw o hundred years has ranged w id ely across N dem bu, and even into Kosa, territory. D uring this tim e it has given rise to no less than tw enty further villages b y fission, only one o f which is n ot still in existence today. Som e o f them have given rise in their turn to further villages. In 1953, 3 o f these offshoots were in Kanongesha Area, close to their parent village, 3 w ere in Chibw ika Area to the south, 2 w ere in N yakaseya and 1 in Ikelenge Area in the pedicle, a total o f 9 villages in the N dem bu region.1 In the K osa R egio n , w here Sailunga is the Senior C h ief under the N ative A uth ority, 5 are in Sailunga Area proper and one village has broken up there, 2 are in the form er C h ie f M pulum ba’s Area, and one is in K akom a Area. O n e village is in N yam w ana A rea and another in Kazem be Mutanda’s Area in the Belgian C o n go , a total o f 11 in Lunda chiefdom s outside the N dem bu R egion . O ther long-established N dem bu villages have similar histories o f fission. There is plenty o f evidence to show that the rate o f fission has increased in the last three decades. A ll the K aw iku villages, containing descendants o f the aboriginal population, claim to have split o ff from N sang’anyi V illage either b y direct fission or b y fission from one o f its offshoots. I have collected inform ation in all the extant K aw iku villages about the period at w hich they broke o ff from N sang’anyi or one o f its derivatives. This inform ation has been checked against genealogical data and cor relations have been made w ith historical events in the D istrict the dates o f w hich are know n. W hereas during the w hole o f the nineteenth century there w ere o n ly five remem bered instances o f the form ation o f new K aw iku villages, there w ere no less than five instances in the 1920’s, three in the 1930’s, and eleven, since 1940, m aking nineteen in the first h a lf o f the tw entieth century. O nly four o f these have been * forms *, all founded since 1947. 1 See Map 2 for these chiefdoms, p. xxtv.
50
'D o ;p
:b
C ') l.;
C;
V (;
./v../• t-r {
Schism and Continuity in an African Society
Thus in the w hole o f the nineteenth century five villages hived o ff from one village w hile in h a lf o f the twentieth century nine teen villages have hived o ff from five villages. I f this rate con tinued it w ould mean that at the end o f the century nearly eight villages w ould come from each village, an increase o f 37*5 per cent. It is, o f course, possible that other K aw iku villages came into existence, only to break up again, during the nineteenth century, but careful enquiry did not reveal that this had been the case. In the nineteenth century m y informants say that several previously independent groups o f K aw iku came together in quite large stockaded villages (which also contained some Ndem bu) for self-defence. The K aw iku, in contrast to the im m igrant N dem bu, have tended to remain in the vicin ity o f the K aw iku Plain where they w ere first encountered b y the latter. O f the 25 villages 14, or 56 per cent, still reside w ithin a few miles o f the Plain, although their local m obility has been high. As the result o f a violent quarrel in one village, M ukanza, many villages seceded and later form ed three villages far from the K aw iku Plain. Several younger men have established farms in Ikelenge Area in the pedicle where they carry on trade. B ut b y and large the K aw iku have tended to remain in approxim ately the same area over long periods o f tim e. Each K aw iku village has m oved its site m any times but always w ithin a m uch narrower orbit than the m ajority o f true N dem bu villages. K aw iku and N dem bu villages are interspersed in several vicinages. Fission m ay occur w ithout the seceding group’s departure from the vicinage o f its parent village. This type o f fission is becom ing m ore com m on today w hen one sometimes finds a village surrounded by a num ber o f farms built b y enterprising younger men w ho desire a measure o f econom ic and political independence from the old conservative headmen. There has been no open breach in these cases and the older m en are beginning fatalistically, i f querulously, to accept the new order o f things w hich is rapidly com ing about w ith the spread o f cash econom y. U ndoubtedly, the accelerated rate o f fission in recent years is related to a considerable increase in population in the District. This increase is the product o f several factors. Immigration from A ngola, and, to a lesser extent, from the Belgian Congo into northern and western M w inilunga is perhaps the most im portant o f these. The ending o f the slave-trade and
T h e V illage : Topography and D em ography
‘ .4
4 -:
51
prohibition b y the Governm ent o f internal blood-feuds are other causes o f die increase. It is also possible that the introduction by missions and Governm ent o f hospitals and dispensaries has led to a reduction in the death-rate. O n the other hand, the virtual exterm ination o f gam e in som e areas, a source o f meat not yet replaced b y cattle w hich could be kept in the fly-free north west o f the D istrict, must have led to a decrease in the consum ption per capita o f anim al protein. This lack o f meat m ay have en feebled resistance to disease and so prevented a m arked reduction in the death-rate. B u t the im m igration from neighbouring territories o f small discrete groups o f kin w ho settle in every part o f the Ndem bu region has undoubtedly provided a stimulus to m ore rapid fission among villages already established there. Cash econom y tends to destroy ties o f corporate kinship within villages. M any small im m igrant groups, each living separately, provide for established villages external models for a new type o f residential group. B ut it must be stressed that these m odern trends do no m ore than accentuate tendencies inherent in the indigenous social system . The local kin group appears never to have been extensive, the spatial m obility o f villages was alw ays considerable,1 fission was relatively frequent and individual m obility was high. M odem changes have not so far struck at the basis o f the indigenous system nor radically disrupted it. Because o f the m obile fissile nature o f its traditional residential units, the system has been able to absorb changes that prom oted m obility and increased the rate o f fission. Individual Mobility In addition to the high rates o f m obility and fission o f villages themselves, individuals also tend to circulate rapidly through villages in the course o f their lifetim es. T h e sociological factors behind this m obility o f individuals w ill be exam ined in detail in later chapters, but they must be m entioned at this point. M en and w om en have different patterns o f m obility. M ost men are bom in their fathers* m atrilm eal villages and some o f 1 See, for example, Livingstone’s remark that * people change from one part o f the country to another * with frequency, quoted in a footnote in Chapter One, p. 5.
52
Schism and Continuity in an A frican Society
them go to their ow n mothers* brothers* villages in their teens. B ut if, as frequently happens, their parents are divorced, they go w ith their mothers in early childhood to their mothers’ brothers’ or mothers’ mothers’ brothers* villages. Some m en remain w ith their fathers until the death o f the latter. I f a village splits, men often have a choice o f m atrilineal villages in w hich to live. T he m ajority o f adult men, how ever, reside avunculocally, unless they found new villages o f their ow n. W om en m ove m ore often than men. T hey are bom in their fathers’ villages ; they m ay return to their uterine uncles’ villages for a w hile ; and they g o to their husbands’ villages after m arriage. Since marriage is brittle am ong N dem bu,12 *they m ay spend their reproductive period alternating between their fathers’ and mothers’ brothers’ villages, and successive marital villages. M any w om en, if w idow ed or divorced after they have passed the menopause, return to the villages o f their nearest m atrilineal kin to remain there for the rest o f their lives. Thus the com position o f a single village varies greatly from year to year as men, wom en and children enter and leave it. This natural fluctuation is supplemented b y other movements dependent on the personality o f the headman. I f he is notorious as a sorcerer, or is unskilful as a ju d ge, or mean and selfish, m any people w ill leave him . I f he has the reputation o f being a w ise and generous man he w ill build up a large village. In Tables III and IV , I attem pt to express certain aspects o f individual mobility.® Table HI shows minimum m obility. It is constructed from m y village census figures and offers an indica tion o f w hat proportion o f adult m en and wom en live in their natal villages. These figures suggest that on ly one-seventh o f adult men and w om en live in the villages where they are bom , w ith no sig nificant difference between the m obility o f men and wom en. H ow ever, since N dem bu are often born in one village and reared in another w e can obtain a further index o f m obility b y setting out a table w hich indicates the proportion o f adult men and 1 See Table VII for divorce ratios. 2 The layout o f these tables is based on Table I in E. Colson’s 4Residence and Village Stability among the Plateau Tonga’, T h e Rkodes-Livingstotte Journal, xii (i 9 5 i), P- 4 Z*
The V illage : Topography and D em ography
53
f women w ho have been reared in their natal villages. B y * village o f rearing * I denote for w om en the village in w h ich they have spent the longest period o f residence from birth to the puberty ritual (Nkanga) w hich marks the entrance into w om anhood at ¿bout the age o f fifteen or sixteen. For m en I have arbitrarily selected a sim ilar age-range, birth to sixteen years o f age, since TABLE III I n d iv id u a i . M
o b il it y
: N
atal an d
P r e s e n t V il la g e s ,
M ales Age
20-39 years 30-39 y ears
40-49 years
Living in natal village
Living in other village
11 (22-9%) 8 (21-6%)
(7 7 -i%) 29 (7 8 -4 % )
60-69 years 70-
years
Total .
Fem ales Living in natal village
48 (lOO'O) % 37
(i 7 -o%)
(8 3 -0 % )
(iao-o%)
5
50 (9 0 -9 % )
55 (100-0% )
9
(9 *1% )
(l0 0 - 0 % )
(18 -7 % )
(10-0%) 0 25 (0-0%) (100-0%) 10 2 (*
29 (100-0%) 25 (1000%) 12 (100-0%)
152 (84 *9 % )
(100*0%)
3
25
3
(89 -3 % ) 26 (90-0%)
27 (is -j% )
L ivin g in other village
Total
(ioo-o%) 28
(10-7%) 50-59 years
37
b y age
179
6
44
Total
53
(12*0%) 2 (9 -1% ) 2 (20-0%)
26 (8 1 -3 % ) 22 (88-0%) 20 (9 0 -9 % ) 8 (80-0%)
(100-0%)
37 (I3 ’ 7 % )
170 (86 -3 % )
I97 (ioo-o%)
3
32 (lO0‘O% )
2S
(100-0% )
22 (l0 0 "0 % ) IO
the boys* circum cision ritual (Mukanda) is often perform ed fo r boys o f ten and under and is not com parable fo r our purpose to the g irl’s puberty ritual w hich is usually follow ed im m ediately by m arriage. T able IV , therefore, shows the com parative pro portions o f those w ho have been reared in their natal villages and those w ho have been reared in other villages. Table IV shows that less than tw o-fifths o f adult m en and rather under a h a lf o f adult w om en w ere brought up in their natal villages. The slightly greater m obility o f boys than girls is probably due to the fact that boys, after circum cision, m ay leave their fathers’ villages for their maternal uncles* villages, o r vice versa, more or less at w ill. Girls, on the other hand, tend to remain w ith their
54
Schism and Continuity in an African Society
mothers, or to m ove when they m ove. A t circum cision a boy is sym bolically severed from his m other ; henceforth he is a full member o f a male m oral com m unity, free on the w hole to choose his site o f residence. B ut a girl does not lose her dependent status until she has borne several children. She passes from mother’s to husband’s village when she leaves m other fo r husband. Thus TABLE IV INDIVIDUAL MOBILITY : NATAL VILLAGE AND VILLAGE OF REARING, BY AGE Females
M ales Age
Reared in natal village
20-29 years
11 (22-9%)
30-30 years
15
40-49 years
(40 *5%) 10 (35 *7 % )
50-59 years 60-69 years 70-
years
Total .
.
Reared in other village
Total
37
48 (100*0%)
<77 *i%) 22 (59 *5% ) 18 (64 *3% )
37
14
15
(4 8 *3% ) (53 *0 % ) 6 (50*0%)
(51 *7 % ) 12 (48*0%) 6 (50*0%)
(100*0%) 28 (1000%) 29 (100*0%) 25 (100*0%) 12 (I00'0%)
69 (38*6%)
xxo (61*4%)
179 (100-0%)
13
Reared in natal village
Reared in other village
21 (39 *6 % ) 23 (41*8%)
(6 0 *4 % )
32 32
Total
53
(100*0%) 55
13
9
(100*0%) 32 (100*0%) 25 (100*0%) 22
(5 9 *1% ) 8 (8o*o%)
(4 0 *9 % ) 2 (20-0%)
(100*0%)
92 (4 6 *9 % )
105 (53 *1 % )
(58*2%)
17
15
(53 *x%) 10 (40*0%)
(4 6 *9 % ) 15
(60-0%)
(lOOO% ) IO
197
(100*0%)
w om en have greater individual m obility than men, but boys have greater m obility than girls. T he natal village is in several respects an unsatisfactory point o f departure for a study o f m obility. A person m ay be bom either in his o r her father’s or m other’s village. I f the father’s and m other’s villages are fairly close neighbours the expectant m other is taken to her ow n village, usually the village o f her close m atriiineal kin, a few days before the confinem ent, and brought hack to her husband’s village after the birth o f the child. O r a w om an m ay go to her ow n village for the ku-tena ivumu (‘ m entioning the w om b *) cerem ony, perform ed at a wom an’s
The V illage : Topography and D em ography
55
first pregnancy, and com e back to her husband’s village to bear her child. These customs underline the basically matrilineal character o f the social structure. B ut they have the effect o f confusing inform ants as to w hat was really their natal village — the village w here the m other actually gave birth, or the village where she was residing w ith her husband. A better measure TABLE V I n d iv id u a l M o b il it y : V il l a g e by
R e a r in g AGE
of
and
Females
Males Age
Living in village o f rearing
20-29 years 3 0 -3 9
years
40-49 years
50-59 years 60-69 years 70-
years
Total .
.
IS (3i-a%) 9
(2 4 *3 %) 5
Living in other village
Total
Living in village o f rearing
33
48
3
<68*8%) (100*0%)
(I7 ‘9 %) 2 (6 -9 %) 6 25 19 (24-0%) (76-0%) (roo*o%) 12 12 0 (00%) (100*0%) (100*0%) 37
(20*7%)
P r e s e n t V il l a g e ,
I
Living in other village
Total
SO
53
45
55
(9 4 *3%) (100-0%) (1000%) 30 32 (9 3 *8 %) (100*0%) 20 25 (80*0%) (100*0%) 22 21 (9 5 *5 %) (100*0%) <8i-i%)
(4 *5 %) JO IO 0 (0*0%) (100*0%) (100*0%)
142 21 179 (79 *3%) (100*0%) (10 *7 %)
176 (89-3%)
197 (100*0%)
o f m obility m ay perhaps be obtained b y showing in a table w hat percentage o f adult men and w om en continue to reside in their villages o f rearing, whether or not these w ere their natal villages. Table V shows this. Table V shows that adult men have a stronger tendency to remain in their villages o f rearing than adult w om en. This is probably due to the fact that m arriage is virilocal. W om en are often brought up patrilocally,1 m arry vixilocally, and after divorce reside 1 ‘ Patrilocal * refers to residence in one’s father’s village. ‘ Matrilocal * refers to residence in one’s mother’s village. It should be stressed that my use o f these terms refers to residence with a parent, and not with a spouse.
56
Schism and Continuity in an African Society
avunculocally until remarriage. It should be noted that a much higher proportion o f men under than over fo rty years o f age reside as adults in their villages o f rearing : 25 per cent under fo rty and 14 per cent over forty. This is probably due to two main causes. T he first is that the m ore able and w ealthy men w ho have not become headmen often establish villages or forms o f their ow n, and the second is the transfer o f men from their fathers’ to their uterine uncles’ villages w hen their fathers have died. In m y sample I found that quite a high proportion o f the popu lation had been born, reared, and dw elt at the tim e o f the enquiry, in different villages. 46 out o f 179 adult men, or 25*7 per cent, and 57 out o f 197 wom en, or 28*9 per cent, had been born in one village, bred in another, and lived at the tim e o f the census in yet another. T o sum up : Tables III, IV , and V show that about one-seventh o f the census population lived in their natal villages and sixsevenths in other villages ; that about three-fifths o f adult men w ere reared in villages other than their natal villages, as were just more than h alf the adult wom en in the sample ; four-fifths o f adult men currently resided in villages other than those in w hich they had been brought up, and nine-tenths o f adult w om en did likewise. A quarter o f the adult male population and nearly three-tenths o f the adult wom en had been born, reared, and dw elt at the tim e o f the census, in three different villages. These figures give little indication o f the extent o f spatial m obility o f N dem bu but m erely show transfers o f resi dential affiliation from one headman’s group to another. The few individuals w ho have remained in the same village all their lives m ay have travelled considerable distances across the region as members o f that village. Those w ho have changed their village affiliation m ay have m oved only a few hundred yards. Some idea o f the extent o f individual spatial m obility may be afforded by Table V I, w hich groups the census population into chiefdom s 1 o f birth, rearing, and affiliation at the time o f the enquiry. 1 The term ‘ chiefdom * here refers to a Sub-Chiefdom under the Native Authority. Mukang’ala, however, ceased to be a recognized Sub-Chief in 1947. His area has been incorporated in that o f Chief Kanongesha.
The V illage : Topography and D em ography
57
Table V I shows that im m igration into each chiefdom in w hich the census was taken, from other areas, has been considerable. TABLE VI I n d iv id u a l M
o b il it y t h r o u g h
C h ie fd o m s
■ Present chiefdoms o f individuals in census
Males
P/R/N %
Mukang’ala . Ikelenge . Nyakaseya Mwininyilamba Total
.
.
p/ r
%
27 40*3 2Ó 57*8 15 45*4 14 42*4
67
45 33 33
100*0 100*0 100*0 xoo*o
82
46*1
178
100*0
O
13'4 24-4 33*3 18-3
2 z
13
9 IX 21 6
0
3*0 2*2 3*o 0*0
55
30*9
37
20*8
4
2*2
6
0/ /o
%
43-3 IS-6 18*3 39-3
29 7
Total
p/ h
X
%
Females Mukang’ala . Ikelenge . Nyakaseya Mwininyilamba Total
.
.
40 9 22 5
57-i 17-0 27.3 17*9
6 6 8 3
8-6 i i -3 18*2 10-7
l X X 0
2*4 X*9 2*3 0*0
23 37 23 20
32*8 69*8 52*1 71-4
70 S3 44 28
IOO'O 100*0 100*0 100-0
66
33*8
23
22*8
3
*•5
103
52*5
*9 5
100*0
Males and Females Mukang’ala . Ikelenge . Nyakaseya Mwininyilamba Total p/ r / n f/ r p/ n
o
. — — = —
.
69
50-4 16 16-3 iS 23*4 XS 29*5 X2 I
32-4
Chiefdom o f Chiefdom o f Chiefdom o f Persons bom
15 17 19 9
10*9 17*3 24*7 14-8
3 2 2 0
2*2 2*0 2*6 0*0
50 63 38 34
30*5 64*2 49*3 55*7
137 98 77 61
200*0 100*0 100*0 100*0
do
20*2
7
1*8
185
49*5
373
200*0
birth, rearing and present residence the same. rearing and present residence the same. birth and present residence the same, and reared outside present chiefdom.
The num ber o f im m igrants in the sample is nearly 50 per cent o f the total population. I w ould like to repeat that the boundaries o f chiefdom s have n ot been significantly altered during the
58
Schism and Continuity in an African Society
period under consideration, although m inor changes have been made b y Governm ent. In the past the areas o f chiefs and senior headmen w ere demarcated by topographical features such as rivers and hills. Thus, in considering personal m obility through chiefHoms, I am no longer analysing allegiance to a spatially m obile unit (village), but am exam ining allegiance to a spatially fixed one (chiefdom). Indeed a person m ay meander through m any chiefdom s w hile rem aining the w hole w hile in the same village. T he percentage o f persons bom and/or reared in M ukang’ala area is m arkedly higher than that found in the other three chiefdom s, all o f w hich are in the north-western pedicle where there is m ost European settlement. M ukang’ala is a deposed Sub-C h ief o f Kanongesha, his area is a * conservative * and traditional one, and m any o f his people are K aw iku, des cended from the autochthonous inhabitants o f the region. M ost o f the im m igration into the pedicle has com e from Angola and the Belgian Congo and has taken place since the beginning o f the present century. Summary N dem bu society m ay be characterized as m obile rather than stable w ith regard to residence. Villages m ove through space and split through tim e. Individuals continually circulate through villages. There are tw o main types o f village m obility : some villages tend to m ove in a relatively narrow orbit, fo r instance, K aw iku villages circulate around the m argin o f the K aw iku Plain ; others, especially those o f Lunda origin, sometimes m ove considerable distances between successive sites. Individuals frequently change their village affiliations and few indeed o f the adult members o f m y census sample w ere livin g in their natal villages and villages o f rearing at the tim e o f m y enquiry. Since villages are seldom large, and there is nowhere to get aw ay from one’s neighbours widiiix the village i f one quarrels w ith them, people are constantly on the m ove. W h en villages reach a certain * critical ’ size they tend to split and the resultant immature villages often m ove some distance from the parent villages. I believe the critical sizes to have been, until ve ry recently, in the neighbourhood o f about 20 huts and 50 people ; since 1950 they are probably less. Just as the membership o f villages is con stantly changing so the membership o f vicinages is also changing,
The V illage : Topography and D em ography
59
although less rapidly. T he rate o f residential m obility con siderably increased in the period between the w o rld wars as the result o f im m igration into the Rhodesian N dem bu region from Angola and the Belgian C on go, and recently attempts b y Govern ment to stabilize the population have slow ed d ow n the rate o f village m obility. B u t m y evidence suggests that high village and individual m obility have always been features o f the N dem bu social system. T h e accounts o f ethnographers and travellers show that other groups o f Lunda origin and affiliations possessed a similar unstable type o f social system .1 A ll these groups have traditions o f m igration ; and although they have settled in more or less defined territories in the last tw o or three centuries, w ithin these territories villages and individuals have continued to range w idely, w hile villages are ceaselessly com ing into being. M y genealogical data suggest that m any sm all settlements have failed to becom e established and have died out. This high m obility in the social system m ay be associated w ith a productive econom y in w hich men hunt fo r m eat and wom en grow crops, and in w hich the men’s role is valued m ore highly than the w om en’s. W om en also tend to accept the m ens valuation o f hunting. T he dom inant characteristics o f the social system at any given point in tim e m ay result from a m aledominated hunting econom y in a bush-habitat, the characteristics being a high m obility, a lim ited degree o f co-operation, and, where there is co-operation, this is m ainly between m en. O ther characteristics, consistent w ith hunting, include personal in dependence, resentment o f political control, and residential dis continuity o f groups and individuals. O n the other hand, there are regularities in behaviour w hich are consistent and constant through these changes. R egularity, cohesion and persistence, in the N dem bu system through tim e, are related to the roles o f w om en both in production o f crops and in reproduction. U ndervalued though it is w ith regard to its actual contribution to the food supplies, the productive labour o f w om en provides the regular staple fo r the physical survival o f N dem bu society. W om en bear and nourish the children, and w here the mother goes, w ith o r w ithout the father, the children go also. The 1 See, for example, Carvalho’s {op. cit.) comments on the Chokwe, Bangala and Songo.
60
Schism and Continuity in an A frican Society
present belongs to the men, and die social order at any given m om ent bears the stamp o f male activity ; the past and future o f the society are dominated b y die m others. T he group o f m ale kinsm en form s the core o f the village today : tom orrow its continuity w ill be found to have depended on their scattered m atrilineal kinswom en. The m en are present and together because o f dead w o m e n ; the w om en are absent because they are m arried to other men, but their sons and daughters’ sons wifi replenish the village. T he conflicts and cohesion in the kinship and residential systems, associated w ith this dichotom y o f pro ductive and reproductive roles between men and w om en, w ill be die subject o f subsequent analysis.
C H A P T E R Ili T H E S O C IA L C O M P O S IT IO N O F T H E V IL L A G E have already noted h ow , w ith the weakening o f chiefly W authority, the village em erged as the significant local unit Iff among the N dem bu. B u t w e have also seen that N dem bu viHages, like the T onga villages described b y C olson,1 * are not necessarily enduring units w ith stable populations tied to par li§ ticular localities . . . villages m ay shift from spot to spot, w hile through them stream a succession o f inhabitants \ Nevertheless, :ü ft at*any given m om ent it w ill be found that the inhabitants o f a I li particular village are not m erely a random grouping o f unrelated i individuals but that the m ajority o f village members are linked to the headman b y varying ties o f kinship and affinity. A n Ndem bu village, like the Lam ba villages described b y M itchell and Barnes 2 and the Kaonde village described b y W atson, m ay i be defined in fact as * a local residential group w hose members claim a genealogical relationship w ith one m an, usually the headman In this the N dem bu, o f course, resemble m any other Central A frican peoples. In theory an individual, provided that he or she is n ot a slave, a m arried w om an, an uncircum cised boy, or an uninitiated girl, m ay reside w herever he or she m ay please, but in practice his or .. - her choice o f residence is usually circum scribed by a lim ited fe number o f modes o f residential affiliation, based on kinship, m affinity and slavery. T he m ost im portant o f these is matrilinea! relationship to the village headman. H ow ever, this type o f affiliation is drastically m odified b y the virilocal residence o f married w om en. I f matrilinea! attachment to headmen and virilocal m arriage w ere the sole determinants o f village affiliation, villages w ould consist entirely o f male matrilinea! kinsmen, their w ives and dependent children. A dult men w ould reside avunculocaBy, adult w edded w om en virilocally, and their young
àie
5¡§K í
1 C olson, op . cit., p . 41. 2 Mitchell and Barnes, passim. 3 Watson, p. 7.
If
6x
<52
Schism and Continuity in an African Society
children patrilocally. B ut structural rigidity o f this order is never found in the m obile Ndem bu society. In practice, there is a relatively high percentage o f patrilocally resident, adult men. A number o f adult wedded wom en m ay also be found living patrilocally. In the past, and in the more traditional areas even today, marriage was uxorilocal fo r the first year and daughters remained during this period in their fathers* villages w ith their husbands. Again, young divorced wom en m ay return to their fathers* villages i f their mothers are livin g there. Indeed, divorce pow erfully affects the residential affiliation o f adult w om en o f T A B L E VII D iv o r c e R
a t io s
(T o t a l M
a r riage
E x p e r ie n c e
Ratio
A. No. of marriages ended in divorce expressed as a percentage o f all marriages . . . . B. No. of marriages ended in divorce expressed as a percentage of all completed marriages C. No. of marriages ended in divorce expressed as a percentage of all marriages except those ended by death............................... ..... .
of
L iv in g In f o r m a n t s )
N o. in census sample
Value
347 /d58
52-7
347/429
8o*i
3 4 7 /5 7 6
61-4
all ages. M arriage is extrem ely brittle (see Table VII i ), and older divorced w om en usually go to live at the village o f their nearest male matrilinea! kin until remarriage. A fter the death o f their husbands, w idow s also tend to live w ith their near male matrilinea! kinsmen until remarriage. W hen they have passed the menopause w idow s and divorcees usually reside permanently w ith their male matrilinea! kin. There is a strong tendency for mothers and children to gravitate together as the mothers grow older. The m other-child bond is perhaps the m ost pow erful kinship link in N dem bu society. As minors both sons and daughters tend to fo llo w their mothers wherever the latter m ay be Hving. M ost children go w ith the m other on divorce. Boys in their teens m ay sometimes leave their mothers and remain 1 See pp. 263, 265 for discussion on small percentage o f w idow s.
The So cial Com position o f the V illage
63
with their mothers’ brothers, but girls usually stay w ith their mothers until the end o f the puberty ritual w hich is custom arily follow ed b y m arriage. M others rejoin their sons, daughters rejoin their mothers after w idow hood o r divorce. In their m inority children live w ith their m other ; in her old age the mother lives w ith her children again. In the intervening years mother, son and daughter m ay spend m ost o f their tim e in different villages, although they m ay visit one another frequently. Patrilocal residence and frequent divorce, then, have further modified the picture originally presented o f a village occupied b y a core o f m ale m atrilineal kin, their w ives and children. W hat we now find is a group o f siblings, brothers outnum bering sisters, and som e o f their children o f both sexes. In addition we m ay find spouses o f the above persons, w ith w ives o u t numbering husbands, and some grandchildren. In long-established villages a num ber o f classificatory as w ell as prim ary kin w ill be found. T h e senior generation w ill con tain tw o o r m ore groups o f m atrilineally related siblings and the junior adjacent generation w ill contain classificatory as w ell as full cross-cousins. It w ill be rem em bered that the mean and median sizes o f Ndem bu settlements in m y genealogical sample are small, io-6 huts and 9 huts respectively. N o w , given the tendency o f full siblings and their m other to gravitate together, and given a tendency for a certain proportion o f adult children o f headmen and o f other im portant elders to reside patrilocally during the lifetim e o f their fathers and w hile their parents* marriages remain unbroken, this means that in m any long-established villages and in m ost recently established villages prim ary kin o f the headmen outnumber their classificatory kin. Thus in a recently estab lished village containing, say, 8 huts, the headman m ay be accom panied b y bis brother and w idow ed sister. I f he has tw o w ives each o f them w ill have a separate hut, so that the sibling group in the senior generation w ill jo in d y possess 4 huts. The head man m ay have a married son livin g w ith him , and his sister a married son and divorced daughter, m aking three huts for d ie jun ior adjacent generation, and seven in all. T he other hut m ight be occupied b y a male or fem ale parallel cousin o f the headman, b y the son or daughter o f another sister residing else where w ith her husband, o r by a sister’s daughter’s child. O r
64
Schism and C ontinuity in an A frican Society
it m ight be occupied b y a member o f a maternal descent group o f slaves 1 ow ned b y the headman’s lineage, or even b y a stranger. Ndem bu give different w eighting to each o f the factors affecting village affiliation. M atrilineal relationship w ith the founder o f a village gives one the righ t to reside there perm anently, other things being equal. A person m ay stay in his or her father’s village until the latter’s death, but must then leave. A free father’s slave children are inherited b y his heir and thus remain. This means that one’s affiliation through one’s m other is to a group, whereas one’s tie w ith one’s father is personal. A man is the end o f a m atrilineage, to paraphrase the Latin m axim , and his son does not belong to that lineage. A wom an’s real hom e is where her close m ale maternal kin (her m other’s brothers, her brothers and her sons) happen to be living, usually together. These male members o f her lineage transfer her econom ic and sexual services to a m ale member o f another lineage, but never renounce their claim on her reproductive capacity. Eventually her sons and daughters’ sons must go to them, and sooner or later she w ill probably fo llo w them. B ut she spends m ost o f her reproductive cycle in the village o f her husband or in the villages o f her successive husbands. In Table VIII an attempt is made on the basis o f data supplied in the village census sample to discover whether m inors, under the age o f tw enty, do in fact tend to reside w ith their mothers wherever the latter m ay be living. Table VIII shows that about two-thirds o f the population under tw enty years o f age, other than married girls living viri— locally, w ere residing w ith both their parents, and about a quarter w ith their mothers only. A negligible percentage were living w ith their fathers only, and hardly m ore w ere livin g in their matrilineal villages w ith neither parent. M ore than ninetenths were residing in the same villages as their mothers, and m ore than h a lf o f those w h o were not, w ere residing in villages whose nuclei w ere their ow n matrilineal kin. N o w let us try to find out at w hat period in the lives o f sons this strong attachment to the mother is severed. O n ly a small 1 Slavery was formally abolished by the British South Africa Company in the early years o f the present century. But it lingered on as a clandestine institution in the villages for many years afterwards. See p. 187 if.
© The Social Composition o f the Village
63
■ T: sample could be obtained b y com paring genealogical and census data, but the results are suggestive.
e s id e n c e o f
U
©
©
©
TABLE VIII R
o
©
M in o r s
nw edded
© A ge groups
©
x viai
if
W ith whom residing
5-9
0 -4 N o.
10-14
O / /o
N o.
0/ /o
N o.
0/ /o
N o.
%
N o.
/o
39*8 9*7
26*1 7*2 2*9
24 10 2
32*4 * 3*5
15
o-o
18 S 2
2*7
9
0
31-2 18*7 0*0
98 34 4
3 3 '3 11*5 **4
10
3*4
xo 0 0
o-o
5
7*2
2
2*7
3
6*2
T o ta l.....................
St
49-5
30
43*4
38
5 **3
27
561
Females : Mother and father . Mother only Father only . Neither ; in village o f matrikin . T o ta l..................... Both Sexes: Mother and father . Mother only Father only . Neither; in village of matrikin .
©
*5 -*9
: Males: Mother and father . Mother only Father on ly. Neither : in village o f matrikin .
©
14O 4 9 *
© © ©
© ©
© o
39 13
O
O
37*9
12*6 o-o 0*0
52 505 80
7 7 .7
23 22-3
27*0 16*2 1-4
9
16*7 i8*8
0*0
20 12 1
2
4*2
1*5
3
4-1
2
4*2
$66
3d 48*7
21
43*9
66*7
44 59*4 23
28 xo 0
40*6
X 39
46
14*5
8
8*7
T o ta l ..................................................... j 103 100*0
69
100*0
74
©
50-4
©
©
5
16
5*4
©'
100*0
48
100*0
294
XOOO
5
6
2*0
10*4
3
0*0
©
6*8
2*9
0
©
2
I8
0*0
148
xs*o 1*0
o
29*7 4 *i
0
<$
32-4
47*9 *93 37*5 7« 2
22
21*7
15 2
95 44 3
65*7
©
© ©
A ccording to Table IX , 21 out o f 37 men in the 20—29 age group whose m others w ere still alive w ere livin g in the same villages as their mothers. In the age group 30-39, 5 out o f 11 m en whose m others w ere alive w ere livin g in the same v il lages as their m others. In the w hole sample rather m ore than
© © # #
©
66
Schism and Continuity in an A frican Society
h a lf o f the men whose mothers were alive were living with them in the same village. According to m y observations the mother-son link o f common residence is on ly definitely snapped by the m other’s death. A divorced or w idow ed m other o f advanced years w ill usually spend her last years in her son’s village. I have recorded a case in w hich an elderly wom an actually divorced her husband in TABLE IX R
e sid e n c e o f
A d u lt M
a les w h o se
M oth ers
a r e st il l a liv e
A ge groups Total
Men in same villages as own mothers
20-29
30-39
40— 49
50-59
Married Unmarried .
i8
3 2
4 0
2
27
3
O
5
Total
21
5
4
2
32
13 3
6
O
4 0
0 0
23 3
Total
i <5
6
4
O
26
Married Unmarried .
31 6
9 2
8 0
2 0
50 8
37
11
8
2
SS
Men in different Married villages from Unmarried . own mothers
Total
Total
.
.
.
order to take up residence w ith her eldest son, w ho had just founded a village o f his ow n. This mother-son tie o f propin quity is all the m ore remarkable when the countervailing ten dencies set up by virilocal marriage o f w om en are considered. The m other-child, and especially the m other-son bond, emerges from these enquiries as the crucial link o f Jdnship and determinant o f village affiliation. W here adult children reside patrilocally this seems to be on account o f the child’s attachment to the m other rather than to the father. In Table VIII only 2*4 per cent o f the 0-19 age group in the sample resided w ith their
The So cial Com position o f the V illage
67
fathers only, w hile 26-5 per cent resided w ith their mothers only, follow ing their divorce or w idow hood. O f the fou r m en w ho resided w ith their fathers only, I knew tw o very w ell, and in both cases special circumstances w ere involved. The mothers o f both w ere dead and their fathers w ere headmen. These factors are, o f course, infrequently com bined, h i one case the son feared w itchcraft in the village o f his near m atrilineal kin. T h e matrilineal village o f the other was in A ngola and as a progressive young man w h o wanted to earn m oney he preferred to remain in M w inilunga in the w ealthier pedicle area. The strength o f the sibling bond and its im portance as a deter minant o f residence also derive from this m other-child tie. In matters o f inheritance and succession it is m atrilineal descent that counts ; the children o f one m other b y several fathers regard each other as fu ll (chikupu) siblings in every respect in the social context o f their maternal village, although they m ay visit in their different paternal villages, and receive assistance from and assist their father's m atrilineal groups independently o f one another. A s the analysis proceeds, it w ill be shown h ow a number o f social mechanisms are brought into play to counteract this tendency for the m inim al unit o f m other and children to cohere too closely and intensely and prevent the grow th o f wider social units. Im portant am ong these— to anticipate— is an emphasis on the social and spatial separation o f adjacent genea logical generations and on unity w ithin each genealogical generation. This emphasis opposes m other to child in a relation ship o f authority on one side and respect on the other, and unites full siblings w ith parallel and cross-cousins against the senior generation as a w hole. T he persistence o f the m other-child link as a determinant o f residence attests to its strength in the face ° £ j e” eration separation and o f other com petitive modes o f affiliation. W e are n o w in a better position to rate the various principles determining village affiliation according to their degree o f em phasis in N dem bu culture. First, there is, as w e have seen, an extrem ely pow erful bond betw een a m other and her ow n children, acting against the dispersal o f this m atricentric fim ily and towards its maintenance as a residential unit. B u t the principle o f virilocal m arriage disrupts the m atricentric fim ily b y separating
68
Schism and C ontinuity in an A frican Society
daughter from mother and sister from brother. Henceforth it is possible to regard the subsequent history o f the original matricentric fam ily as an effort to reconstitute itself as a local unit. M any sons remain w ith their mothers whether the latter live virilocally or avunculocally. I f sons and mothers are tem porarily separated, divorce or w idow hood o f the mother sooner or later brings the old w om an back into the village o f her son w ho b y this tim e m ay be fo lly mature, and w ho may indeed have founded a new village o f his ow n. Sisters return to brothers after their marriages have terminated and m ay often fold their mothers there. T he m atricentric fam ily, united in its early grow th but divided as the children grow up and themselves reproduce, seldom comes together again in its full membership, but it always strives to do so against virilocal marriage and the m any other centrifugal tendencies o f N dem bu society. Another important factor influencing residence is, as w e have seen, the tendency o f males to com e together in kin groups to form the persistent core o f villages. T heir collaboration is couched in the dominant matrilineal idiom o f Ndem bu kinship so that the village is regarded at its inception as a group o f uterine brothers. B ut a w ider group than a pair or handful o f brothers is required .for effective collaboration in economic activities and in legal matters, and for offence and defence in feuds and raids. The concept o f uterine brotherhood is extended to include in the residential unit, m atrilineal parallel cousins. V irilocal marriage, w hich had broken up the matricentric fam ily, n ow appears as the ve ry means b y w hich the * adelphic group * o f male uterine parallel cousins is enabled to remain together. U xorilocal m arriage w ou ld disperse this group,1 1 Among the Nayar castes o f Malabar in India, sisters remained in their brothers’ house. The Nayar kinship grouping was a matrilineal lineage living on its own property. All Nayar women were married before puberty to men o f their own caste and divorced after four days, so that children they bore subsequently to lovers o f their own caste or higher castes would have the correct caste on the whole for ‘ father \ Thus uterine siblings and matri lineal parallel cousins remained together throughout life, for though each brother might have several * mistresses * in other houses, he resided with his sisters. In a sense, the frequent divorce among Ndem bu women parallels the Nayar situation, since there is hardly a village in which, at a given moment, one or more sisters are not found spending the intervals between successive marriages with their uterine brothers anti matrilineal parallel cousins.
T h e So cial Com position o f the V illage
69
primary uterine kinship as the basis o f local association w ould prevent its classificatory extension. T he unity o f the m atricentric fam ily, virilocal m arriage, and classificatory adelphic co-residence, are principles o f local group ings w hich give rise to continual conflicts w ith in the village. W e have seen h o w the ecological system o f the N dem bu, w ith its com plem entary and opposing poles o f hunting and cassavacultivation, m ay itself be a direct source o f residential instability. This instability is considerably aggravated b y conflicts w hich arise out o f the social structure. In subsequent chapters I shall examine in turn each o f the principles determ ining residence and show h o w each interlinks persons and groups divided from one another b y other principles o f social affiliation. In the course o f the analysis it w ill be shown, fo r exam ple, h o w conflict between the principles o f the unity o f the m atricentric fam ily and virilocal m arriage tends to break up the m arital group, especi ally in the early years o f m arriage, and on the other hand, h ow stable m arriage tends to retard the grow th o f villages b y keeping men apart from their uterine sisters w ith their children. It is to a considerable extent b y divorce and w idow h ood that a village is enabled to persist through tim e. W ith a w om an com e her children, and i f she is divorced or w idow ed after she has passed her menopause she and her sons com e back perm anently to the village o f their m atrilineal kin and replace the w ives and children o f their m ale kin w ho have gone out o f it. Thus divorce and w idow hood act sim ultaneously as principles o f village recruit ment and attrition. W hat are the visible effects o f these principles o f local organ ization on the social com position o f villages that I have actually observed ? Is it possible to infer regularities in residential com position from field data ? Is it then possible to construct models o f different types o f village structure on the basis o f such observed regularities ? A n attem pt to answer these questions is provided in Tables X and X L T h e universe on w hich these tables are based consists o f hut-ow ners in villages w here I have collected full genealogies. O f the 77 villages in w h ich I collected genealogies (see p. 37), I have used m aterial from <58 villages fo r the purpose o f m aking these tables. T he nine discarded villages are those in w hich the genealogical data are too scanty and fragm entary to G
V
0 0
©
@ © & e
©
0 © © © @
© © © # « # ©
70
Schism and Continuity in an A frican Society
i
be suitable for use. Since in the great m ajority o f villages most .5 members are kin o f the headmen I have classified hut-owners according to the kind and degree o f kinship or affinal links they have to headmen. W here a person has m ore than one tie o f consanguinity or affinity to the headman o f his village I classify him b y the relationship w hich N dem bu hold to be the dominant one in the context o f village membership. In establishing which m em ber o f a m arital pair or o f a group o f im m ature boys or girls inhabiting a single hut is the ‘ hut-ow ner ’ I em ploy the criteria used b y Ndem bu themselves. I f a w om an is livin g in a village where she has a prim ary or classificatory m atrilineal rela tionship w ith the headman, and her husband has not, she is considered to be the hut-ow ner. I f both are matrilineaJly related, as in the case o f a marriage between m atrilineal kin o f alternate genealogical generations, the husband and not the w ife is taken to be the hut-ow ner. I f a w om an is livin g in her father’s village and her husband is not related or is a distant cognatic relative o f the headman she is the hut-ow ner. In a cross-cousin marriage, the spouse m atrilineally linked w ith the headman, regardless o f sex, is die hut-ow ner. In a boys’ o r girls’ hut the oldest, linked b y w hatever tie o f kinship to the headman, is classified as the hut-ow ner. I f no link o f consanguinity or affinity can be traced between a hut-ow ner and his or her headman, he or she is classi fied as a * stranger ’. Strangers fall into tw o categories. They m ay be first-generation slaves (sing, ndungu, pi. andungu) or maternal descendants o f slaves, w h o even today m ay still be regarded as belonging to free members ; or they m ay be im m i grants from other areas in Rhodesia, A ngola or the Belgian C on go w ho have attached themselves or have been attached by a Governm ent Sub-C hief to a particular headman. I have not been able to determine in every case w ho was or was not a slave and have consequently been com pelled to classify all hut-owners fo r w hom it was impossible to trace genealogical connection w ith the headmen o f their village as 4 strangers \ I have called a fem ale stranger a * hut-ow ner * i f she is so described b y the headman o f the village in w hich she is living. B y em ploying the descriptive system o f kinship nomenclature used b y antnropologists I was able to distinguish no less than 7z different categories o f kin, in relation to headmen, living in these 68 villages. These are subsumed in the Ndem bu
TABLE X R
e l a t io n s h ip o f
H u t- O w ners t o H ea d m en (male and female ) Primary
Category
Ego (headmen themselves) mm m
Matrilineal kin
3Ä
mG older G younger G zC zdC zddC Total
Own descendants of male headmen
. . . C dC SC
Total
Children of male matrikin
mBC older BC younger BC BC (unspecified sibling rank) zSC Total
Patrilateral siblings
. . .
older FC younger FC Total
IS -
Children cf patrilateral siblings
FdC older FSC younger FSC FddC Total
. . .
T o ta l............................ m Mother G Sibling
in
Classificatory
M.
F.
Total
67
i
68
—
—
— 2
0 O
0 4
— --
13
2
7
—
35 56 II 0 105
4
1
71 176
83 58
70
53
23
76
3 3
0 3
6
0 2
Z
3
I
84 20 0
23
21 2
49
M. F.
6 0 16 7 28 9 26 22 5
9 58
28 9 0
—
—
3 3
0
68 S et t l e m e n t s
—
Total
M.
F.
Total
67
i
68
i 16 0 14
i 16
23
— 7 18
37
<53
32
32 95
188 X29
317
Total —
i 12 6 48
II 3
141 —
—
9
3 0 0
2 0 0
0 0
0 0
0 2
6 3
1 3
7 6
17
3
20
12
6
6
i
0 2
8
4
2
7
2
9
2 3
I 0
1 J 0 0
3 4
6
2
9
0
255 102
z Sister C Child
3
82 16 2
7
50 132 15 31 I 3
49 I 3
21 2
70
0
3 3
S3
23
76
6 3
3
8
6
9
6
I
7
5
3
18
29
9
38
i 0
3 4
3 10
ï 2
12
6
i
7
13
3
I6
0
z 0
0 0 1
0 0 0 0
0 0 0 1
2 3
I I
I * O 0
3 4
8
i
0
I
7
2
9
102 65
167
3 S7
l6 7
524
357
d Daughter S Son'
9
2 0
I
L egend
5
—
8
4
I I
B Brother F Father
72
Schism and Continuity in an A frican Society
classificatory system under 14 kinship categories, or 18 i f cer tain terms are qualified b y adjectives denoting sex. In Table X I discriminate between matrilineal and patrilateral prim ary kin according to descriptive criteria, and subsume all other kin under N dem bu classificatory categories. A number o f inferences m ay be drawn from the material presented in Table X I. The first is that headmen and their kin outnum ber strangers in villages by m ore than four to one. The second is that headmen and their matrilineal kin account for nearly TABLE XI T h e S o c ia l C o m p o sit io n
of
68 S et t l e m e n t s
Males Category o f hut-owners
Headmen and primary matrilinea! k i n ......................................... Classificatory matrilineal kin Own descendants o f male headmen Children o f male matrilineal kin . Patrilateral siblings and their children Unspecified kin and affines
Wo.
0/ /o
N o.
0/ /o
Wo.
%
172 83 53
25’7 12*4
72 58
10-8 8*7
244 141 76
21*1 S’ 7 3;7 2*9
29 20 16
Total kin and affines . . . . Strangers.................................... Total hut-owners
.
.
.
.
.
Total
Females
7-9 4 '4
23
3*4
2-9 2*4
9 5 3
1*2 o*8 o*5
38 25
373 97
55*9 14*5
170 28
25-4
543
470
70-3
198
296
4-2
19
125
36-5
n *4
81-3 18.7
668 ÏOO-O
three-fifths o f all hut-owners in the sample. Children o f head m en and o f their m ale matrilineal kin account fo r ju st under onesixth o f all hut-owners and one-fifth o f all related hut-owners. A nother feature is the numerical preponderance o f prim ary over classificatory matrilineal kin in a ratio o f alm ost tw o to one. 130 out o f 385 m atrilineally related hut-ow ners, or 33*7 per cent, are w om en, an indication in a society practising virilocal mar riage o f the strong tendency o f w om en to rejoin their male m atrilineal kin. Since nearly all w om en are married ju st after their puberty ritual this means that w om en found in their matri lineal villages must have been divorced, w idow ed, or recently separated from their husbands, or w ith husbands w ho live
The S o cial Com position o f the V illage
73
uxorilocally. T he evidence o f m y village census suggests that divorce is easily the m ost com m on reason w h y w om en reside w ith their m atrilineal kin. W om en live in uxorilocal marriage only for special reasons, except that a few men still live uxori locally during the first year o f marriage. T h e reasons behind the m atrilocal and avunculocal residence o f adult w om en w ill emerge in the course o f the analysis. Headmen and their prim ary and classificatory brothers in both lines o f descent account for 161 out o f 542 related hut-owners, or nearly three out o f ten. W ith them are 49 w om en w hom they call * sisters o f w hom 32 are their ow n sisters, m aking up a sibling and classificatory sibling category w hich accounts for tw o out o f every five related hut-ow ners. In this category w e find that tw o out o f every three are prim ary siblings. This tends to confirm the previous statement that uterine siblings rejoin one another in the same village as they gro w older. The table also demonstrates that whereas the ratio o f m ale parallel cousins to brothers is 53 :4 7 , the ratio o f fem ale parallel cousins to sisters is as lo w as 35 : 65, a fact that indicates the importance o f the principle o f classificatory adelphic co-residence. T he bilateral character o f the ju n io r adjacent genealogical generation o f hut-ow ners em erges from the analysis o f the data. Children o f headmen and o f those w hom headmen call ‘ brothers * total 94 persons, w h ile those w hom headmen call * sisters* children * total 134, a ratio o f about 2 : 3. For males the ratio o f * sons * to 4sisters* sons * is about 7 : 8, and fo r ow n sons to o w n sisters* sons approaches one to one. I found in the villages from w h ich the figures w ere com piled, that there was an approxim ate num erical balance between hutowners in the opposed generation categories. Thus i f hut-owners in the headm en’s generation are added to those in the second descendant and second ascendant generations from the headmen, members o f this linked generation category total 269 hut-owners. I f members o f the first and third descendant generations and first ascendant generations from the headmen are added together, this opposed linked-generation segm ent possesses 273 hut-owners. This num erical balance between the linked-generation segments is to be expected, since in a given population, a hut-ow ner is equally likely to belong to either generation-segm ent on a random basis. Each linked-generation segm ent in Ndem bu
D O
b 0 {..' ■ O c O a c C:
O e
lo ,c o
b
C; 'O O iCJ 't "
ÏO ■ O
74
Schism and Continuity in an A frican Society
society occupies its own. section o f the village, and, since villages are generally circular in shape, inhabits a sem icircle o f huts facing the semicircle o f the opposed generation segment. In each village the number o f members in each segm ent is seldom equal, but i f a sufficiently large sample o f villages is taken it appears that the linked-generation categories approxim ately balance one another num erically. In order to bring out significant differences in social structure between long and recently established villages I com piled a further table illustrating such differences. For the purposes o f this analysis I defined * a long-established village * as a village in w hich there had been three or m ore successive headmen, including the present incumbent, and ‘ a recently established village * as one in w hich there had been less than three successive headmen. This distinction does not precisely correspond w ith the distinc tion previously made between registered and unregistered vil lages and farms. According to the earlier distinction both unregistered villages and farms are o f com paratively recent establishment (cf. p. 34 and p. 36) since they consist o f settlements not y e t recorded on the Governm ent T a x R egister. Thus by this earlier reckoning I included as * registered villages ’ settle ments that had been founded m ore than about fiv e years before the enquiry and w hich had been granted official recognition as tax-paying units. Several o f such villages indeed had been founded m ore than thirty years previously, but each village had been under the authority o f a single headman throughout this period. A long-established village w ith three or m ore successive headmen is, in every case I collected, a village that was in existence before the European governm ent began. There w ere 21 such villages in m y sample and I compare the social com position o f these villages w ith that o f 47 * recently established villages ’ in Table X II. Table X II shows that in long-established villages the ratio o f prim ary to classificatory hut-ow ning m atrilineal kin is about 1 : i *i , whereas in recently established villages the ratio is about 3 : 1 . In other words, recently established villages tend to con sist dom inantly o f uterine siblings and their fam ilies, w hile in long-established villages uterine kin are ju st outnum bered by their classificatory matrilineal kin. In long-established villages also w e find a significantly higher percentage o f cognatic hut-owning
The So cial Com position o f the V illage
75
kin than in recently established villages. This is because such villages are regarded as som ething m ore than a group o f uterine siblings. T heir membership is looked upon as a group w ith a persistent m atrilineal core to w h ich a h in ge o f other cognates TABLE XII T
he
S o c ia l C o m p o sit io n
o p L o n g - E s t a b l is h e d w it h t h a t o f R e c e n t l y E s t a b l is h e d
M ates
V il l a g e s C V il l a g e s
o m p a r ed
Total
Fem ales
Category o f hut-owners
se £» ^ .2 'S
•3 £
N o.
%
N o.
%
N o.
52 53
18*2 l 8*6
23
8*0 n*6 3*2
75
33 9
26*3 30*2 10*2
%
H.M. and primary matrikin. Classificatory matrikin Descendants o f H.M. Cognaric, patrilateral and un specified kin, and affines S tra n g ers..........................
20
70
38 41
9 7
3 *2 2 *4
47
14*4
48
16-5 16*8
T o t a l ...............................
204
71 *6
81
28*4
285
100*0
H.M. and primary matrikin. Classificatory matrikin Descendants o f H.M. Cognatic, patrilateral and unspecified kin, and affines . S tra n g ers..........................
X2 0
31*3 7*8
12*8 ö *5
169
8*6
49 25 14
44-1 14*4
7 *i 14*6
8 21
2*1
56
5*5
35 77
20*1
T o t a l ...............................
¿66
6 9 *s
117
30*5
383
100*0
H.M. and primary matrikin. Classificatory matrikin Descendants o f H.M. Cognaric, patrilateral and unspecified kin, and affines . S tra n g ers..........................
172
25*7 12*4
10*8
7*9
72 58 23
8*7 3*4
244 141 76
21*1 11*4
«5 97
9*7 14*5
17
28
2 *5 4 *2
82 125
12*3
T o t a l ...............................
470
70*3
198
29*6
668
100-0
30 33
27
83 53
13-3
3*7
86 29
55 47
12*3 9-2
30*5
18*7
attach themselves. In these villages the m atricentric fam ily has matured into the matrilmeage. T h e principal factors influencing the persistence o f settlements beyond their initial phase as uterine sibling residential groups is discussed on pp. 198-203. In a matrilineal society w hich practises viriiocal m arriage, i f there is
76
Schism and Continuity in an A frican Society
virtually unlim ited access to resources so that settlements are under n o constraint to occupy specific tracts o f land in perpetuity, the transition from m atricentricity to m atriliny as the basis for local groupings m ust always be hazardous and uncertain. It is dependent on such factors as the ability o f the headman to keep his follow in g together, the maintenance o f reasonably good rela tions between the m en o f the m atrilineal core and their brothersin-law , and the biological accidents o f fertility and freedom from disease. The core o f the N dem bu settlement consists o f a group o f uterine or m atrilineal kin, o f w hom the m ost senior is the head man. V iiilo cal m arriage separates fem ale and m ale matrilineal kin and at the same tim e enables men often to retain their children, especially their sons, in their ow n villages after the latter have matured. Nevertheless the dominant attachment o f children is not to their fathers but to their mothers. Thus a man can o n ly retain the residential affiliation o f his children w hile his m arriage to their m other endures. O n divorce a w om an usually goes to live w ith her nearest m ale matrilineal kin, and her children tend to fo llo w her. T he crucial bond o f kinship is between a m other and her children, w ho form a m atricentric fam ily. This fam ily m ay be attached for longer and shorter periods to a wom an’s husband o r brother. As the children m ature, daughters are separated from mothers, and brothers from sisters, b y virilocal m arriage. B u t there is a constant tendency, though it is seldom com pletely realized, fo r the m atricentric fam ily to recon stitute itse lf as a local unit. This is an im portant factor behind the m obility and instability o f the individuals residence. It is also a cause o f conflict between husbands and brothers-in-law and between sons-in-law and m others-in-law. V irilocal mar riage is related to the tendency for uterine brothers and male matrilineal parallel cousins to reside together. This is part o f a general tendency am ong Ndem bu to build up villages around a fram ew ork o f close male kin : brothers, sons, sistere sons, and sisters’ daughters’ sons. Evidence from matrilineal societies practising uxorilocal marriage such as the Bem ba and Y a o 1 1 Richards, A. L, * Mother-right among the Central Bantu \ Essays Pre sented to C . G. Seligman, ed. E. E. Evans-Pritchard, R . Firth, B. Malinowski,
The So cial Com position o f the V illage
77
suggests that w here the link between male village residents and the headman is one o f affinity, and these men are linked to one another not b y kinship but m erely b y propinquity, they co operate very little am ong themselves. This is probably because in most prim itive societies social control at the local level is associated with position in the kinship structure. Senior kin exert authority over and com m and respect and obedience from their juniors, and betw een each category o f kinsmen custom has prescribed an intricate and specific pattern o f behavioural expectations which facilitates co-operation and inhibits dispute. A m ong Ndem bu, m ale villagers co-operate in hunting and bush-clearing, and form erly assisted one another in offensive and defensive war fare. There was no strong central political authority and no national arm y, and a man’s first lo ya lty was to his village. For all these purposes d ose and ready co-operation betw een men was necessary. It was therefore appropriate that m en liv in g together should be linked b y consanguineal kinship, the m ajor vehicle o f social control. B u t the local attachm ent o f m ale kin probably owed som ething o f its intensity, as w e have suggested in previous chapters, not on ly to such em pirical factors, but also.to masculine resentment o f m atrilineal descent form ally expressed, as w e shall see, in a num ber o f ritual contexts, and the decisive econom ic im portance o f w om en. It is as though there w ere a general though unconscious male conspiracy to exclude their female kin from their local units. Nevertheless, men needed w om en for their personal and corporate survival. T h e individual man ow ed his care and nurture to a wom an, his ow n m other. Sim ilarly the group o f male villagers needed their sisters to ensure the survival o f their village as a social entity. T o bring them back into the village entailed the male group’s com ing into conflict w ith other men. So that the v e ry exclusion o f fem ale kin from villages in order that m ale kin m ight live together provided the principal source o f conflict and unrest between N dem bu men as a w hole. A gain , since in the econom y w om en’s w ork provided the staple o f subsistence, w om en w ere needed in the and I. Schapera (1933), p. 267, and Bemba Marriage and Modern Economic Conditions, Rhodes-Livingstone Paper N o. 4 (1940), pp. 33 fF. ; Mitchell, J. C ., ‘ The Yao o f Southern Nyasaland \ Seven Tribes o f British Centra/ Africa, ed. Colson and Gluckman (195*). pp. 328 fF.
78
Schism and Continuity in an African Society
village. M en ow ned gardens, it is true, and w orked in them sporadically, yet w om en had not only w orked m ore regularly in their ow n as w ell as in their husbands* gardens but also dug up, carried, processed and cooked the daily, never-failing supply o f cassava. I f sisters w ere excluded, w ives had to be included in the village personnel. In all their conscious statements about the relative im portance o f wives and sisters, N dem bu have continually stressed to m e that w ives outrank sisters, and that indeed the m arital tie is the strongest and deepest they know. In ritual, too, husband and w ife are regularly initiated together into curative and fertility cults. In a decisive aspect the girl’s puberty ritual is a drama o f marriage, emphasizing die importance o f the conjugal tie. In spite o f this verbalization and ritualization o f the im portance o f the marital relationship, in fact Ndem bu marriage is extrem ely brittle. This is due to the very real im portance o f the brother-sister tie w hich is the socially, although not biologically, procreative link in matrilineal society. M ale kin can o n ly live together through die exclusion o f their sisters and sisters* daughters and the im portation o f w ives linked to other corporate groups o f male kin.1 B u t having thrust their female kin out, the m en must reclaim these w om en and their families i f the village is to persist through time, and is n ot to die out for lack o f replenishment. Unfortunately fo r the men, their ow n w ives, closely bound to their children, are subject to the same * gravitational * compulsion from their male matri lineal kin. E very adult Ndem bu man, under the operation o f the same set o f social forces, is in the ambiguous position o f striving to retain his ow n w ife and children by his side and o f simultaneously endeavouring to w in back the sister whose absence is the price o f his privilege o f livin g w ith his brothers and sons in a matrilineal society. Put succinctly, to live w ith the male kin he loses a friend to m arry an enemy. It is all part o f the N dem bu male fligh t from the inexorable reality o f matri lineal descent and fem ale control o f the econom ic basis o f sur vival. Perhaps this is one factor in the Ndem bu preoccupation 1 In this discussion I have been considering some o f the factors relating to the question o f why Ndembu are virilocal. I have taken the fact that they are matrilineal for granted. I am also assuming here that there is a brothersister incest rule.
The So cial Com position o f the V illage
79
v with ritual, one aspect o f w hich is escape, and another, compen sation, although these constitute only a m inute part o f the w hole story. There is no guarantee, therefore, that a settlement w ill be per petuated and w ill survive. I f the sisters o f the adelphic group have successful marriages, and providing that their husbands have long lives, it m ay be that their sons w ill live m any years w ith fathers ; and that w hen these fathers die the sons w ill found sibling villages o f their ow n rather than return to their uterine uncles. It m ay be that conflicts am ong the brothers themselves, arising from jealousies over succession and inheritance, w ill break up a village. M uch depends in this society o f individualists, only too ready to take offence, on the tact and diplom acy o f individual headmen, whose best hope is not to browbeat or domineer but to persuade and reconcile. B ut once the essen tial link-up between first and third generations after the founding o f a village has taken place, new sources o f conflict w ithin the setdement com e into being. W hen a village contains only tw o generations o f hut-ow ners it m ay still be regarded as a bilateral extended fam ily. B u t w hen three generations o f adult hutowners form its membership, incipient cleavage along lineage lines becomes detectable. I f a headman has tw o fertile sisters each has becom e the founder o f a lineage. The principle o f the unity o f the m atricentric fam ily com es into conflict w ith the unity o f the m atrilineage, and each m atricentric fam ily or alli ance o f m atricentric fam ilies w ith a com m on grandm other is a potential source o f village cleavage and the potential startingpoint o f a new residential unit. H ere the unity o f siblings and the unity o f classificatory adelphic m ale kin m ay be insufficient to hold the village together b y cutting across lineage affiliation. It m ay be that the original sibling-group w hich founded the village has been reduced in numbers b y death. In any case its authority is continually threatened w ith in the village b y the maturation o f the ju n io r adjacent generation, the leading men o f w hich are eager to obtain the headmanship. T o maintain its authority w ithin the village on the one hand, and to prevent the disruption o f the village as the result o f struggles between lineages on the other, an alliance develops between the senior and the second descendant generations, an alliance sometimes cemented b y m arriage between persons w h o stand in the classificatory
8o
Schism and C ontinuity in an A frican Society
relationship o f ‘ grandfather ’ (nkaka) and 4grandchild ’ (mwijikuiu) to one another.1 W here such marriages occur they often link a * grandfather * o f one lineage to a * granddaughter ’ o f another. Thus the grandparent and his siblings secure the allegiance o f the granddaughter’s siblings against the thrustfui m iddle generation. The m iddle generation, o f course, exercises authority over the generation o f grandchildren and also threatens the authority o f the senior generation. Grandparent-grandchild marriages also bind together potentially conflicting m inim al 23matrilineages and uterine sibling groups. T h ey are on ly made between classificatory kin and never between prim ary grandparents and grandchildren. Thus they tend to occur m ainly w ithin longestablished villages. Nevertheless, in all villages containing at least three generations o f hut-ow ners the alliance between alternate generations, whether interm arried or not, finds physical and spatial expression in the fact that both generations build their huts in the same village semicircle o f huts facing and opposed to the semicircle o f huts built by the intervening genealogical generation. Ï use the term 4alliance * and not 4unity ’ or 4equivalence ’ in describing the relations between alternate generations, since their m ode o f association approaches m ore closely that obtaining between cross-cousins (asonyi, sing. musonyiY w ho interm arry, rather than that between siblings w ho do not. Alternate generations, like cross-cousins, jo k e w ith one another and interm arry, although the jokin g is less ribald and m ore gently affectionate than the pleasantries o f cross-cousin intercourse. Joking, according to R adclifle-B row n, implies a fairly even balance between hostility and friendliness, and in fact tension exists in grandparent-grandchild relations. In the first place grandparents and grandchildren belong to different, and therefore potentially opposed, matricentric families, the basic 1 It should be mentioned here that grandparent-grandchild marriages take place not only within villages but between villages o f common origin (see p. i7<5). Most, but not all, o f such marriages are between a man and his granddaughter * (see p. 246). 2 In this book I regard a minimal matrilineage as a group consisting o f the descendants through women o f a common grandmother, a minor matrilineage as the matrilinea! descendants o f a common great-grandmother, and a major
lineage as those o f a common great-great-grandmother. 3 See Chapter Eight for analysis o f village kinship.
The Social Composition o f the Village
8l
corporate kin units o f N dem bu society and the starting-point o f new residential units- In the second place, grandparents, especi ally the senior m ale grandparent, belong to the generation w hich exerts gérontocratie political and ju ral authority in die matrilineal village or in other villages in w hich the grandchildren m ay reside. A uth ority is ultim ately in their hands. A t the same tim e, the position o f these tw o generations is essentially similar vis-a-vis the m iddle generation as w e have seen above. Again, the older needs the younger generation to replace as allies the siblings w ho have died, and the younger needs the older to m itigate the severity o f the intervening generation o f uterine uncles and senior affines. T heir m utual inter-dependence, strengthened b y m arriage and brother- and sister-in-law ties o f fam iliarity and friendship, acts to bind together a residential group constandy liable to disruption from lineage and adjacent generation conflicts. In the subsequent analysis I shall isolate each dom inant principle governing village residence, and exam ine Its operation, illustrated b y case m aterial, in prom oting cohesion or generating conflict w ithin the local unit. This exam ination w ill concern itself not only w ith secular behaviour but also w ith ritual. Afterw ards, the w orking o f each principle in the field o f inter-village relations w ill be analysed. Finally, there w ill be a discussion o f modes o f social integration, other than those derived from kinship, and these w ill be exam ined in terms o f their congruence w ith and opposition to the kinship principles already considered.
© ©
ô
© © ©
© © ©
© ©
© &
©
CHAPTER IV M A TR ILIN E A L D E S C E N T : T H E B A SIC PR IN CIPLE O F V IL L A G E O R G A N IZ A T IO N H ERE is sense in regarding N dem bu villages as colloca tions o f matricentric families interlinked by varying ties o f Idnship and affinity. In addition, som e measure o f continuity is provided by maternal descent w hich determines succession and inheritance. In long-established villages w ith respected head m en something like hierarchically organized lineages o f the type made familiar in British anthropological literature b y the re searches o f Evans-Pritchard and Fortes begin to em erge, although such lineages are shallow b y comparison w ith those found among such matrilineal and uxorilocal people as the H opi and Ashanti. '1 able XIII shows lineage depth in 64 village genealogies. Farms have been excluded from the sample and also several villages for w hich m y data are unsatisfactory. Table X V appears to indi cate that there is a significant increase in the m agnitude o f a village i f six generations o f matrilineal kin o f its headman can be recorded on a genealogy. Thereafter village m agnitude seems to remain fairly constant regardless o f recorded lineage depth. T he differ ence in genealogical recall between the older villagers o f ninehut and thirteen-hut villages respectively tends to correspond w ith the difference previously made between * long and recently established villages * (see Table X II). It tends to correspond also w ith differences in social structure. R ecently established villages, villages w ith less than three successive headmen, consist mainly o f uterine siblings, their adult children and a few mature grand children. Long-established villages, on the other hand, m ay con tain m any adult members o f tw o or m ore segments o f equal status w ithin a minor or m ajor village m atrilineage (see footnote on p. 80). B u t few long-established villages, whatever the generation depth o f their nuclear lineage, attain to a great population size. For fission, occurring m ost frequently between uterine sibling groups and minimal lineages, and virilocal mar riage, and various kinds o f individual m obility, all prevent their expansion.
T
M atrilineal D escent
83
A lthough there is no absolute correspondence between the lim it o f recall and the span o f the residential segm ent, one tends VILLAGE LINEAGES TABLE XIII L ineage D e p t h
in
N o. o f g en era tio n s in v illa g e g en e a lo g ie s
TABLE XIV
64 G en ea l o g ies
G e n e r a t io n
N o. o f v illa g es
G en era tion o f headm an
4
4
5
6
19 21
7
*4
3rd 4th Sth 6th 7th 8th 9th loth
8
9
10 11 12 Total . . . .
2 2 I
0
of
H eadm an
N o. o f v illa g e h eadm en 15 26 16 4
2 O O I
r Total
64
Mean lineage depth . . Median . Mode . . .
6-1
.
.
.
Mean generation level Median . . . Mode . . . . .
6
<54
. .
4-3 4
.
4
TABLE X V V il l a g e S iz e
and
N o. o f g en e ra tio n s in v illa g e g e n e a lo g y
L in e a g e D e p t h
M ea n n o. o f h u ts in v illa g e
4
8-4
5
9*4 13*5 I 3 -I
6 7
8 & over
Mean 6*r
14*0
I z '9
to find a greater depth o frecall b y the headmen o f long-established villages than b y those o f recently established villages. This is because it is o f im m ediate interest to distantly connected kin in
84
Schism and Continuity in an A frican Society
long-established villages, to trace precise genealogical relation ship to one another as members o f a single local com m unity. O f course, one m ay find that headmen o f recently established villages m ay be able to trace back their matrilineal ancestry to quite a considerable depth. This is especially the case i f they belong to the maternal descent group o f a ch ief or senior head man, and therefore are possible successors to an im portant office. B ut it is usual to find that headmen tend to trace back their descent from that ancestress w hom the matrilineal core o f the village have in com m on, and to forget her predecessors. Struc tural amnesia typically begins above the ancestress whose genea logical position is structurally significant for the living members o f the village matrilineage. I f rem oter ancestresses are recalled, one finds oneself in the field o f supra-village relations. The lineage am ong the Ndem bu is not, as am ong the Tallensi,1 * the skeleton o f their social structure’, so m uch as the endproduct o f a number o f social tendencies w hich under specific circumstances place a curb on the m obility and instability o f residential groupings. I discuss these tendencies in the subsequent analysis. A m ong them m ay be mentioned a time-honoured village title, continued proxim ity to an area where gam e is abundant, and the good reputation o f its headman at the crucial transition period from sibling group to structured lineage. A m ong the Tallensi the lineage is ab initio a basic premiss o f social organization. A m ong the N dem bu the local matrilineage is a goal, and an ideal to w hich am bitious headmen aspire. O nce a village has becom e established and consolidated, great efforts are made not on ly b y its headman, but also b y those w ho hope to succeed him , to prevent the disruption o f its widest span lineage. T he form and direction o f such efforts w ill be considered in a series o f detailed studies in this and the follow in g three chapters. In m ost cases these efforts are ultim ately unavailing since the ecological and structural pressures m aking for fission are too strong, but the efforts are nearly always made. The basic unit in Ndem bu society is not the lineage, but the matricentric fam ily w hich in its life-cycle becomes, after the death o f the mother, the uterine sibling group. This and not the elem entary fam ily is the basic unit, since frequent and easy divorce, often consequent 1 Fortes, op. cit., p. 31.
0
M atrilin eal D escent
85
on the conflict between a w om an’s husband and her brothers for custody o f her and her children, renders the fam ily unstable and impermanent. T he uterine sibling group is a constant source o f danger to the continuity o f the village lineage ; the narrower unit is the foe o f the w ider. It w ill be show n below that the most frequent unit o f secession from a village is the uterine sibling and n ot a lineage o f w ider span. Bach xnatricentric / is given an early autonom y from other m aternally-linked matricentric fam ilies b y the institution o f virilocal marriage. Hence w hen the m atricentric fam ily fin ally returns to the village o f its nearest m atrilineal kin, it is seldom assimilated fu lly into the latter, and tim e and again its narrow er loyalties prevail against the value set on village unity. W e have seen h o w vinlocal m arriage disperses the members o f the uterine sibling group for a time, and h o w they tend to divest them selves o f other attach ments in order to becom e reunited as a local autonom ous group. The success o f the narrower against the w id er unit as a residential nucleus is attested b y the small average m agnitude o f N dem bu settlements. I cannot say w ith any certainty w hether this structural w eight ing in favour o f the narrower maternal descent group is directly associated w ith an ecological emphasis on hunting, w ith its con com itant m obility and productive individualism , or whether it is linked to hunting on ly through virilocal m arriage w hich prim arily consolidates adelphit relations. It m ay w ell be that a whole com plex o f ecological and historical factors is involved. B ut it is clear that the tension betw een the w idest village lineage and its com ponent uterine sibling groups, overlaid though it m ay be by sundry other sources o f tension, nearly always asserts itself in situations o f crisis and determines the dom inant m ode o f fission.
a
N dem bu do n ot now possess structurally significant clans (nyinyachi o r nyichidi) although they had them in the past, and a few older people rem em ber their names. M ost o f the peoples w ho came from the Luanda hom eland o f JMwantiyanvwa possess a set o f clans in com m on, several o f them named after some legendary hero (such as Saluseki or Sachingongu). A m ong Lw ena today individual members o f these clans still have an obligation to offer hospitality to fello w clansmen, even those K
0
0 0
e
0 0
0 o
0
0 Q
© © © © © C ©
Cl ©
©
© © ©
©
© © Q €>
©
©
fT) t
j
-■ o o (
f ■"*; O o 0 o o o o o ( ”) C)
(3 o
c
c.; C
0
O •O * t ;..J
V:.-'
■ %l.y &•
V
86
Schism and Continuity in an A frican Society
o f other tribes, w ho travel across Lwena territory. B u t for m any years N dem bu have had no clan in the sense o f a group consisting o f a num ber o f m en and wom en, bearing a common name w hich passes m atrilineally, dispersed w id ely over the country, forbidden to interm arry, and claim ing vague kinship w ith one another, such as exist am ong other Central African peoples like the Bem ba, Kaonde, M azabuka T onga, Lamba, Lunda o f Luapula, and m any other tribes. N dem bu m ay marry a prim ary (though rarely) or classificatory sister’s daughter’s daughter, a m arriage w hich w ould have been forbidden b y clan exogam y. N dem bu also have a fairly narrow range o f incest prohibitions. Exam ples o f marriages w hich occurred between reputedly forbidden categories o f kin and w hich have been cleansed and made publicly acceptable by ritual, are given below in case histories. In one case a wom an m arried her matrilineage brother descended from a com m on great-grandm other and the m arriage, w hich served as an index o f village fission, speedily achieved general recognition (see p. 208). T he disappearance o f clan organization is a further example o f the persistent ten dency towards the consolidation o f narrower groups at the ex pense o f w ider groups in all sectors o f the N dem bu social system, except in the field o f ritual where, as w e shall see in Chapter Ten, the reverse holds good. Indeed, even villages whose nuclei claim com m on maternal descent cannot trace precise genealogical connection w ith one another, unless one has split o ff very recently from the other. As w e have seen above, villages said to be inter-related b y mater nal descent m ay be scattered far and w ide across the N dem bu region in different Governm ent chiefdoms and in different vicinages. Nevertheless, N dem bu possess a lineage group w ider than the village lineage w ith in w hich genealogical linking is possible. This group extends beyond the village m atrilineage and includes all, o r nearly all, o f the livin g matrilineal kin o f the village m atrilineal core. T h e w om an w ho has married out o f a village, and her children, are not forgotten by her kin, even i f she lives m ore than fifty m iles aw ay in her husband’s village. T he man w ho has quarrelled w ith his matrilineal kin and has gone to live w ith a paternal o r other cognatic relative is sim ilarly remembered. Sisters’ children w ho as adults are living w ith their fathers are kept
M atrilin eal Descent
87
in mind. T h e continual flo w o f visits between m atrilineal kin, how ever far apart in space, serves to maintain their connection. These dispersed m atrilineal kin constitute a pool w h ich m ay at some tim e be draw n upon to supplement and replenish the village o f the nuclear maternal descent group. A gain , i f a man wishes to start a village or firm o f his ow n he m ay attem pt to attract to his new settlem ent such scattered kin. A man I kn ew w ho was thinking o f starting a village told m e o f 17 m en and boys and 10 w om en and girls o f his lineage livin g in his village, and o f 19 males and 19 females o f his lineage livin g outside it. H e had in the period during w hich I knew him paid a series o f visits to m any o f these m atrilineal kin in other villages to sound them about com ing to jo in him in his contem plated n ew village. This exam ple illustrates the point that N dem bu regard matrilineal kin as the prim ary and essential elem ent in the social com position o f a village. It is true to say that both a man and a divorced wom an have a choice in the m atter o f residence in that they m ay decide to live either w ith their m atrilineal kin or w ith their father, or, indeed, in other villages under special conditions ; but they have a life-lon g and autom atic right to reside in the village o f their close m atrilineal kin, unless they have lost this right through some sin or crim e they have com m itted such as sorcery, w itchcraft or m urder. A man has the right to remain in his fathers village w hile the latter is alive, but after his death he must go to his m atrilineal village, unless his m other was a slave, in w hich case he w ould belong to his father’s uterine heir. Succession to Office M aternal descent determines succession and inheritance. In succession to headmanship uterine brother succeeds uterine brother, and i f the village is a large one w ith a deep matrilineage it sometimes happens that m atrilineal parallel cousins succeed one another. W h en the ro w o f brothers and classificatory brothers has died out, o r i f there are no suitable candidates am ong them , the right to succeed passes dow n to the next senior genealo gical generation o f sisters’ sons. D ue to this tendency to confine succession w ithin the membership o f the senior genealogical generation, it n ot infrequently happens that sisters’ sons becom e im patient fo r office from w h ich they are barred b y adelphic succession until * hope deferred m aketh the heart sick \ They
88
Schism and C ontinuity in an A frican Society
hive o ff from the village w ith their ow n uterine sibling gtoup and the children o f that group and found new settlements. Table X V I, based on data I have collected about succession to village headmanship, supports this statement. TA BLE X V I R
e l a t io n sh ip s
B e t w e e n H e a d m en
a n d t h e ir
S u c cesso rs 1
Relationship o f successor to previous incumbent
No. o f instances
B
zS
zdS
S
20
15
4
3
mBS mB 1 z 0
O
X
zd
Total
0
43
Legend B Brother zS Sister’s son zdS Sister’s daughter’s son S Son mBS Mother’s brother’s son mB Mother’s brother z Sister £d Sister’s daughter 1 Excluding the instance o f succession in Chikang’a Village discussed on p. 211.
From T able X V II, p. 205, it w ould appear that sisters* sons and not brothers o f headmen are the most frequent founders o f new villages. This tendency w ould be less marked i f there existed, as in m any other matrilinea! societies, a clear-cut norm eryoining nepotic succession to office. I f w e exam ine the pattern o f suc cession to headmanship, Table X V I shows that brothers succeed m ore often than sisters’ sons. W hen it is realized that the number o f m en in the senior genealogical generation is far less than in the jun ior proxim al generation, w hich contains men o f a low er average age, it is clear that the chance o f succeeding to office is m uch greater in the form er. Adelphic succession ensures that a number o f aged office-holders w ill succeed one another, few o f w hom w ill live long. Even i f a sister’s son does succeed, he probably succeeds much later in life than he w ould have done i f he had been his m other’s brother’s first heir. A value is attached in N dem bu society to the wisdom w hich comes from age, and
M atrilin eaí D esceñí
7í,
llS lf
Si ■'t u
>::
89
age-seniority is an im portant qualification for headmanship. The older an heir is when he succeeds, the shorter the tim e he is likely to hold office. Consequently, there tends to be a high turnover o f adelpfaic successors, w h o manage to obtain office before they die. I f heirs manage to succeed young, they remain in office longer, w hich means that potential successors have to wait for m any years, and m ay eventually be passed over. For these reasons, it is often a headman’s sister’s son. w h o leads a dissident group from a village. I f he wishes to enjoy a long period o f leadership, he m ay w ell prefer to give up his chance o f succeeding to office in a long-established village, despite the greater prestige o f such an office, than to w ait until he is old. In this and subsequent chapters I present a num ber o f casehistories (social dramas) w hich throw light on the problem s o f succession to headmanship. It is indeed distinctly unusual for sisters’ sons to succeed w hile a m em ber o f the mothers* brothers* generation remains alive. T able X I shows that out o f 385 matrilineal hut-ow ners on ly 7 w ere prim ary or classificatory mothers’ brodiers o f the headmen, and o f these o n ly one was a headman’s ow n m other’s brother. This man, Biscuit (III, E 7), lived in K am aw u V illage near M w inilunga Bom a ; he had spent most o f his life in Southern Rhodesia, and. had returned as an old man to live w ith his sister. She was a w id o w livin g in a village which her sister’s son had founded. W hen her brother returned he had no w ife to cook for him and no other kin except her to look: after him . H e did not contest the headmanship o f the village w ith his sister’s son, as he was feeble, forgotten b y m any people, and an inveterate drinker. B y and large, how ever, it m ay be stated that in practice as w ell as in principle brothers succeed brothers and cousins succeed cousins before nephews obtain the right to succeed. Conflict over Succession within the Matrilineage Thi&-bDokis_dom inantl^a..sjtu.dy_Qf^ the so^d^al.iiiechaiihm s.hroughtm tcr play..toreduc_e_}_e2ccludeQr resolve that conflict.__Beneath all other conflicts in Ndem bu society is
th e concealed opposition between men and w om en over .descent and in the econom ic system.-_Influenced.by .tMsBasic apposition, but-possessing their o w n autonom y, sets o f struggles an sew ith in the sQciaL.strueture..:_ ..conflicts... between, .persons and. .between
n
h o
90 ■
*—• o
c o 0 n T"1 / o D
O C
j ¡j j:
0
j | ij j| 11 !j ;i 1 1
Schism and Continuity in an A frican Society
group&jgho invoke different principles. o£. residentiaL afifiliation to support _and justify their ow n specific interests,., political, jural andeconQmio_;_.struggles_.be.tweerL,perspiis_aiKl,gmups_CQUched in_te£ms_Qfia_cQmmon-J3orm..whicli each party claims the other has_broken..;„ and_.CQnfUcts.„betwe.en.-pjsrsQns,.„wniteti„Ky:..a„single principle o f descent.. and_xesidence,„fbr_pQsitions_o£-auihority determined by that principle. Struggles-aro.und_s.uccession to village headmanship are instances. o£.the.last.typeL-o£x:onflict-and it Js w ith these that I wish to. commence die analysis-of w hat I propose to call * social dramas*. Eormal._analysisL-.of-a social system enables us-to locate.and.isolate_criticaLpoints_and_areas in itsL_structure where one. m ight expeot, on a prior¿grounds, to fmcLconilicts between the...occupants of-social-positions-carried in jthe structure. Im the examination . o£.the_,Mdemhmsystem I have_in„this..xhapter_isQlated_the.matrilineal_descent_group and shown .how the office o£village-headman.-is. .vested-in_this_group. I have looked at different categories-of.matrilineal.kimand_sh.own howi^struggles fo r succession m ay be expected ,to take__place between adjacent generations- and. b.etween—specific-. kinship positions, notably .between m other's-.brother and- .sistecs son. Xl<:remains to., test out these., hypotheses._in..a-number_o£ cases, regarded as typical, and to see whether struggles, do.in_fact_take place. But the task does not end at this point. I f conflicts occur w e want to see in what, w ay they are. handled .by the. memhers o f the society. ImNdemhu..society..cQndu.ct.has..b.een regulated over what_w.et_can-assume_tO-.have been a,_very .long period o f tim e by norms, values,-beliefs., and sentiments associated, ’with.., kinship. Conflicts o f interests arising_o.ut.of-.the .sociaLstructure are per petuated by the observance of. thes.e.norms.... Hence, the.conflicts must_also .followr a. regular course ..dic.ta£e.d.par.dy_.byjhese..norms, and take a shape grow n fem iliar to..the.peapleJhraugh_repetition. W e can expect to. find, in fact,...a mimber,..o£ social..mechanisms, o f institutionalized ways o f behaviour, whidbu-havearisen in response to .an almost endless reduplication o f such-conflicts, and which have been designed b y group._.experience_tQ-mitigate, diminish or repair them. C o rn e t .and..the.xesoiution-0.£xoiiflict hafce_ejfects.wm ch„areohservable in statisticaLand -genealogical data— But~the hints, and indications, afforded by,such.. data must be .follow ed up by a close study o f social dramas.— There we observe the interlinked and successive events which fo llo w breach,
M atrilin eal D escen t
9i
andLxnake visible, the. sources o£ conflicts^„ _Xhis im tum -leads to action'which may. restore theearlierset.o£relatians,.or.recanstitute them in a different pattern, o r even ,recognize- an-irreparable breaking o f relationships betw een particular persons-or~ groups. These last, nevertheless,!« into the w ider pattern ofthfi-N dem bu system.
The Concept o f the Social Drama O n a num ber o f occasions during m y field-w ork aware o f m arked disturbance in the social life o f the group I happened to be studying at the tim e. X h e jw h o j^ ^ ^ ^ l m ight be radically cloven .in to. tw o conflicting - fa c t io n s t h e quarrdling parties might-.compriseL„.some—b u t-n o t...a ll-o f its members ; or disputes m ight be-m erely interpersonal iru.character,.... Disturbance in . short....had. a~ variable -_range_o£_sodal inclusiveness. A fter a w h ile ! began tO -detec^a-pattem .in these eruptions o f c o n flict: I noticed phases_m.„their„develo.pment which seemed to follow-one„another-in_a-m ore-or,-less-regular sequence__Xhese~em ptions,._whkh„IxalLlsacial-dram as./, have * proeessionaL.£ormj>._XJiav.e_-provisionally^.dmded .-the- social process^which.constitutes_the-..sociaLdram a „in.to__four major phases : (1) B reach o f regular- norm -govem ed social relations occurs between persons or groups w ith in the—sam e system—o£. social relations. Such a -breach is. signalized-hyi. the_.pnbKc„breach_ or nonfulfillm ent o f som e crucial, norm xeguIating..the- intercourse o£the-parties. (2) F o llo w in g-b reach -o f . regular .social relations, a phase o f m otm ting crisis supervenes, durm g w hich, ^ can-be sealed o ff quickly w ithin a lim ited area o f social.interacrion, there is a tendency.for...the.breach to a^ -extend.-r'Mntil it becomes co~extensive with-som e dorninant.cleavageinth.e widest set -oX felevant social relations to w hich the conflicting parties belong. T h e phase o f crisis exposes _the_ pattern_ o£..cvirrent factional- stxu gglew ith in .th e relevmt_socdal grou p, be it village, neighbourhood o r chiefdom ; and beneath, .it—there—becomes visible the less plastic, m ore durable, but nevertheless ,gradually fhangin g basic social structure, m ade up o f relations w h ich are constant-and-consistent. (3) In-order to lim it the spread o f breach certain adjustive and
92
Schism and Continuity in an A frican Society
regressive mechamsms,.in£GrmaLor£oxmd,are-speedily.jbrought into-operation b y leading members of. the rele_vaxit_s.o.ciaLgroup. }\ These mechanisms vary in characterjw th su c h ..fa ^ depth i; and social significance o f the. breach,.the. sociaLinclusiveness .o f :j the_crisis, the nature o f the .sooal^gro.up-vdthimwiuch-the.. breach j| took place and the degree o f its autonom y .with referencero. wider !( systems o f social relations. T h ey m ay range. from .personalad vice andinform alarbitration, to form al juridical and legal machinery, and^to resolve certain kinds o f crisis, to theperforman.ee o f public ritual. (4) The finaLphase. I have distingmshed. consists either in t jj reintegration o f die disturbed social group, o r in- the social recogJ| nirion.ofirreparableL_breach_betw:een_the.cQn.testing. parties. |j in ' short,. the processional form o£ the-social-dram a-m ay be ¡j form ulated as (1) breach ; (2) crisis ; (3 ). redressive.. action ; •j (4) re-integration or. recogniti.on.-of schism. lL_must. be. recognized,... o f course,-..that- in- different—kinds o f group, in different.societies,, and under..varying circumstances in theisame kinds o f group in die.sam e _society, the prQcess m ay not run- sm oothly or inevitably from , phase to.-phase.... Failure, for exam ple, in the operation o f xedressive machinery, m ay result in regression to crisis. In recently-form ed groups, institutional ized ieg a l or ritual means o f handling, social, disturbance -may be lacking, and breach m ay be succeeded -im m ediately- b y the irreversible fission or ficagmentetion-.of-.the group. | In hidem busociety,although.villages- arise and..penshrthe ideal j form-o£the village persists. M eanw hile, in order that any village life ^should be possible, it is necessary, that members~o£-a-village should observe certain com m on. ..values,—and-tha.tL_the._norms governing behaviour between village..memhers^jcnast„o£.whom arejmterlinked by ties o f kinship-and affinity, .shouldLbeupheld. W here custom ary, norms .and..values .are,deeply.entrenched it is usual to find institutionalized. m achinery. o f redress. Each instance_o£-breach Jn-sociaL-relations-is-made-the-OGcasion o f a restatem ent-oftheh-xegulative. norms.— The.iiature_o£xedressive machmeiy-.an.d..the..way.m.which_itLfunctions-itx_&p.ecific situations is discussed later. I have found the social drama a useful descriptive and .analytical tool W-hen taken in conjunction w ith rn o re orthodox techniques o f analysis. such as the genealogy,-the-census.and.theh.ut diagram.
M atrilineal D escent
93
AnaLysis-of_numeri.caLmateriaLobtaiiied_hyJ;he^use_ofls.uctLjt:eck“ niquesLj:eveals^ regularities in social relations., that _.we...may call structural. A m ong N d e m b u ,fo r example^..we^fmdLby-. these means^tbatLthexore^o£.yiHage&.tends^to_be..tlie„inatemaLdescent group^that_mariiage..is...predominantlyi-virilocal^-tha£..tbere is a ffighfrequency o£chvQ:rce5.thataltemate_genealagicaLgeneEations tend to build adiacently.and. adjacent ^enealogicaL-eenerations in opposite sections o f a.:village, .and,, so xm l W e....find..that.jh^re is a tendency..towards.addtahic..succession.and that sisters* sons tend to th e .m ost frequent founders o^ new ..villages.— This leads us to..suspect tension in the relationship .of_mo.therV.brother- and sister s-son. T he social drama shows vividIy..how...these- social tendencies operate in practice ; how , in a givem situation^som e m ay .support and others oppose one..ano.then^.andJhow_coiiflict between persons or groups, in. terms~of_a_common_norm,or in term s-of contradictory norms...may..be xesoLved-in.a-particular set o f circumstances. In the social drama latent conflicts-of.interest become^manifest, and kinship..ties,...whose..significance...-is-not obvious in genealogies, em erge irito_key^ im portancc. I f w e exam ine ft sequence o f social dramas arising w ithin the same .social unit, each one affords..us..a glim pse, .as.it. w ere^ of the contem porary stage o f m aturation o r decay o f the .socialstructure o f that unit. Xhope. ta..demonstrate ..thisin_pxesenting. a set o f five consecutive social, dramas. ..in., a single long-Æstablished village. T he social drama is a lim ited areaoftranspareiicy on the otherwise opaque surface o f regular, uneventful so cial life. .. .Through it w e are enabled to observe the crucial principles.of.the social structure in their- operation., and_jtheir„relative dom inance at successive points i n tim e. O f'th e five social dramas based on material collected at M ukanza Village,-thelastLthreeLcame.undeJLmyLdirect.ohsetvation. T he first-two.rest.on.data-CollectecLfrom ..a.large.num her.of interviews/and conversations w ith livin g . persons w h o .actively par ticipated in them . Social Dram a I illustrates the conflict that m ay arise between m other’s brother and sister’s son, and between m ale parallel cousins, w hen on ly a few men rem ain in the senior, office holding generation in a village, and there are several m iddleaged men ripe for office in the ju n io r generation. O ther kinds o f conflicts becom e overt w ithin the fram ew ork o f these crucial
Jb.e
94 | !
; \ j j
| | i I j jj j|
¡1 !
j
| | :
1
Schism and Continuity in an African Society
conflicts ; but the form er w ill not be analysed here, since they involve other principles o f village organization than m atriliny. This social, drama is one o f a series, .each o f j^ H ch .contains the same principal characters, and. each o f w hich, reflects,,.different aspects o f the same structural conflicts. It m ay be-objected that such-factors as innate psycho-biological constitution and. personflity_ vananons..determ ined.bydifferenti3L.traiiiing.in,the-early years of,childhood take precedence,over sociological-factors in shaping_the.. events, .to. be. described. .- B u t it is d e a r - t h a t , the d i f ^ e n « L _ p e r s Q n a h d e s . i n y o . l y e d . . occupy. s.adalp-Qsition&..tha£..must inevitably com e into conflict, and. each, occupant .of.,a.-position must presentJhis_case in terms ..of.generally-acceptc
M atrilineal D escent
95
SO CIAL D R AM A I The Bewitching o f Kahali Chandenda by his Nephew Sandombu (compiled from informants) One day, in 1947, Sandombu (I,Gio) trapped a duiker and divided its meat between his village kin. His own mother’s brother Kahali Chandenda (LF5) was headman o f the village and should have received by custom a back leg or the breast o f the animal as his share. Sandombu, however, gave him part o f a front leg only. Kahali refused to receive it, saying that Sandombu had shown that he despised his uncle. A few days afterwards Kahali went to a village in C h ief Sailunga’s area about eight miles from Mukanza Village and snared a bush-buck. He sent his daughter back to Mukanza with the meat. Sandombu took the meat and proceeded to divide it, retaining the breast, liver, a front leg and the head, for himself and his wives. N ext day Kahali returned, and finding Sandombu away, asked Sandombu’s wife Malona, a Lunda from Angola, for some food, for he had no wife o f his own at that time. She was insolently slow in preparing the food, and in the end he went to his classificatory sister Nyamwaha (I,F7), who gave him beer as w ell as food. A t night Sandombu’s sister and Kahali’s own niece, Mangalita (I,G n ), came to him in private and told him, in anger and shame, how the meat had been divided. N ext morning Sandombu set o ff early to go to Hiliwood Farm, twenty-five miles away, where he had seasonal employment as a capitao in charge o f road maintenance. Meanwhile Kahali spoke bitterly in the village forum (chota) about Sandombu’s action, and the latter’s wife Malona wept tears at this public shaming o f her husband. After a week Sandombu returned to Mukanza Village and Malona reported Kahali’s remarks to him. A fierce dispute arose between unde and nephew, in the course o f which each threatened the other with *medicine ’ (yitumbu, a euphemism for sorcery, w uloji). Sandombu ended b y saying, ‘ I am going to Sailunga Area. The people o f this village are worthless. Some people must look out.’ B y this people took him to mean that he was going to seek out the services o f a notorious sorcerer, thought to be Sakasumpa o f Shika Village, to kill Kahali. It was believed by Mukanza people that Kahali kept his familiar, an ilomba, or water-snake which possesses the face o f its owner, in a stream in Sailunga A rea; and that Sandombu had gone to pay Sakasumpa a fee to shoot the ilomba with his wuta wawufuku or *night-gun \ a piece o f human tibia carved in the form o f a muzzleloader and primed with graveyard earth and decomposing pieces o f
H m Hi £
o
2 ~ w 2
52 I
ì o
§M « £
3 3
I« IS
'i.
T"
ô
-•< S2 < 9 <3 gs si »< S*
? 9
Q O
¥
iu
o_ *§ 2:* §
? ?
L fJj j K > i.Hw. N Y A M U K O L A — *•
SAKAZAO
S\
U I.Dî. MÀLABU------ ® r,Ej.»---------------- « ----------------------® [,Gi6_________ f * ^
Lü LU U U...I,J.,.ü
O 0
0 O O
© G O
© © © O Ô
©
© © Cl
© ©
© © ©
© ©
© ©
m a la b w
M AP 5. H ut P l a n o f M u k a n z a V h x a g e m 1947-4 Illustrating Social Dramas I and II (Line down centre shows the separation o f adjacent genealogical generations)
© © ©
©
98 ::o so
Sr. ■c ■ o
Schism and Continuity in an African Society
human bodies. When an ilomba has been killed its owner also dies. After a few days Kahali fell ill, and. died shortly afterwards. A rumour came to Mukanza Village that Sandombu had boasted in Shika that he would kill his uncle by sorcery. N o divination was made into the death since the people feared prosecution by the Govern ment for making accusations o f witchcraft. Besides some, including Mukanza Kabinda (1,FS) and Kanyombu (I.Fp), classificatory matriIineal brothers o f Kahali, said that Sandombu had condemned himself out o f his own mouth both in Mukanza and in Shika Village. Sandombu returned to his place o f employment and Mukanza people said that Sandombu must not succeed Kahali, for he had shown himself to be a man with ‘ a black liver * (muchima bwi), a selfish person and a sorcerer. The question o f succession was left over for a time. Sandombu was not expelled from the village because there was no positive proof o f his guilt, such as might be obtained from the diviner’s basket. He had only spoken in heat as had Kahali himself, although good men did not speak in this way. Mukanza Kabinda was made headman with the approval o f all, and confirmed in his office by the Government Sub-Chief Mukang’ala. Analysis T he account presented is a digest o f inform ation given me, four years later, by most o f the senior residents o f Mukanza V illage including Sandombu himself. Sandombu, how ever, denied that he threatened to bewitch Kahali, although he admitted that he w ent to Shika V illage after the quarrel. H e said that he did this in order that peace m ight be restored in the village. Kahali, he said, was an old man, and old men always die some tim e or other. This rationalistic attitude was not, how ever, typical o f Sandombu in other situations, as w e shall see presently. A ll the other members o f the village concurred in the view that Sandombu had bewitched Kahali and asserted that he had publicly avowed that he w ould do so. It is impossible to get any closer to the actual facts o f the case fo r the events are no longer susceptible to enquiry and the account has acquired a m ythical character. W h y has it b y the consensus o f all save one acquired this character ? T he answer lies, I think, in the genealogy o f M ukanza V illage itself, and in the operation o f the principles o f residential affiliation discussed in Chapter Three. The m yth o f the bew itching—o f. Kahali Chanjfencbt in a. short „period„ o £ tim e becom e the
M atrilin eal D escent
99
, ‘ m ythological, .charter .to.use M alinow ski’s expressive phrase, the-sodal. justification, for „the. exclusion...of. Sandom btuftom the succession. W h y had. he been so excluded ? W as it sim ply because he was regarded as a selfish and a quarrelsome fellow and a sorcerer to b o o t? I do not think that this explanation altogether fits the facts. Sandom bu was a m ost generous host to m e as w ell as to others. H e was a diligent agriculturalist, w ho grew m uch finger-m illet and gave aw ay beef brew ed from it free o f charge. A lthough he m ight easily have follow ed the grow ing practice o f selling beer, he preferred to give it to all and sundry. B ut there is no doubt that w ithin the lim its o f his little w orld he was a h igh ly am bitious man, eager fo r headmanship. „and the prestige that even today attaches to., that, .office- in.-the m ore conservative areas. This is w h y he gave aw ay beer and also food, for he wanted to put m any people under an obligation to him, and in tim e to build up a follow ing w ho w ould com e and live under his leadership in his village. This, am bition was a little^too obtrusive fo r his relatives to. stomach,., since, he w ould boast,, even w hen I knew him , that one day he w ould becom e headman o f M ukanza V illage. This meant .only one thing to Ndetitbu, that he w as im p a tie n tfo r office... and would._s.top at nothing, n ot even sorcery, too-btain bis ends.. B m Ms hospitality, his obtrusive, am bition,, .and even, his w ild boasting„in „hi? cups, suicidal though it was at the tim e. to. hisL.hopes,- ..all stemmed from his position in the. sod ai structure. It m ay be noted from A ppendix I that M ukanza is a lon gestablished village in w hich Kahali Chandenda (I,F j) was the sixth successive headman, and in w hich there w ere tw o w elldefined lineage segments w hose com m on ancestress was in the third ascendant generation from the present headman. Succestitinhad^beemadelphic-between-the. secondhand, third headman and im:the_next-.genealogical.generation-thre.e,matrilineal parallel cousins.had follow ed, one .ano.ther.in the.headmanship*-,™In other w ords,thesocial.com p_ositionof.theviEage.m usfJ^vesuccessfidly undergone the_qualitative change from „bilateraLextended-fem iiy (i.e. .uterine siblings and their children) to matrilineage. and had developed a .precise, and. extensive, system^of-complementary and opposed genealogical generations.. B u t w ith the onw ard flo w o f strutturai tim e and th e ..passage. ..of-successive-genealogical generations the danger o f fission, becam e..progressively greater.
io o
Schism and Continuity in an African Society
T he lineages o f w hich nyachjntang ’ a 1 (I,D i) and malabu (I,D a) w ere the respective ancestresses w ere structurally ready to divide from one another. A few years previously (1928) another m atrilineage had split o ff after actual fighting in the village between the senior m en o f M ukanza ; and it was touch and go, so to speak, w hether the lineage (ivumu) o f malabu , the senior man o f w hich was Sakazao (I,H9), w ould not follow suit. Sakazao, although in the second descendant generation from Kahali Chandenda (I,F5) and M ukanza Kabinda (I,F8), and in the first descendant generation from Sandombu (I,G io), was older than the latter, and not m any years younger than M ukanza Kabinda. M ukanza Kabinda was in fact the person w ho held the tw o lineages together, since his w ife, b y w hom he had m any children, was o w n sister o f Sakazao, head o f the malabu lineage. H e was thus a matrilineal relative o f his w ife by grandparent-grandchild m arriage. His w ife N yam ukola (I,H io) was the link that united the tw o halves o f the village and stood in an intercalary relationship between them. H er children are children o f the tw o lineages : outside M ukanza V illage they had now here to live. A t the same tim e, Mukanza Kabinda (I,F8), his sister N yam w aha (I,F7), and his brother Kanyom bu (I.Fp) w ere the last surviving members o f the senior genealogical generation after the death o f Kahali. N o male o f the m alabu lineage, to w hom the headmanship m ight pass, remained alive in that generation. O ld Kajata (I,Fio) was senile, and at that tim e seemed lik ely to die in K afw eku, his w ife’s village. M ost o f the senior members o f malabu lineage w ere in the second descendant genealogical generation to M ukanza Kabinda and his siblings, and w ere m oreover allied to him by affinity and sibling-in-law ship. O ne man and three wom en o f malabu lineage and o f G generation (to w hich Sandombu and Kasonda also belonged), w ere alive at that time. T h e man, Chinem a (I,G ai), was w orking on the Copperbelt as a road labourer, and it was then considered that he w ould never return to M ukanza V illage. H e had been headman o f Kahali Chandenda’s village in Ikelenge Area fo r a number o f years after Kahali had returned to M ukanza. B u t since his village there had been com posite in character, containing m any members o f 1 The
names o f lineages are in capitals.
p
iIOI oi
another lineage and few o f his ow n m atrilineal kin, he had been ¿^involved in m any disputes w ith the other lineage. H e returned v to M ukanza Village about 1945, perhaps w ith the hope o f suc ceeding Kahali Chandenda. B u t he had quarrelled w ith N yam uw ang’a, the notorious w itch o f Social D ram a FV, and it was said that his fear o f her led to his flight to the Copperbelt. N yam uw ang’a herself was barely tolerated in M ukanza Village, and, as w e shall see, was subsequently expelled from it. Her sister Shim ili (I,G i8) was at the tim e o f Social D ram a I living '‘«hi, virilocally w ith her husband in Chim bila V illage in Ikdenge Area. O f her other sisters, Kalusa (I,G i9 ) was livin g v in locally at Shika V illage, w hile N yam alita (I,G zo) alternated between her virilocal village Chim bila and K alene Hospital where she w orked fo r lo n g periods as a ward-m aid. N one o f the livin g members o f G generation in maiabxj lineage entered the reckoning at the tim e o f Social D ram a I as candidates for office or supporters o f a candidate. B u t Sakazao and Ins uterine sisters and their children could be regarded as m ore or less permanent residents in the village, and Sakazao’s sisters w ere married to M ukanza and his brother K anyom bu. M ukanza K abm da.byanum bero£reckonihgsQ ccupied-a_pivQ taI-position with-Jtegarri-_to. JtbcLjtilUgitJK^..a..whole.^.,most_Q£.its^stmctural relations converged on his, position ,in the.kinsbip„,system. B efore considering die position o f Sandom bu in this situation, another potential source o f danger to the continued existence o f the m ajor m atrilineage should be m entioned. It has been noted that fission m ay, and frequently does, occur as the result o f the secession o f a fam ily o f mature uterine siblings and their children. The possibility o f such fission existed in M ukanza V illage at die tim e o f KahaH Chandenda’s death. T w o w om en, C haw utong’i (I,G i2) and M anyosa (I,G i3 ), both sister’s daughters o f M ukanza, had large families, one w ith six and die other w ith five children. B oth these w om en w ere livin g avunculocally w ith w eak com pliant husbands. T he husband o f one, nam ed N deleki, in fact provides an illustration o f the fate o f a man w ho has n ot pressed Ids claims to the headmanship o f a village. H e was a quiet and selfecontained person, a devoted husband and father, but w ithout that stubborn individualism , barely concealed b y the veneer o f politeness and sociability, w h ich is typical o f N dem bu o f both sexes. W hen the headman o f his m atrilineal village had died,
C Cc o e c C GCOQCCCCOOGOOOOOOOOOOOOOO®
Matrilineal Descent
102
O
IV-, l ■ U
c
(■ )
fe" /
O .'O Q
Q 4 )
Schism and C ontinuity in an A frican Society
he had n ot put him self forw ard as the successor. As a result people tended to despise him or at best to ch ivy him in a goodhum oured w ay in his ow n village. H e phrased his social re jection there in terms o f w itchcraft accusations against his kin and w ent to live at his w ife’s village. His w ife Chaw utong’i (I,G i2 ), on the other hand, was an aggressive and ambitious wom an, high ly conscious o f her im portance as a m other o f many children. H er m other’s sister’s son, her ow n parallel cousin, Kasonda (I,G i5) b y name, was an im portant man in the litde village com m unity, shrewd to observe the social norm s o f N dem bu life but also am bitious and seeking b y devious means either to succeed to the headmanship, or better, to start a village o f his ow n. The other m other o f m any children was his ow n sister M anyosa (I,G i3), also a dom inating personality. A ll three, Kasonda, M anyosa and Chaw utong’i, w ere fu ll sisters* children o f M ukanza Kabinda (I,F8), w ho in form er times w ould have had the right, as their m other’s brother, to sell them into slavery. He still exerted considerable authority over them . B u t one factor in particular at that tim e m ilitated against Kasonda’s claim to succeed— his relative youth. A lthough he had lived fo r more than ten years in urban areas, five o f them in B ulaw ayo, and although he had acquired a little education in M ission schools, traditional N dem bu notions rated him too young, at the age o f thirty-five, to becom e a headman. In frank conversations I have had w ith Kasonda about the question o f succession, he has told me that he saw his best hope in exerting his influence over his pow erful sisters to support the claim to succession o f his ow n uterine uncle, M ukanza Kabinda. T h e latter was b y that time an old man o f about 66 and could n ot be expected to live verylong. I f he died in a short tim e Kasonda w ould support Sakazao also in his sixties, fo r the headmanship. I f M ukanza Kabinda’s life lasted a further ten years Kasonda w ould press his ow n claims after his uncle’s death. Kasonda, although he was careful to maintain the outw ard show o f friendliness towards Sandombu, was privately jealous o f him , hated him as a bar to his ow n advancement, and feared him on account o f his widespread notoriety as a sorcerer and friend o f sorcerers. I f Sandombu w ere to succeed at the age o f about fifty, it m ight be m any years before Kasonda w ould get the chance to follow him into office, and therefore Kasonda was determined to keep
M atrilineal D escent
103
Sandombu out o f the running at all costs, b y the secret and devious ways o f whispered slander. H e, m ore than anyone else, was responsible fo r the story that Sandom bu had bew itched Kahali Chandenda, and he never allow ed the episode o f the quarrel to becom e forgotten. For various reasons it suited him w ell that M ukanza Kabinda, and not Sandombu, should succeed. M ukanza Kabinda had a retiring and unaggressive character and at that period at any rate did not possess Kasonda’s skill in the advocacy and judgm ent o f cases. Judicial skill (wukaku) is rated high among the accomplishments o f N dem bu men, and Kasonda hoped that through his ow n gifts o f eloquence and deliberation he would becom e the de facto headman o f M ukanza V illage. Kasonda could also speak w ith facility and assurance to such Europeans as D istrict Officers and missionaries w h o came sporadically ‘ to trouble the people * (nakukabisha antu), as N dem bu put it. His know ledge o f English, and his fam iliarity w ith documents, w ould prove useful in this respect, and provide a sharp contrast w ith M ukanza s illiteracy and ineptitude w ith mapaperi ( papers , a general term fo r tax receipts, gun and d og licences, bicycle tax, permits to buy gunpow der, etc.). Kasonda saw himself^ in short, as the G rey Eminence o f M ukanza V illage, recognized b y N dem bu as the effective village head, but delightfully free from official responsibility for the delicts o f its members. T h e links w hich bound M ukanza Kabinda, his brother-in-law Sakazao (I,Hp), and his ow n sister’s son Kasonda closely to gether, w ere extended to include the close lineal kin and de pendants o f these im portant men. Three linked pressure-groups w ere thus form ed, all o f w hich supported M ukanza Kabinda’s claim to succeed K ahali Chandenda (I,F5). Sandom bu (i,G io ), although anxious to succeed, was the odd man out. W h y then was he unable to obtain support from the other members o f the village ? T he principal reason fo r his exclusion from considera tion la y in his structural position. K ahali Chandenda was Ms ow n m other’s brother and i f Sandom bu had succeeded, members o f one lineage segm ent o f different generations w ould have succeeded to headmanship, w hile a senior o f the other lineage segm ent was available. A principle o f great im portance in N dem bu political structure w ould thus have been broken, nam ely that no single lineage segm ent should obtain a m onopoly over a long-established office, for such positions are few , highly
104
Schism and Continuity in an A frican Society
valued, and jealously contested. For instance, the chieftainship o f Kanongesha is vested in a dispersed maternal descent group w hich is scattered over the N dem bu region to form the nuclei o f diffèrent villages. There is a rule o f succession w hich states that no tw o successive incumbents should com e from the same village. This m ode o f succession ensures that the honours should g o round and not becom e the m onopoly o f a single village m atrilineage belonging to the chiefly maternal descent group. In a sim ilar w ay in long-established com m oner villages, classificatory adelphic succession, often cutting across affiliation by lineage segment, is a means o f interlinking discrete uterine sibling groups. I f the leading men in each consider that they stand a reasonable chance o f succeeding to office they m ay w ell hesitate before seceding from the village. T he basic member ship o f long-established Ndem bu villages consists o f a rather loose association o f free and independent elders w h o are not really constrained b y econom ic exigency or political directive from above to remain together. Nevertheless, i f the village is relatively ancient, pow erful ties o f historical pride m ay buttress those o f kinship to retain the elders’ jo in t allegiance. Perhaps their unity tends to be further stiffened by the possibility o f succeeding to the headmanship. T h e prestige attached to livin g in a longestablished village puts them under a certain am ount o f moral pressure not to secede from it and * k ill the village as Ndem bu put it. Since the headman, am ong the w atchfully egalitarian Ndem bu, is hardly m ore than a primus inter pares, he cannot coerce village members to remain in the village. T he unity and cohesion o f villages depends on the fraternal association o f generation mates, not on the dominance o f a single lineage segment. A nother reason fo r the prem editated exclusion o f Sandombu lay in the history o f M ukanza V illage. In m ost Central African tribes the nam e o f a village is a hereditary title through w hich flow s a succession o f individual incumbents, w ho take the name o f the founder o f the village. In the case o f senior headmanships this is also the case am ong Ndem bu. Here the names o f the founders becom e inherited tides. B u t com m oner villages are constantly com ing into being as the result o f fission, and also villages frequently becom e extinct ow in g to the secession o f the m ajority o f their inhabitants. There is often, am ong
M atrilin eal D escent
105
N dem bu, a period in the developm ent o f a village during w hich die village is generally know n b y the personal name o f each successive headman* B ut w hen a village has becom e unquestion ably established as a persistent social unit, the personal name o f an outstanding headman tends to petrify into a title w hich is in herited b y his successors. Thus in the village n ow know n as M ukanza, the founder was Kabonzu his first successor was N g ’ unji (I,E z), w hile the next three headm en (I»E3 ; 1,^4 ; I, F5) bore the tide o f Kahali* T he establishment o f this tide m arked as it w ere the social consoli dation and m aturity o f the village, and it is lik ely that the village w ould have continued to bear this tide had not the British Govern m ent intruded upon the continuity o f N dem bu life. W hen taxation was introduced b y the B om a in 1913» d ie village o f Kahali Saluyi (I,F4), as it was then know n, fled to w hat is n ow ikelenge A rea to seek the protection o f the first C .M .M X . m issionary, D r. W alter Fisher. There a great hunter from the village, M ukanza Kandulu (I»F6), killed a roan antelope and quarrelled over its division w ith Senior Headm an Ikelenge, a descendant o f one o f the tw elve headmen w h o had accompanied the first Kanongesha from M w antiyanvw a. Ikelenge demanded a back leg o f M ukanza’s kill* as his right, since he was a senior headman whose ancestor had been appointed b y Kanongesha, ch ie f o f the N dem bu. M ukanza refused to give him the meat and thus refused to recognize Ikelenge’s authority in the area. B u t Ikelenge had m ore follow ers in his ow n area than M ukanza could m uster ; and M ukanza, having heard that the British South A frica C om pany’s adm inistration was less harsh than had at first been feared, suggested to Kahali Chandenda (I,F5), uterine u n d e o f Sandombu, w h o had ju st succeeded Kahali Saluyi (I,F4), that the village group should return to their old site fiv e m iles from the Bom a. Kahali refused to g o and the village divided, some returning w ith M ukanza and some rem aining w ith K ahali. In 1919 M ukanza K andulu died, and some years later K ahali Chandenda, w h o had him self quarrelled w ith Ikelenge’s people, cam e back to the village and resumed his position as headman. B u t in his absence, the village had generally becom e kn ow n as M ukanza, and although K ahali refused to take the title o f M ukanza, the village continued to be called b y that name both b y its inhabitants and b y other N dem bu. It was argued, I
f '- 'H
V - ■/
f)
106 l .= OrC>-
i ... -
o o f'
< C . ••
\ .■
(:) f /"• O
r"> «1...•
Schism and Continuity in an A frican Society
understand b y the siblings and sisters’ children o f M ukanza Kandulu, that the name o f M ukanza was n ow established as the title o f the village. In fact, Kabinda (I,F8), w ho eventually suc ceeded Kahali Chandenda as headman and was still headman w hen I was in the field, w ent so far as to succeed in 192S to the name M ukanza. B y that act h e secured fo r him self, in N dem bu belief, the tutelage o f the spirit o f M ukanza Kandulu (I,F6), in anticipa tion o f his later accession to office. In a similar w ay, Sandombu, w hile I was residing at M ukanza, form ally succeeded to the name o f Kahali. Thus, as early as the twenties o f this century, a cleavage had already developed between the follow ers o f M ukanza Kandulu and those w ho remained in Ikelenge Area w ith Kahali Chandenda (I,F5). Am ong the form er w ere Sakazao (I,H9) and his sister N yam ukola (I,H io), w hom M ukanza Kabinda m arried, thus linking the matrilineages o f nyachln tang ’a (I,D i ) and m alabu (I,Da). From this union m any children w ere bom both o f whose parents belonged to the m ajor matrilineage o f M ukanza Village, and w ho could be counted upon b y their father to support him when he claim ed the headmanship. Sandombu (I,G ro), on the other hand, remained w ith his uterine uncle K ahali Chandenda, and stayed for m any years in Ikelenge Area, w here indeed he secured paid em ploym ent as a road capitao at H illw ood Farm. H e visited M ukanza V illage often, how ever, and when he saw that m ost o f Kahali’s original follow in g w ere trickling back to M ukanza, decided to build his ow n huts there, and eventually to try to succeed there. B ut by the tim e he had settled there a netw ork o f ties already united the members o f the tw o lineages w ho had follow ed M ukanza Kandulu (I,F6). N o t only affinity, but also the alliance o f alternate generations (c£ p. 80 above), held together the senior members o f these tw o groups. In the course o f tim e genealogical generation and biological age had ceased to coincide as between members o f the tw o lineages. This gave rise to a situation in w hich several members o f the grandchild generation o f malabu lineage w ere chronological contemporaries o f several members o f the grandparent generation in nyachintang ’ a lineage, and w ere in fact older than most o f the senior members o f the intervening generation in the latter. D ue to the friendship that existed between these alternate generations it was quite possible that Sandombu, and indeed Kasonda (I,G i5), w ould
PLATE II S u c c e s s io n
to a
N
ame
Headman Mukanza pours out white maize beer at the base of a newly-planted muyombu sapling (see Social Drama II) to the spirit to whose name Yana (see Chapter Ten) is succeeding. Yana is on the left. She will later wear the white head-cloth draped over the muyombu. She will also wear a white bangle. Manyosa, to the right, is anointed with powdered white clay (mpemba), and holds some in her hand with which she will later anoint the whole ritual assembly. White stands for health, good luck, strength, fertility, good will between persons, atonement with the ancestors. The muyombu tree has white wood and exudes white gum.
M atrilin eal D escent
107
be passed over in the succession, and that Sakazao (IjHp), the * grandchild * and brother-in-law o f M ukanza K abinda (I,F8), w ould fo llo w the latter i f he succeeded to K ahali Chandenda (IJF5). This situation gave to Sandombu’s am bitions an edge o f desperation that caused bis desire fo r office to becom e a little too obtrusive fo r N dem bu tastes : he made statements publicly w hich proclaim ed this desire and brought upon him the suspicion that he w ould n o t hesitate to use sorcery to realize it. A n im portant factor operating to defeat Sandom bu’s ambitions in the period im m ediately p rio r to Social D ram a I, w as the absence from M ukanza V illage o f close maternal kin descended from his m other’s m other N yakapakata (I,E i). His sister M angalita (I,G i i ) appeared to be sterile and his m other’s sister’s daughter’s daughter Bibiana (I,H4) was residing o n the C opperbelt w ith her husband. H e could thus invoke no local support from close m aternal k in to counter the pow erful local pressure-group mustered b y M ukanza Kabinda and Sakazao (I»H9). Other.factors .arism gfrom biological and psychological accident conspired to..hardm suspicion agaim t_Sm aom biiJhto_what for N d en ffiu .w as.vk tu d ..^ Param ount amQag..ffig^.was the fact that: Sandombu- had .no. children, . save fo r o n e. daughter— w hom scandal hinted ..was n o t his . child. Indeed, it is certain that w hen he had gone to an urban area to w o rk in 1927 he contracted gonorrhoea w hich seems to have rendered him perm anently sterile. N o w fo r N dem bu .a. sterile m an is often regarded as a sorcerer. It was said that his semen (matekela) was not * w h ite ’ {atoka), a colour sym bolically associated in m any ritual and cerem onial contexts w ith purity, health, strength, piety towards the ancestors, and good w ill towards one’s follows. Instead, his semen had becom e * red ’ (achinana), a colour often linked w ith w itchcraft, aggressiveness, and evil pow er obtained through a w ilfu l breach o f the social norm s. Sandombu, ow ing to his infection, had yello w semen (called * red * b y N dem bu) ; and this was com m on know ledge am ong his follow -villagers, perhaps because he had married successively tw o o f Mukanza Kabinda’s daughters, the first o f w hom died. H e had often beaten both o f them , and caused them to seek the protection o f their parents, w hom th ey had doubtless inform ed, in the circum locutions (iku-didyika) em ployed betw een members o f adjacent generations, o f their husband’s raisfortune.
xo8
Schism and Continuity in an A frican Society
.Sterility is also a misfortune, for an. ambitious-man-in -other respects.. W hen-a man. wishes to..succeeci .tO -office. oE-to found a village o fh iso w n , he looks fo r foe baching o£his-o-VNm childxeri in these.ventures,as w ell astohis.uterine.kin. A man’s major unit o f political support is the circle o f his closest kin, called by Ndem bu ntanga, or b y the plural form antang a. In addition to his ow n and his sisters’ children, these kin include his brothers and their children. Such a group contains the nucleus o f a new generation, the ju n ior adjacent genealogical generation over which he andhis siblings exercise authority and control. Sandombu was doubly unfortunate_in...that his_oxdy .full~sister_JMangalita (I^Qi i ) . was barren, so that. he., had .neither...children.jrior own sister’s children to. support him. fam ily, the fact, that Sandombu . was..a...di3igent_and .tireless gardener, set the seal on the v iew that, he was a.sorcerer^a i««/cyi or mukwa kulowa....It.was .said that he. hacLa familiar, a wooden figurine ,o f the type called katotajiy vAnch was activated—by. the blood o f his previous victims, .and whicb- w orked-beside him , invisible to others, in his cassava gardens, heaping up countless mounds. Sandombu from the fruits .o f his labours, washable to...dispense to many people, most o f them strangers, a-lavish hospitality, and to earn golden opinions fronx them. Thus he aimed to lay the basis for the mclusion .in his futur& village o f strangers w ho m ight be attracted thither b y diis- reputarion for open-handedness. This again constituted a threat, tojhe_closelyknit and partly endogamous traditional com munity o f Mukanza Village. In summary, Sandombu was from m any points.of view an out sider, an atypical, m arginal ,man.-in_Mukanza.. Village. H e belonged to the Kahali faction w ho had remained in Ikelehge Area when Mukanza Kandulu (I,F6) and his close kin and the sibling group o f Sakazao (I,Ht>) had returned to the traditional area where Kabonzu founded die village. H e was outside die mesh o f interlocking reladonships that united die m ajority o f members o f n y a c h in ta n g ’ a ana malabxj lineages. He~was a m em ber^ofthe intervening genealogical.generation,-which at the time., was out o f pow er in die village.. He.was, sterile ,,.and in dustrious, both indications .o f sorcery. l£he_succeede
M atrilineal D escent
109
Finally, he w a s.a ch a n n elth ro u g h jw h ich .s flo w intO-.the tightly, orgaiuzedLand jflghly^xonservative. com niiinity o f a jy p ic a l K aw iku. village. W h o better than, he could, serve as a scapegoat fo r m isfortiines befalling the village ? W h o h etter than h e'co u ld serve to umte- the-potentially-flssile~com ponents o f -thfe remaining-jmembersMp,.. in_.teans_QL-Opp.osiJdotLJteiL. his am bitionsandto. hisJmmoderate_way-of life ? H e tried to attach him self to the chineng ’a sub-lineage o f n yach in tan g ’a lineage and to malabu lineage b y m arrying daughters o f M ukanza, one b y N yatungeji (II,D 3), and one b y N yam ukola. B y this last m arriage hem ade him self the tw o lineages’ *son-in-law * (Ndembu tnuku, properly speaking an affine belonging to an adjacent generation to ego ) . B u t this. very reladonship-is one-involving avoidance .o f fem alem -daw s, _and~unwilling-subm ission- to male in-law s, o f the senior adjacent generation^dtism otLa.xelationship o f fnendship a n d fm u n a rjty. ,-.^et ..stifl^jhis, odd m an, out was audacious enough to aspire to be the mandat the helm . W h y was it, then, that (he m arriage o f M ukanza to N yam ukola proved to be a structurally effective alliance between lineage segments, whereas that between Sandom bu and Zuliyana was fraught w ith conflict ? Part o f the answer lies undoubtedly in the fact that the form er m arriage w as fruitful and the latter barren. It was a source o f chagrin to Sandom bu that Zuliyana had no children and he was w on t to blam e her infertility and not his sterility for this m isfortune. In fact, he paid several ritual specialists to treat her w ith medicines to render her fertile. In Chapter T en I discuss one such ritual in w h ich Zuliyana under w ent treatm ent. B u t other factors w ere, in m y opinion, no •less im portant in rendering the alliance structurally ineffective. T he marriage between M ukanza and N yam ukola did not m arkedly increase the social distance betw een the spouses and members o f the intervening genealogical generation, since the parents o f M ukanza and o f N yam ukola w ere dead, and neither had close in-laws in the senior adjacent genealogical generation. B ut Sandombu, and fo r that m atter Kasonda, ranked as classifi es to ry affines o f an adjacent generation to both spouses, and this relationship b y m arriage, added to the tense relationship between adjacent generations, did nothing to bring Sandombu closer to M ukanza and his w ife. O n the other hand, Mukanza*s marriage to N yam ukola strengthened the antecedent tie o f alliance between
no
Schism and Continuity in an A frican Society
alternate genealogical generations that connected M ukanza and Sakazao, head o f malabu lineage. T he children o f their union w ere Sakazao’s sister’s children and ciassificatory * children * o f N yam ukola’s sister N yatioli (I,H n ). B iological and structural factors, interacting in the process o f social life, lim ited the efiectiveness o f Sandom bu’s cross-cousin marriage and increased that o f M ukanza’s grandparent-grandchild marriage. T he mifniitfulness o f the form er led to frequent quarrels between the spouses and Zuliyana was usually able to run to her parents fo r solace and protection against her husband. The fhritfulness o f the latter gave M ukanza and d ie senior members o f malabu lineage a jo in t interest in the children bom in it. T he relative duration of. the respective m arriag^ at_the_time o f Social ..Dram a .I m ust also_.be taken into _accoxmt. M ukanza’s m arriage had. .stood the exacting test. o f .tim e....Nearly...all the children o f M ukanza and.Nyam ukola w ere fu lh grow n adults, and some had children o f their ow n. „T he m arriage Jwas.everywhere held up as a shining exam ple o f m atrim onial happiness {umluwt) . O n th e other h a n d ,...Sandombu.. had. o n ly recen tly.. married Zuliyana, a pretty and rather_dighty...girl, and., the first years o f m arriage am ong N dem bu. always. ..possess., an.-experimental character, frequently term inating rin divorce. E ven in 1953, Sandombu was still almost pathologically jealous o f Zuliyana. O nce he showed m e a spear that he kept in a com er o f his hut. He said that he w ould stab anyone w ith it w hom he found m aking advances to Zuliyana. H e did not trust her for a m om ent out o f his sight. B ut he was determined to keep his marriage in being, fo r although it was unsatisfactory in m any respects, it still gave him a footing in the village com m unity. I have. set .out. the sociological and historicaLbackgrouncLto this first-social drama, a t.some, length, and deJineated-the-personalities o f its m ain protagonists,, because ,w ithout- this. preMminary- w ork it w ou ld be impossible., to make., a...satisfactory-analysis- o f the subsequent social dramas, each o f. which--represents -a- further stage, in the conflict over succession, im plicit in the matxilineal structure o f M ukanza Village. Itris evident from the events., described, „at. theL.b.egfrming o f SoriaLJDrama I, that tension had existed in. .the., relationship
M atrilineal D escent
Ill
between K a h a h C h a n ^ for sQmejdm e hefo the,.episodes in vo lvin g meat disiribution had.,,precipitated a situation o f open hostility. M y inform ants told m e that Sandoxnbu had neglected fo r several years before this to supply his w idow er uncle w ith cassava meal from his ow n gardens. N yam waha (I,F7), M anyosa (I,G i3) and Kasonda (I,G i 5) had com bined to feed thè old man. Kahali Chandenda, even b y Ndem bu standards, was a very poor man. Part o f his leg had been ; amputated, and w hen he travelled he had to enlist the help o f a vounger man to push him on a bicycle. O n the credit side o f the i- balance, he had. been a famous chiyang’a, a hunter w ith a gun, and í was in addition a great 4law -m an ’ (ihaku), ju d gin g and advo cating cases w ith forensic skill and know ledge o f custom . B ut he w as old-fashioned and could n ot cope w ith the m odem duties o f a headman. H e was also feared as a sorcerer. Hunters are thought to possess exceptionally pow erful fam iliars ; their strength and skill in killing animals is acquired initially from killing their ju n io r relatives b y m edicine. Thus KahalTs accom plishments belonged to a rapidly passing social order, and physically he was a burden to the village and a reproach to Sandombu. H is reputation as a man w ith bad medicine was all dnat restrained the hostility, felt b y several persons against him, from breaking out. Sandom bu, w ho had lived and w orked in Ikelenge Area most o f his life, had also gained a certain notoriety as a sorcerer there. I have been told b y members o f a village in Ikelenge Area in w hich he lived fo r a lon g tim e, that w hen he left it finally, a large hole w as found under the door o f his abandoned hut, in w hich it was supposed he kept his ilomba, a serpent fam iliar w ith a human face. I f there really was a hole, it was probably used fo r urination at night. I have seen similar unhygienic devices in huts else where in M w ind unga, and N dem bu consider that people w ho have these holes are disgusting. But, w hatever the * circum stantial evidence \ it is certain that Sandom bu had a bad reputation as a practitioner o f the black art am ong the local inhabitants o f Ikelenge Area. I m yself have often discussed sorcery and w itchcraft w ith him , and although he always spoke indignantly against it, he had undeniably a w ide know ledge o f its putative techniques and materia medica. LLjs_passibIe„, that. Sandombu, fm ding . him self dubbed a sorcerer, on .account_o£ his „.sterility,
t) n
o D e o o o
F
o O o
c o
0 t? 0 c :o c o
Id £:■
X €.) X "D
X)
112
Schism and Continuity in an A frican Society
.deliberately, fostered the belief,, by innuendo. when.sober,.vand by the, violence ofhisbeha'viour when in lus cups^uLoxderL_to inspire feat, in people and so to get his ow n w ay in a number o f situations. B ut-it was even m ore in the interests o f his rivals. £bt.,thc,village succession, - particularly MuJkanza. Kabinda and Kasonda, to harden suspicion o f ms sorcery into popular-certainty_„and to establish it as a local-dogm a. N dem bu headmen pray to their ancestors before boys" .and girls* initiation rituals : * G ive us in this village o n ly the meat o f animals, not the meat o f m en/ This means, * Exclude sorcerers and witches from our m idst/ i f the headman him self is a sorcerer, in N dem bu belief, the village is in terrible danger, fo r it is thought that headmanship adds additional pow er both for good and evil to w hat a man already has in the w ay o f m ystical acquirements. I f Sandombu w ere branded w ith this evil reputation, his chance o f succeeding to office w ould be eliminated, except under special circumstances w hich I w ill mention in the analysis o f a subsequent drama. B y -g iv in g Kahali an inferior .share o f his ow n .meat and by eating a m ajor share o f Kahali*s meat, Sandombu made it clear that.Jb.e no longer respected Kahali. as an uncle and a-headman, and...would be glad to see the last o f him . .T h e fact-that in the subsequent quarrel M ukanza people alleged...tha.t-Kahalk.too. used threats o f sorcery against Sandombu^indicates that they- con sidered that uncle and nephew were-eq.ualiy .evil.-..From-this the people argued that their lineage..should no lon ger enter.into the succession. It m ay be asked w h y, i f Sandombu had the reputation o f being a sorcerer, people thought it was necessary for him to go to the black-hearted old Sakasumpa to enlist his aid against Kahali. The answer lies in the belief that w hen a man wishes to k ill an office-holder, such as a headman or a chief, he does n ot w ish to take the fu ll responsibility him self fo r the sorcery. It m ay be also that the diviner, to w hom the kin o f the dead headman may repair, w ill be deceived b y this device. There are innumerable Ndem bu tales, told round the chota fire or in the privacy o f huts and kitchens, o f claimants for office w ho have sought fo r sorcerers to slay b y m edicine the present incumbents. It is told h ow pay ment o f guns, cloth and m oney, and form erly slaves, w ere made for these services. It is also told how a great diviner w ill detect
M atrilin eal D escent
113
this device and h o w he w ill lay the blam e for the sorcery on the claimant and not on the sorcerer w h o actually * shoots the medicine ’. He was only doing a jo b for w hich he had been hired, the people say, a w icked jo b it is true. B u t the real guilt lay * in the liver *, as the seat o f the w ill and source o f pow er, o f the relative w ho wanted to succeed. In such divinations the name o f the hired specialist is never m entioned, perhaps because o f the m ode o f questioning em ployed b y diviners w hich makes the initial assumption that kin bew itch kin. O n ly i f they receive negative answers, on consulting their oracles, to questions em bracing the w hole gam ut o f kinship, w ill they seek fo r a sorcerer or w itch outside the victim ’s kin-group. This point is relevant to the analysis for later w e shall find Sandombu accused b y the headman o f a neighbouring village as the sorcerer called in b y that headman’s relative to bew itch him. In jfcjcta rra tive .o f. thebcw itching,.Q £K ahali,.coloured,:though it is by>the inclusionjQf.m ythical. m aterial, .we.detect...the .presence o f those elements considered- ty p ic a lo f aULSQciaLdramas._Th.ere had keen ,a breach in the custom ary xegularit^..pf social relations between Kahali .and_.Sandombu,.and betweerL the-m em bers o f M ukanza village as a w hole. This breach- took on. decirive~and dramatic form as th e . breach-of_a cruciaLnorm ..-governing., the relations o f m other’s brother and sister’s, som in -bidembu society. A sister^.on.shauid^give.hia-Uterine_ uncle„a spLe_cific_j.oiiit„of meat w heh-heJias m ade.aJrill.-... This, is o n e o f the. m ajor. obligations inherent in the relationship. I f i t is n ot .carried, out,, w ithout the excuse ..of exception al. circumstances, this is tantamount to a refusal on the part , o f the nephew ..to_ accept.. the_continued authority .of..his. .uncle... B ut it. goes, further, than „that. ...It is a challenge to one o f the., m ost. deeply , entrenched.„principles o f N dem bu social organization, that b y w h ich the .senior adjacent genealogical generation is given authority, over..the.._.junior generation, and the latter m ust-respect and . obey the .former. W here generation authority, is linked w ith m atriliny,. as. in the uncle-nephew relationship,.the.authority o f the .senior.generation is exceptionally strong. It is n ot sim ply a breach in the relation ship-betw een tw o persons, but a m atter involving, the w hole group ; for it is a challenge to the authority o f the. senior genera tion as-a w hole and ultim ately to the system itself.
114
Schism and Continuity in an A frican Society
H ie challenge is also one w hich not only jeopardizes the-general authority structure, but w hich involves succession-to-ofÉce in the group. It is argued by N dem bu that no~one woulcLmake such a challenge against gérontocratie authority, and assert his claim to o ffic e , unless., he .possessed-special-powers,...beyond the pow er inherent in his kinship position alone. T o m ake such a breach in the social order, an order validated h y N dem bu ani m istic religion, a man m ust have powers o f an anti-social kind, w hich in this society are hostile to kinship. T o counteract these powers, the forces o f the constituted order must be m obilized. This is done b y a conscious, and public statement of-official-norms o f correct behaviour and b y a public condemnation o f the-actions o f the delinquent. B u t w h y, in this case, was no recourse made to a diviner,, w ho w ould alm ost certainly have given the M ukanza.people the answer they desired, that Sandombu was a sorcerer and-should be_expelled from the com m unity? T h e answer given b y the com m unity that Governm ent w ould get to hear o f it and take action-against „the, accusers, is im portant, b u t it. is .not .the. w hole tru th .,. A n ..active, and., aggressive—m an. like Sandom bu. m ight havereported to the Borna an accusation o f sorcery ,made .against him. I say he * m ight have reported ’ the matter because this betrayal w ould have led to his banishment from the. village, w hich he did not w ant. B ut the people feared he m ight so report .them, in w hich case the other senior men o f the village, w ho made the charge, w ould undoubtedly have been sent to prison. A nd it is fo r this reason that N dem bu rebuke. Govern m ent in private discussion for * protecting witches and allow ing them to m ultiply*. B u t over and above this cogent argum ent there was the im por tant consideration that, objectivelyspeaking,- Sandonibu^was an asset_to.. foe—com m unity. JBe was. in. regular em ploym ent and hisr position as capitao enabled him to obtain labourers*, jobs at H illw ood Farm fo r other men in the village. He was said.to.have the ear-of M r. Fisher, the em ployer, and-could flnd-places-for his ow n kin.. Again, there w ere several-boys -in the village.w ho had passed Standard II at the local mission out-school.and required m oney fo r their boarding foes at the Middle. School , at. Kalene Mission. These boys w ere partly financed by Sandombu...... From Kalene they m ight go on to Mutanda School at.Solw ezi where,
M atrilineal D escent
115
ifth eyp assed StandardV I,. they_^otddjobtam_a^ fam ily lines had been at-least tem porarily-averted. JSandombu haiLno.t-been- expelled, from . the. wllage^.otit-had-beeti -excluded from headmanship. T h e village had ..maintained its unity and even strengthened it b y finding a scapegoat fo r-its . internal dissensions. B u t at this stage Sandom bu was still sm arting from his defeat and further outbreaks o f hostility on his part occurred against those w h o had ousted him . T h e m ost serious o f these led to Social D ram a II, w h ich brings out clearly the. lines o f tension in thejvillage structure o f M ukanza.
o n
ii6
Schism and C ontinuity in an African Society SO CIAL D R A M A II The Expulsion and Return o f Sandombu (compiled from informants)
( (.;;
i..
v
G
Sandombu bad married Mukanza Kabiizda’s daughter» Zuliyana (IJ2). It may be supposed that he hoped by this means to attach himself securely to the tw o village matrihneages o f CHIKENg’a (r,E4), and malabu (I,Da), since she was connected to both through her mother and father. He also hoped that she would give him children, for although he had married eight times, no wife, except his first, had borne him a child, and this was thought not to be his. He had divorced Malona (see Social Drama I) before marrying Zuliyana, because she was not acceptable to Mukanza people, being reluctant to give them food and hospitality. .One _ojfthe__psychological drives behind Sandombu’s persistent . and openly , expressed desire for headmanship probably consisted in the-£act..that_the_names o f .sterile persons are not given to children, -or rather, to~ put it in Ndembu idiom, do not * come back.to. children.to-gLve_them a name *. Their spirits are called ayikodjikodji (sing, chikodnkodii) and in ritual concerned with female fertility are given special offerings o f beer and food by way o f placaidon at one side o f the temporary shrine, to keep them away from the afflicted woman patient. Sterile people * die fojr ever \ in that their names are not remembered -and..no..i»«yowi« shrines o f quickset saplings are planted to honour them... Sandombu probably feared that his name, i.e. his social personality, would be forgotten, unless he became a headman, in which case, ho-would be remembered as the occupant o f a poUticaL position. After a year Zuliyana had borne no child although she came from a fertile family, and Sandombu, while he was living at Hillwood Farm, used to vent his spleen on her by frequently beating her. Towards the end o f 1948 an Nkula ritual was performed at Mukanza Village for Nyatioli (I,H n), sister-in-law o f Mukanza (I,F8), own sister o f Sakazao (I,H9), and classificatory mother o f Zuliyana (IJz). Nkula is most often performed for women with menstrual troubles, and it appears that Nyatioli was experiencing a difficult menopause, with much menstrual bleeding. Sandombu, who was by this time working as a Road Capitao for the Public Works Department in a camp about a mile from Mukanza Village, was under the impression that an Nkang a girl’s puberty ritual was to be held that weekend, and not Nkula. He ordered Zuliyana to brew many calabashes o f beer for the occasion, for at Nkang*a nearly all the people in a vicinage attend. A t the concluding phase o f Nkula many people also usually attend, but this was only a * small Nkula *
n8
Schism and Continuity in an African Society
confined to M ukan za V illage and a few external kin an d neighbours. Z uliyana, w h o w as living in M ukan za V illage itse lf a t the time, brew ed on ly tw o calabashes fo r this sm all ritual. W h en Sandom bu cam e o f f d u ty— m o st N d em b u rituals to d ay are h eld a t weekends to accom m odate those at w o rk — he fo u n d on ly tw o calabashes o f beer. W ith ou t w aitin g to hear the explanation he began to beat Z uliyana, w h o called o u t to her kinsfolk fo r help. K aso n d a ( I ,G i 5) entered San d o m b u ’s village hut, and accordin g to his o w n account, rem onstrated in a broth erly w ay w ith San dom b u. T h e latter’s only response w as to take a b am b o o cane and thrash Z u liyana w ith it. T h en a y o u n g m an called Benson, a m em b er o f the neighbouring village o f N g ’o m b i, ju st back fro m the C o pperb elt, cam e n , snatched the cane fro m San d o m b u and broke it across his knee. Sandom bu lost his tem per an d bis head com pletely, accordin g to B en so n and K ason da, and accused B en son o f bein g Z uliyana’s lover. Others, b y this tim e assem bled in the d oo rw ay, heard him g o o n to say that he w o u ld m ake the m em bers o f M ukanza w h o w ere interfering with his m arried life p a y fo r their intervention. N e x t day, San d o m b u w as observed b y K a n y o m b u (I,Fp), M ukanza K abin d a’s y o u n g er brother, and by their sister N y a m w a h a (I,F 7 ), as h e w as stoo p in g ov er the footpath betw een M ukanza and B en son ’s village o f N g ’om b i. T h en he circled the huts o f M ukan za and N y am w ah a. H e w as heard, it w as alleged, to have in v ok ed the spirit o f his deceased parallel cousin Lu pin da (I,G y), a hunter an d pow erful personality in his life-tim e, to punish his enem ies. H e is then alleged to h ave said to those sitting in the chota, ‘ I a m n o w returning to Sak eji [i.e. to Ikelenge A rea]. T o m o rro w som eone w ill die in this village. M ukan za peop le h ave n o sense.’ E a rly n ext m orn in g w hile the aged N y am w ah a, sister o f M ukanza and m other o f K aso n d a, w as p ou n din g cassava, she felt a spasm o f pain and fell d ow n in a helpless condition. K ason da, w h o h ad been w o rk in g as a gard en b o y fo r the D .C ., w as in form ed o f this and obtained perm ission to return to the village, five m iles fro m the B o m a . N y am w ah a w as gravely ill. H er last w ords before she died w ere ‘San d om b u has killed m e. K aso n d a w rote an. an gry letter to San dom b u, retailing his m other’s last w ord s an d d em an d in g th at he return to M ukan za and g iv e a foil explanation o f his suspicious behaviour. San d o m b u cam e back and all the senior m en o f M ukanza assem bled in the chota to hear w hat he had to say. It w as p u t to him b y M ukan za that he h ad cursed the village b y in v okin g. L u pin d a’s ghost. T h is fo rm o f curse is called hushing*and b y N d e m b u and is classed w ith sorcery ( w uloji ). San d o m b u denied that he had cursed the village in this w ay , b u t he ad m itted that he had been an gry, and expostulated that he w o u ld never,
©
©
M atrilineai D escent
0
119
under an y circum stances, have killed his ‘ m other \ w h o m he loved. H e to o k o u t a 10s. n ote and said : * H ere is m y m on ey. G o yourselves to a divin er in A n go la, and find o u t the truth.* (T h e P o rtu g uese, accordin g to N d e m b u , d o n o t regard w itchcraft accusations as a crim inal offence, so that divination is le g a l inutheic territory.) K an y o m b u (I.Fcj), M u k an za’s brother, to o k the m on ey a n d dem anded .£ 3. xos. in addition, sayin g that the com pen sation he sh ould p a y fo r causing N y a m w a h a to die m ust b e set at ¿ 4. It w as n o use, he w ent on, den y in g that he h ad threatened the village w ith sorcery. M an y people h ad heard h im , so that there w as n o need fo r a divination. B u t San d o m b u insisted w ith tears in his eyes that under n o circum stances w o u ld he k ill his m other, o r in deed, an y o f his kin . K an y o m b u then said that after the m ou rn in g ritual (Mudyileji) San d o m b u m ust leave the village, and stay aw ay until he sh ow ed b y his beh aviour that he could liv e p rop erly w ith his kin. T h e others agreed, an d San dom b u left the village fo r a b o u t a year. I 11 the end he h u m bly asked M ukanza fo r perm ission to return. In the m eanw hile, in view o f San d o m b u *s evident sorrow at N y a m w a h a ’s death, public opin ion w av ered as to the cause o f death, and secret divination into it w as m ad e b y m eans o f a m edicated po u n d in g -p o le {ngotnbu yamwishi), and blam e w as fin ally ascribed to the husband o f N y a m w a h a ’s d au gh ter’s d augh ter fro m C h ibw akata V illage in the sam e vicinage. It w as alleged that N y am w ah a h ad tried to keep her gran ddaugh ter in M ukan za w hile her husband w an ted to take her to his ow n village. In an ger he h ad * sh o t m edicine ’ at her and killed her. S an d o m b u w as absolved fro m b lam e and allow ed to return. B u t in order to sh o w finally that ‘ his liver w as w hite ’ tow ards the people o f M u kan za, d ead an d alive, he p aid M u k an za a g o at and to o k part in a village ritual perform ed to placate the restless spirit o f N y am w ah a w h o had caused m an y p eop le to d ream o f her and caused a m inor epidem ic o f illness. It w as th o u g h t th at she w as disturbed by the * troubles in the v illag e \ S an d o m b u plan ted a tmtyombu 1 tree to his ‘ m o t h e r ’ in fro n t o f M u k an za’s hut, and M an yo sa (I,G i 3) inherited N y a m w a h a ’ s nam e. B o th S an d o m b u and M ukan za prayed to N y a m w a h a, and m entioned that th ey w ere n o w reconciled. T h ro u g h the plan tin g o f the tree, th ey said, she w ou ld be rem em bered b y her relatives. Finally, three lines o f p ow d ered w h ite clay w ere sprin kled o n the g ro u n d fro m the base o f the tree b y M ukan za, and all the m em b ers o f M u kan za V illage w h o w ere present, m atrilineai,
0 0 m
0-
@
0 © @ o
©
© © © ©
Q Q
Cl
©
© © © 0
o o Q o
1 Trapneli and Clothier, p. So, give the botanical term for this tree, which they call mwyombo, as Lannea antiscorbutua. C h eck L ist, issued by die Forestry Office, Northern Rhodesia, identifies mwiombo [sic] as K irltia acuminata.
o
o
c ©
120
Schism and Continuity in an African Society
patrilineal and affinal, were anointed by Mukahza with white clay, symbolizing the basic values o f Ndembu society— good health, fertility, respect for elders, observance o f kinship dues, honesty, and the like. The prodigal nephew had returned. Analysis In,, this drama the breach o f relations w hich ..inaugurated the series o f events was a quarrel between a married ,couple w ho not only stood to one another in the relation o f husband and-wife but_were also cross-cousins. Furthermore, since. Zuliyana (I,j2) was jaxhild o f both the village matrilineages whose -members had effectively prevented him from succeeding, the cleavage between Sandombu and the rest o f the village entered., into...his. m arriage relationship w ith Zuliyana. As usual, the .breach .of relations received visible expression in the violation o f. a cruciaL norm governing those relations. In this case the norm , governed marital behaviour and can be stated as follow s.;. * N o husband should beat his w ife violently w ithout good reason.or .without a..fault on her part/ Znliyana.-had.brew&cLa^sinal 1enam ount o f heexJthaa her. husband-had.otdered^not-because-sho.was lazy or,Li,wilfully, disobedient, - but—because-Jhen. husband- had mis understood. w hat cerem ony was to be perform ed. Sandombu probably.thought, that she shouldhave.brew ed ..aJarge amount of-beer in any case, whether the ritual w as a m inor one. or not, sim ply because he. had ordered her to do so. T he fact, that she had not done so w ould almost certainly have been interpreted by him as agreem ent w ith her .village relatives .and-jdisohedience to him, her husband. B u t to the other members o fth e . village, his beating o f his w ife must have seemecLa- breach o f the norm that a m an must n ot beat his w ife -without-good. reason. T h ey felt m orally justified in intervening^ although actual physical inter vention was left to an outsider from another village. Their dis approval o f his action was im m ediately interpreted b y Sandombu, and probably not w ithout reason, as a recrudescence o f their hostile attitude to his claims for headmanship. He lost his temper, as I m yself have know n him to do on m any occasions, and reviled everyone w ithin range. It is impossible now to find out whether he m erely reviled (ku-tukana) the villagers or cursed
M atrilineal "Descent
121
(ku-shingana) them . It is allow able, though frow ned upon, to revile in N dem bu society. O utlets fo r suppressed feelings are sometimes venial in a small-scale society w here personal relations are intense. B u t cursing, since it involves the raising o f a m alevolent ghost (musalu) to kill, is illicit, a form o f sorcery. Sandombu’s rivals say he cursed the v illa g e ; he him self admits only to having reviled them . U nluckily fo r Sandombu, the death o f Nyam waha,...like the deatbuof Kahali ( fF j) on a previous occasion, occurred ju st w hen it was m ost lik ely to confirm die w orst suspicions.ja£.his_.£ellow villagers. H e was publicly denounced as.,.a-«i«/q/L.(sorcerer), and in spite o f his genuine and obvious, g rie f wJhich-madeLa-profbund impression, he was tem porarily banished.. fr.Qm Jb.e_.village. Unlike_Dram a I,., the., mechanisms .used, to seal. .up. the. breach between Sandombu and ..the rest o f .the village w ere of. a., m ore form al-and institutionalized nature. In..Drama..I,. it-w ilL b e re~ callecLthat although.the village, inform ally. discussedJSandombu’s behaviour in the chota, he him self was not present and na..action was. taken against him. JustL. enough m obilization..o f ..public opinion was made to bar effectively his w a y to the headmanship and.-ensure that M ukanza Kabinda should succeed. .. But. in Dram a II Sandom bu was form ally excluded from the. village fo r a probationary period by a general assembly o f the village. Later. doub.tarosem them in dso£ village..m em beE sw hether-tbe. grave a«LUsation m ade against him was.just^and-by. one-m ode o f divina tion- he was adjudged innocent. Therefore he was. received back into the fold b y means o f a ritual w hich reasserted the. solidarity o f the lineage of. jstyachintang’ a (I,D i ), to w hich .not only Sandomb.tLb.ut_.alsQ..headman.. M u kan za. Kabinda,^.Kanyombu, Ka.soncla, and. M anyosa, among.. the ..living...belonged,. aiid exheadm an.Kahali Chandenda and. the victim Nyam^vaha. am ong the.dead. It-w U lhe n o te d th a t ..although, action. .was„.taken. against San dom bu, the m atrilineal lineages o f M ukanza V illage were_cateful to appear to be m orally justified in their actions and.t.o_avoid..o.pen violence, against him — F.or.example,..it.w as B.enson, an ojutsider, w ho—broke -Sandom bus. .cane. Kasonda. S^ d om b u ’s classificatory_bro.ther,. confined h im self,to.., mere. remon&tration.—..Kanyom btbL M iikanzas..brother- and n ot the h eadman M ukanza him selfo. pronounced the sentence o f. exile ; . and._.this__was a
122
Schism and Continuity in an A frican Society
tem porary, and not., a permanent,,..exile._T h en.^ decision. was obtained from d^\dnatiQn_whicK allocatecl blam eTorM yam w aha’s death to an outsider, and not to Sandombu.....Mo. attem pt was made to..bring,, a divorce suit, on, Zidiyana!s.b.ehal£.agamst Saxidombu. b y her. relatives .o f. m a ia b u .lineage.— T h ed aor-w as left open for. the reconciliation that fin ally occurred....-At-the same tim e it . was. m ade clear.. to..alL-who_.might have-supported San dom bu in the village that he was com pletely unsuitable for the position o f headman. B y ju ra l and. ritual ,means the. crucial norms.of.kinship-and affinity had been reaffirmed,.and Sandombu had.been, shown not to have conform ed w ith them. Sandombu had been com pelled to accept his ju n ior status in the new order o f things in which M ukanza Kabinda was village headman. Sandombu had been reintegrated into the./village com m unity, but .the. condition o f his -re-admission .was that he had to return ,as a-penitent., and to renounce his claims to office. Conflict.had .been,...temporarily m itigated but. not,...of co.urse,..eHiiiiciated__ItsJnevitafale revival la.yjCQiled-withm~the.sociaLstrucfcure.-Thc-i'/omiiii^-the-poisonous snake, w ith. the..human. .face...which, is .used, by- Ndembu-sorcerers as a.familiar,.^vividly_symb-olizes..the.-secret.malignity.„and_s.udden ,| em ergence o f .masculine jealousy, o ver, succession-to office. pQcess o f both jhese, dramas has been shown to. run from | : '/ :br<^cft@arough crisis and redress-to. reintegration._Inform al and i^ S^ ^ ^ B ijiiral.-m echam sm s, and ritual, w ere -brought--into play to.restore the..equilibrium of:the. disturbed. so.ciaI..gro.u.p.s.._Is there myJwayimwHch...we-mayjdis.tinguish.b.etween.the..impIicit..ainis and-functions o£.these..difietentmodes-o£redEessing-and.adjusting | conflict,-—ju ral.an d ritual ? It-would-seem ^at-the-present stage o£.anthrop.ological -enquiry, - that .jural... m achinery_is„em ployed when conflict between persons, and.groups. is-couchedJln-terms o f an appeal b y both contestmg_parties..to.-a..com m on-norm ,1 or, when norms conflict, to a com m on fram e o f values.which_o.rganize a society*SJiorms into a hierarchy. Thus, when Sandombu spoke angrily to his relatives, they all accepted a com m on norm govern ing social intercourse— nam ely, that no person should curse 1 C£ Professor Gluckman’s discussion o f how Lozi judges work with norms in cross-examination in T he Judicial Process among the Barotse o f Northern Rhodesia (1955), pp. 49~SO.
M ain lin e d D escent
123
another b y raising a pow erful ghost. In other w ords, human beings w h o w ere members o f a single society had an interest in one another’s w elfare and should not w ish one another dead. In N dem bu society the w ish is thought often to be father to the d eed ; in fact, w hen it is publicly expressed, it is stated b y custom to have the m ystical p ow er to do the deed. Thus, Sand0mbu..argued...dbathe. had w hile his reladves held that he.had-not reviied^butJha
124
Schism and Continuity in an African Society
w ith the norms must be purged o f their socially disruptive-quality ifth e group is to remain integrated. B itual is the saciai.mechan ism/by which a group is purged of. the.. anarchic_and.disruptive impulses which threaten its crucial norms and values.- These impulses are present in the m ajority o f .its members.-.and.~come dangerously near to overt expression i f there, has-been.a_.!ong series o f quarrels between its members. Hut this temptation to rebel and even revolt, may also..be.ied..by constant disputes between individuals and factions,-which in fact manifest deeper conflicts between different sociaL.pmcess.es. In Ndem bu society, we.have..discemed_.thatn.cer.tain~-principles o f organization, which appear to give rise to consistent-norms, in practice originate processes which., w ork against _.one_another. O ver-a-period o f time,. conflict-o£process_Jeads_tc^breaking o f particular relationships and fissionL .o f particular-grQups^-but new relationships-and--gro.up&..on.the old-patteoi-areuestablished— The contrary processes- compensate.one another, within. the-.s.tnucture o f the changing,. but-, repetitive,-...social- .field.— untiL.the processes have worked out their mutual._compensation.^_the conflict between . them w ill- continue_to ._produce_naggingIy persistent quarrels, which are ineradicable- untilne-w-relationships are_formed. In a society, which is. not.changjng,..the_sa.ai6_pro cesses w ill then produce similar quarrels, in the-new_relationships.
Q m this view o f society, norms and. their., supporting values can only appear to be consistent, since they must-.co.ver the presence o f contradictions w ithin .the., structure, itself. The contrary. prQcess.es.-are.likej^.each_to„be..s.tated-imnQrms....Hence situations must arise where the. norms..which—determine the course, o f action, to be taken ..cannot be. clearly ..and ..consciously afiirmed for the acceptance o f all parties, since, each _can_ claim some'support from customary values.- It is here-that intrigue may? become rife and. disruptive. In Ndem bu .society,....based as_it is on close interpersonal ties, it is also at. this stage-tiiat illfeeling becomes charged w ith the malevolent, mysticaI_power o f witchcraft and sorcery. Ill-feeling is not merely, im m oral: it-is charged w ith the danger o f disease, -death-and-.other mis fortunes to one’s fellows. Here judicial ..decisiomcan .condemn_6ne. -OX,mor.e..Q£.th^.. dis putants, but it.cannot always, relieve the quarrels so .as. .to preserve the threatened relationship. Accusations o f sorcery_and witch-
M atriîineal Descent
125
craft m ay tem porarily emphasize for. tke, parries_the„m ystical danger, o f their ill-feelings, as happened to Sandomhu- w henJiis * m other * died. Ritual,, like beliefs in w itchcraft, is a custom ary device w hich lifts the emotions o f peapieand-the. vahies..o£ society to a. m ystical plane where they have power-beyond-their-secular effects. It m ay he invoked to affirm th e uniryoTa-relationship or group over and above the quarrels which-are rooted-in-confficts between their organizing social principles. It is significant-that as accusations o f w itchcraft and sorcery follow . orL._what. we: call natural misfortunes,—which._ate.__as.crib.ed_ to_. illdfeeling.. in the group, so .similarL.situations,-Eather..than.-open. dispute,JeacLto-the perform ance..of- rituals.1 -H ence, .it_seems.to_ me,_.afterL_amajor crisis^roduced b y these conflicts, o f process,. rituak ace-employed to affirm that reconciliation has been achieved. _These__rituals may-'even be em ployed after a ju d icial decision appears -to have settled rights and wrongs,, when...in fact the cause_o£_.disp.ute is beyond settlement. -And finally,., w hen all-attem pts to preserve existing relationships have failed,, final breaches, -often-provoked b y w itchcraft charges, are confirm ed b y rituals whicburestate the norms as consistent and enduring, even ,though ,n ew .relationships have been established. Those w h o disputed bitterly fo r headmanship, w ithin a village, m ay-becom e helpful relatives-when, they reside in tw o diffèrent villages w hich each appear...to_confbmi to the JSfdembu ideal. In the tw o dramas set in M ukanza V illage, w.e-hav.e_ seen a series o f these canflicts-and_a-series o f attempts-by. members, o f the village m atrilineage to resolve them , , first .by.inform al,, then b y form al, jural measures. W ith the expulsion of-Sandom bu a new leuel o f .conflict had been reached. D oubts as to the..justice o f their action in banishing him seem to have beset th e. leading members o f the. village. T h ese.found_.expressian.Jn..the...result o f .the divination and in the b elief that the spirit -o f N yam waha was -haunting the village. T h ey appeared-to..represent the uneasy stirrings o f a sense o f collective guilt. T he people o f the village had exiled their kinsman, in w hom -flo w ed-the blood o f 1 C f. the view s expressed b y M . G . M à rw ick in * T h e Social C on text o f C e w a W itch Beliefs \ Africa, x x ii, 2 and 3 (1952)» and b y J. C . M itchell in * A N ote on the African Conception o f Causality \ The NyasahndJournal, v , 2 (July, 1952), pp. 51-8.
iz ó
Schism and C ontinuity in an A frican Society
collective life, a man w ho had in m an y w ays.treatedthem gener ously ; and perhaps they felt that theyhad sirmed agam stjhe.deep value set on kinship itself as an organizing principle o fso cia l life. I go t the impression from the w ay they talked, that...they felt gu ilty lest, in order to further their sectionafaiidpersonal interests, they had excluded a brother and^a .benefactorsJroixL_participation ..in their jo in t existence as a com m unity o f hlo.o_d.and place. W hen Sandombu expressed his sense o f penitence and asked to return, there follow ed not m erely an expression o f his ow n con trition at the tnuyombu tree o f his * m other but prayers to her b y her brother the headman, and b y her very ow n daughter that aÚ o f them should be forgiven. Xhe life .o f the. com m unity should begin anew w ith Sandom bu..in...its..m idst-once more. T h e -tnuyombu tree,, plan ted ...because- there, . had- been-conflict between people and w ithin people themselves, _now.stands, and w ilkgrow . and .branch-out .as. a sym bol, that peace, is. restored and the values-.ofJkinship are once m ore renewed w ith fervour. Judicial mechanisms., tend ..to be. Jn.vo.ked_ tou.rcdr.ess...conflict, where the conflict is overt,, and these judicial mechanisms-involve rational investigation into the m otives and . behaviour o f the contending parties. R itu a li mechanisms tend to. b e utilized where’ conflict is at a d eeperlevel. Here, conflict-expresses itself through projection— that is ,. in the collective association o f misfortunes w ith ill-feeling and the w orking o f m ystical beings and. forces, w ith dreams, and w ith answers to divination,. T he conflict is between norm and impulse in each individual.m em ber o f the group, but since in a tightly-knit group similar impulses assail com m on norms in situations w hich em brace.the, w hole o f its membership, this conflict attains social recognition. ..through repetition dow n the, years, and ultim ately cultural techniques are devised to handle it. It appears to the. members o f the.sQ.ciety as though mysterious forces are attacking the. v.ery_foundations o f the .moral and social order, not from w ithin, but.by.projection from w ithout, in the form o f witches, spirits and mystical, powers w hich penetrate individual members or some representative individual in the group in the form o f dreams, illness, infertility, madness, etc. V iew ed in this w ay, the tnuyombu tree planted to the dead * m other * N yam w aha m ay be regarded as a means by which the forces which are potentially disruptive o f the village
M atrilineaí D escent
127
are drawn o f f ; and die conflicting members o f die group unite, purged o f anger, in am ity o f com m on worship, around it. The very.tensionsan daggressionsjhat arise.in this social system thus become means to give its norms a new charge.of^energy so that nat_anly._are.they, rein stated h u t.alsoth ey.areth ejn orefferven tiy accepted. Thus, Sandombu m ay w ell have w ept because he believed, w ith other Ndem bu, that his norm -breaking behaviour and disloyal self-seeking impulses m ight have caused Nyam waha’s death. Pressure is^exerted b y such beliefs ..on the individual to conform to. the norm s o f N d em b u so ciety, .because,- i f h e per sistently breaks them , people w hom . he. lo ves-w ill-d ie. The overt expression o f ill-feeling, especially during periodsjafxrisis in the affairs.pf a group,, is thought to. render, the. ohjectsjof.anger vulnerable to m ystical danger. It is thought,..for.example^Jthat the fomiliars o f w itchcraft and sorcery are emb.oldenedLby._the anger/of their owners to dem and the * m eat ’ ó f tliose_against w hom their owners are. incensed— If.hasty_..anger_iipens into a ptatracted grudge (chítela) the fam iliar cannot be. gainsaidin its demand to kill. V iolen t persons are thus constrained not on ly by the-public opinion that they are witches, but also -b y them ow n fears that inadvertently their anger m ight k ill ..som eone, they love. A_further...diferen.Ge_betwc.c.n. the. kinds. ,_o£.situatiQn. ^ jud^cial..md_ritual machinei^_.o.f..ady.us.tment..ar.e respectively: em^ ployed, m ay here be considered. .R ecourse is often mad^To law when the livin g quarrel. T h e nature o f the conflict brought to lig h t is usually specific attemptS-are..made.to findLsL.remedyr and i f possible, to bring about a reconcihationbetw een the opposing parties. T h e conflict is man-made and im m ediately intelligible. B ut- w hen a breach in social regularity is made b y _some naturalm isfortune such as the death .o f a. m em ber,of the .group, or a .fam ine, or a plague, and if. the naturaLorder-is.thought-to,be sensitively responsive ,to the m oral condition ,o f society, .then .the calam ity allow s o f a num ber o f alternative interpretations. A w ide-range o f conflicts between persons and. factions... in -th e disturbed group is brought to light. I take up this point in greater detail in the follow in g chapter, in w hich I attem pt to analyse the pattern o f intrigue in M ukanza V illage. B u t I w ould like here to emphasize w hat seems to m e to be a significant difference. W h en rules are broken b y livin g -persons, judicial
128
Schism and Continuity in an A frican Society
action can fo llo w and. this, action can speed ily seal o f f conflict w ithin the orbit o f a single relationship or w ithin a small sector o f th e'social system. B ut when., a., severe.. naturaLrais&rtune precipitates crisis, practically every, .latent „source_Q£..conflict in the.system is made manifest. These m ay be disclosed in accusa tions o f w itchcraft/sorcery, or in confessions o f guilt b y those w ho feel that they have broken some crucial norm governing the intercourse o f the living w ith the living or w ith the dead. Since nO-.specific..norm has.been .patently..broken-,it_~is im possible to introduce judicial machinery.o£redress.~.--Eitherxitual o r the ascription ..o£.witchcraft/sorcery .to .one,o£its- menihers is resorted to.by the stricken, group. InJboth eYents.the.fundamental norms_and ..values o f .the .community.are-.emphasizedJn-symbol, m im e and_p.re.cept....¿kid. by.„the .p.urging,,.o£.a...witch., the group reaffirm s.its solidarity. . ..In. such situations irremediable__c.onflict has been felt .to exist between the m ajor principles b y w hich the group. .is organized., There is n oth in gjforjt but to_.lay stress on values to w hich all men subscribe,..regardless-of.their.particular loyalties and interests. It should be pointed out that w hile judicial m achinery is often deliberately- .and consciously— brought- in to . play w hen men quarrel, ritual seems, .to. emerge, as .the spontaneous response to the .moral discom fort experienced ,b y a group-disturbed by-som e natural...disaster. L aw is thus directly related to man-made breach in social, .regularity,...while. ritual_is„ohliquely..related to natural breach in it. Finally, when natural disaster.strikes^,there.is .n o.legaLrem edy for it. O ne cannot ,sue m ystical beings, and powers in .any court and obtain - compensation. Y e t . the disaster -h as-to .heum ade intelligible in terms o f a m oral .order . o£_the.._universe^^This can. be done, as I have mentioned^ b y x confession .o f one’s ow n g u ilt in.a ritual context, or b y a projection_o£-the-g.uilt-outwards onto an external enem y, a w itch or a sorcerer.1 Summary It-has been shown how iiLa.long-estabKshedwiHage,xnatrilineal affiliation constitutes the -cruciaLm eans-of.providing-village inte1 1 am grateful to Professor John Barnes, of the University o f Sydney, for making this final point.
M atrilineal Descent
129
gration_B oth dramas reveal.how pow erful .matrilineal des can. be,-andJhow they resist-disruptive forces- put in to a ctio n -by con flicts _of interest .w hich arise .withm jh e m^ of m atriliny itse lf People live together, because...they-,,are„.matrilineally rdlated, but ju st because-they are- m atrilineally related they come into conflict over officeandover the inheritance o f property.1 Since the dogm a o f kinship assemtha.t-m atrilineal-.kin,participate in. one another’s existence, and since the norm s _of.kinship„state that--m atrilineal .kin., must at all times help. one.another, open physical violence between them seldom takes, .place. . Their struggles are phrased in the idiom o f sorcery/w itchcraft and animistic beliefs.- - .Yet..in. spite-of. the..fear. arou&edJby-witchcraft and sorcery, m atrilineal kin m ay go to strenuous.lengthsJto avoid finally ostracizing one o f their num ber w h o is believed -to be a sorcerer or w itch. ConfHct is endentic m.the-SO.cial ,structure but a set o f mechanisms exists w hereby conSict-itself-is -pressed into the service o f affirm ing group unity. . In the n ext chapter I go on to show h o w the conflicts over succession to headmanship 2 inherent in the m atrilineal structure o f M ukanza V illag e broke 1 When a person dies, his or her most senior matrilineal kinsman, usually an own mother’s brother or uterine brother, becomes executor (nyamufu, literally ‘ mother o f the dead o n e '). Traditionally, the standing crops in the dead person’s gardens were consumed by the funeral gathering (rhipenji) that came nightly to mourn for a month or more. The funeral camp con sisted o f kin and neighbours o f the deceased. At the end o f the chipertji (or ckibimbt) the executor divides the dead person’s movable property, including such items as guns, tools, utensils, cash, small stock, clothes, bows, spears, beds, stools, etc., among the matrilineal kin o f the deceased. Uterine kin o f the executor’s own genealogical generation, i.e. uterine siblings, receive the largest share ; next, his o r her sisters’ children, then sisters’ daughters’ children. I f any property still remains it is allocated to more distant matri lineal kin in order o f seniority. It depends on die good will o f the executor and the agreement o f the matrilineal kin whether or not the wives and children o f a dead man receive any o f his property. The deceased’s hut is burnt down and he is buried in his clothes and blanket, while many o f his cups, plates, and other utensils are nailed on trees in the graveyard as a memento mofi. 2 Headmen in the past enjoyed a few privileges, such as gifts from their villagers o f first fruits and the first calabash brewed o f each kind o f beer. At ritual gatherings and other public occasions when beer and food were distributed headmen were entitled to the first share. But the primary value o f the headmanship o f a long-established village, such as Mukanza, lay in its prestige—it is a great achievement to build up an enduring residential unit from Ndemhu individualists.
130
Schism and Continuity in an A frican Society
out afresh in a new series o f social dramas. Throughout my second tour o f fieldw ork it became clear to m e that the village was rapidly approaching the major crisis o f fission, w hich will probably take place w hen M ukanza Kabinda dies. I then go on in subsequent chapters to discuss a num ber o f situations in w hich fission actually occurred in M ukanza and in other villages and analyse the tensions that led up to it in different villages. In .spite o f all efforts to maintain the unity of. village .members, fissiah frequently _takes_.place._^The_.circumstances.-in_which it occurs and the different form s it assum ed.w illbe described and analysed, and the relationship between, the- subsystem o f the village and..the inclusive system, form ed b y N dem bu-society w ill. be exam ined in a variety o f aspects.
CH APTER V M A T R IL IN E A L S U C C E S S IO N A N D T H E D Y N A M IC S O F V IL L A G E IN T R IG U E N the last chapter I exam ined tw o social dramas which, made visible and explicit the underlying social structure o f M ukanza V illage, and .exhibited in. action the lines o f tension and* struggle w ithin the m ajor m atrilineage. I.now . wish .to. .analyse a further set of-dramas w h ich arose out o f the same fundam ental structural situation after a passage , o f three. years in, M ukanza V illage, as that-village becam e the scene o f new conflicts, for-pow er. B ut first I m ust outline the events that preceded the next o vert rupture o f relations between the leading persons in the village, social, dram a has passed a clim ax, and the group in v o lv p ^ ^ ^ p to have-been reintegrated, there often follow s, in N d e^ h u h^ei: a period o f apparent equilibrium during w hich, in the mami interpersonal behaviour is in consonance w ith generally accepted norms o f conduct. B ut beneath the surface conflicts, o f interests go on and private intrigues provide m eans-whereby-individuals seek, to realign the social structure in pursuit o f their ,own-advan tage. Beneath the manifest pattern, o f daily, interactions, a re shaping. ,o f the social group—is taking...placet-transferences o f persons from one faction to another, loss an d .replenishment o f personnel in the follow ings o f leading..m en^so-that-in- effect a new and at first hidden set o f pow er relations gradually comes into being. There com es a tim e w hen this reaUgnment-becomes visible in.afresh, social.drama. A t the culm ination o f this drama, i f unification is -re-established,...the^gro.up_may_exhibit^a„.marked shift-in the balance o f power.-between.. variau$__components. In Jhèse intrigues., the diffèrent, principles. of-N d erabu.social organization are m anipulated b y the m ajor intriguers to. further their o w n ends. - These interim periods Qf.se.emmg.peace.hetween struggles...are really the., con tin u ation .of struggles-irt-a—different and-indirect guise....In th em .w e.fh id attem p ts Jbeing—rnade_by parties-who have secured social, gains, in preceding, social-dramas to consolidate their position,, and. .attempts b yi defeated..parties to repair their dam aged prestige, attach to. themselves- fresh
I
131
132.
Schism and C ontinuity in an African Society
supporters, and w in over to their, side, members o f the - oppos ing.- groups. A fter the reconciliation between Sandombu and the other members o f the major lineage o f M ukanza, several years elapsed during which there prevailed on the w hole an appearance o f outward harmony between M ukanza Kabinda (I,F8), Sandombu (I,G io), Sakazao (I,H9) and Kasonda (I,G i5), the leading men o f the village. It is true that, on occasion, bickering took, place between Sandombu and the others over m inor matters. One such dispute arose over the allocation o f a goat paid to Sandombu b y the husband o f his classificatory sister’s daughter as a customary gift to the local head o f the nearest maternal descent group when a w ife bears her first child. M ukanza Kabinda, as headman o f the village, said that the goat should be killed and divided im m e diately between all the village members for * there was meathunger in the village ’. B ut Sandombu pointed out that the goat was in kid and that i f they were patient they w ould have several goats instead o f on ly one to eat. M ukanza became angry, but Sandombu, claim ing that he was the senior livin g ‘ m other’s brother ’ o f the wom an and thus had the right to do as he pleased w ith the animal, sent it to a village some miles aw ay to run with the flock o f his blood-brother. V illage opinion was divided on this issue, some siding w ith M ukanza, others w ith Sandombu. D uring this period o f peace, Sandombu was careful on the w hole to avoid public dispute w ith M ukanza, and, in fact, sought to ingratiate him self w ith the other village members. From 1948 to 1951 he did not live in Mukanza V illage but in a camp o f road labourers a m ile aw ay, where he was capitao or ‘ foreman H e managed to obtain the em ploym ent by the Public W orks Departm ent o f a number o f M ukanza m en and youths, thereby putting them under an obligation to him . In his camp o f mud huts he occupied a position analogous to that o f a village headman and at w ork he enjoyed greater authority than N dem bu headmen norm ally possess. This situation was brought to an end in m id-1951 when an African R oad Inspector recommended to the D istrict Comm issioner that Sandombu should be dismissed on account o f the poor condition o f the stretch o f road for w hich he was responsible. Dismissed he was, and it is interesting to note that in this situation all the Mukanza people, including the headman, took Sandombu’s part
M atn lin eal Succession
133
in attributing his m isfortune to the repaym ent o f an old grudge by the R o ad Inspector, w h o had quarrelled w ith Sandombu when both w ere livin g in C h ie f Ikelenge’s area. Sandombu n o w decided to build near M ukanza V illage. He did not build his new house in the village itself since there was little space in the hut-circle fo r such an edifice as he intended to erect. H e selected a site about a hundred yards from the main village, next to the main Bom a—Kalene road, where he w ould be in constant contact w ith the daily stream o f people passing; along it, could offer hospitality to the wealthier and m ore im portant, and he aufa it w ith any D istrict news he m ight turn to his advantage, such as advance inform ation about paid em ploym ent. W ith his savings he em ployed an O vim bundu bricklayer to make fo r him a large three-room ed house o f K im berley brick and a carpenter to m ake him a handsome door, w indow-fram es and furniture. H e then applied his indefatigable energy to the task o f m aking large gardens o f m illet and cassava, and urged on his tw o hard-w orking w ives to do likew ise. His senior w ife at this tim e, K atiki, was the daughter o f a Lw ena w om an slave o f Governm ent C h ie f N yakaseya, and had a w ell-grow n fam ily, one o f w hom , a son b y a form er husband, built a pole-andmud hut beside Sandombu’s. Sandombu’s.im m ediate aim was to build up b y any means a follow ing, to settle, w ith him , and whenJMukanza died to make, a renewed bid fo r the headrnanship o f the- village. It is at this point that w e m ust pose the.question: w h y did Sandombu still desire to becom e a traditional village headman w hen through his participation in the^modem casheconom y he m ight have expected to prosper as a petty .trader-and cash-crop farm er ? A ll over M w inilunga, and especially in the northern pedicle where Sandombu had spent a large part o f his w orking life, the traditional village was givin g grou n d before. jfo e sm a ll ^farm * asjthfc typical, form o f settlem ent..... JEarmJheads. w ere disencum bering themselves o f m any o f the.obligations. o fk in sh ip , .and retaining for their o w n use and for. the. use o fth eir. elem entary families m oney they earned as .wages and jb y the_ sale_ of_.cashcrops o r surplus subsistence-crops. B.utJS.a.ndo.mbu .still hankered after, the h ead m an sh ip o f a _tm ditionaL village.— W hy.?.-. The answer to .this question lies, ! think, in the -fact^tliat-the-kind o f response made to socio-econom ic. ..change...im..this_xegion_was 1»
134-
Schism and C ontinuity in an A frican Society
ia£ge.ly_„.a.. function .q£, relative . age. Q ld men. lik eM u k an za remained obdurately conservative,.and„deplored.,the.new w ays, although they w ere indulgent towards the young people, who practised them . Y ou n g m en lik e . Sakazao's uterine nephew, Pearson (IJ5), capitao o f a European-owned, store at the Boma, and.JVianyosa’s oldest son Daudson (fH y) w ho had. w orked in Chfogola, had accepted the new order o f things., andJWOOS. smart European clothes, ow ned bicycles and. gram ophones,.„played guitars, used Copperbelt slang, and attended-traditional rituals only to jo in rings o f young, people, w ho danced the latest,..dances im ported from the Rhodesian or Congo urban areas an.d sang.the latest local or urban ‘ song-hits \ Re.tween .these_.tWQ. extremes, Sandombu, ..and .to...a. lesser degree.Kasonda,.„though., th ey had grown, up .in .the .era....o f. m oney-earning, .. still_ belonged ..£o a generation_.which saw. success in life as.measured_by.the,numher o f followers , a man could acquire, and. not b y the irisignia.of .conspicuous w ealth that could, be. purchased—b y „m oney._.True, Sandombu, and also Kasonda, had large ..houses. well..equipped with' furniture. T hey had mosquito-nets and o il lamps. . B ut these._signs o f wealth .were rather indices o f success in the traditional order than signs o f an altered w a y o f life, in vo lvin g the acceptance o f entirely new modes o f . behaviour, and .o£.a .new scale o f values. R oth these m en continued- to -w ork..in their gardens, to gossip and discuss cases. in .the yillage, chota, tgjp&rticipate in ritual as cult-m em bers and patients, to exercise their traditional rights and fu lfil obligations as kin, and to interact w ith the older generation in terms o f traditional .norms. T hey were not ashamed, as m any o f the younger people w ere, o f the ancient w ay o f life ; they lam ented that, so much, o f it. had already passed aw ay, and m ore was passing beneath their v e ry eyes._ B ut th ey felt that the royal road to eminence .within. .theJm lage w ay o f life now la y through the acquisition o f cash._ Possession o f cash gave them large houses,-bride=wealth for-several-w ives w ho m ight g ive them children, and en.able .them .to -offer-hospitality, and the means o f retaining their chddren.and..givmg. .thes€ea g o o d education. T h ey wanted m oney to-better- their-position w ithin the .traditional system, not as a means .o f loosening-thek-ties w ith it. For instance, for these men o f m iddle.years,_a.gunwas better than a bicycle, and to obtain, .the wher.ewithal..to_buy one they w ould save up fo r m any years, keeping...the money_in_a..bag or
M atr ¡lineal Succession
135
boxJburied secrectly in the ground, where.even.their.3sdv.es could, not-find it. I woidd_like_ta.-emphasi2e a$ this_p.oint_ the..main^diiierences between, tw o m odes o f incorporation in ^the.. .m odern. cash economy. O n .the. one. hand, a m an. .cm ..acquire...wealth hy w o rk in g in th e .W hite economy_as^a.wage labourer,-eitherJocally or by- m igrating to the urban centres, on the R hodesian line^ofrail. O n the other hand, a m an can. obtain m oney Jocally by selling surplus subsistence crops,, b y growing_.cash_.cro.ps,_such as rice onground-nuts,, for. sale..to. European or African, traders,.and by setting, up .in business as a petty trader, tailox. or. ‘ tearrxoom * proprietor_-(i,.e. b y .selling.ibod and_drink..to..travellers .along a labour, route). It seems often, ...even today,, to b e t h e a i m o f returned, labour-m igrants „or ...ofLpaid ^workers. in .local .employ ment to obtain influence, and. subsequently. .office,. in .traditional villages. M any o f them see the village as their, ultimate, hom e, andjregard their w age-labour as. a_means._ofaeqmring.the.. wealth that w ill give them prestige in the. village.sphere.._..Bju£.i£jhey havejsubstantial savings, it m ay happen thanthey wish^to. invest these in. capital equipm ent, such, as a sewing-m achine,--with the aid o f w hich they m ake up store cloths in togarm ents for pay ment, .a bicycle, on w hich they can travel , to buy. goods, cheap and sell them elsewhere, at a profit,_or. a, storehouse-or.tea-r00m madefirOm sun-dried bricks, to serve.as.business.premises. Once they have taken this step and have invested m oney to .m akem ore m oney, th ey find themselves increasingly embarrassed b y the demands o f their k iir fojipresents in cash and kind. If they w ish to become..-p.etty:-capitalists,.- they -must-separate--themselv:es~firom the.village-Sphereuandthe v illa g e .w a y o f life— JXhatris_.why it is usuak to fin d traders and tailors hving w ith_theh iam ilies, in different vicinages, and even.chiefdom s, from the. viUages.^£ their close m atrilineal kin. . T o 'a lesser extent, the same ..set,, o f con ditions holds, good. for. .pettyxom m odity. c.ultiva.tors^.)dxeincipient * kulaks-, w h o. grow crops specifically fbr sale— If.the_.making o f m oney tends to supplant, as a._major. aim. in life.._the._acqirisition o f a-follow in g, people ,try .to., accumulate..money,...t.Q.„he, turned into capital and the visible signs o f a. higher .status_in.. .terms o f European values. A large fo llow in g then tends,.to_.become an embarrassment rather than an asset. In essence, the..wage-eamer, w ho acquires w ealth at the expense o f the W hites to .strengthen his
136
Schism and Continuity in an African Society
position in the traditional village, must be „distinguished, fmm the local ,m oney-m aker w h o grow s rich at the expense o f his feUowrNdem bu. B l4t.-$andomt>u, and f o r m a n y years Kasonda sought to usejheir..wages gained in w orking.for Europeans.. (Kasonda since his__return from the Im e-of-raii had. w orked as-adomestic-servant for a succession o£D istrict O ificers).iii acqmringjHends^and sup porters am ong their kin and neighbours to prepare the w ay for succeeding to headmanship. Later on, towards jthe_.endLoF my second period o f fieldw ork, both menjwere_cQming_to_realize that the old order ..was doom ed,, and .that-tO—b.ecom.e_eminent they/ m ust. ...commit, themselves w hole-heartedly—to .the cash econom y.... Hut. at the_time .of SociaLDxama^III„they- .were still caught up in the contradiction between.parrid.patianin.am odem w ay o f getting a living. and_aUegiance_to_arL.jo.hsoles.ceJitjs.et o f values. The cash econom y. to..which .theyL..w.ere..c.ommitted was breaking down. the structure .o f the. village_^-aiid_ye.t.itjwas their am bition to obtain authority oyer a-disintegratmg_s.acial unit o f this kind. Sandombu,_then,.-used_mjon.ey.-to~increaseJhis~influence-wtthin a„system w h ich , was.gradually-breakmg-do-wii-as-the-result o f the introduction o f m oney. .He..was..not satisfied rto..remain_alone in his h ig house, or even to attract Qthers,to„fosL.settlement,_ He wanted eventuaUy to becom e headman o f M ukanza and to spend the evening o f his life as a Tespected eldet.o£.the_n.eigh.bourhood com m unity, supported.b y emQluments_fr.Qm the_salaried,_young merL Jto .whose educatioix he had con.tributed- w lien .they were boys. M eanw hile bis prospective rivals for the office had been by no means inactive. A member o f malabu lineage, w ho had spent m any years on the Copperbelt, Line (I,H i 7), returned in 1951 w ith ms w ives. Sandombu prom ptly found him a jo b as a road labourer, but Line soon quarrelled w ith him . Sakazao (LH9) then persuaded Line to bring his uterine nephew Aram (IJ7) w ith him and found a * farm * ju st behind Sakazao’s own huts. N yam uw ang’a (LG 17), the suspected w itch o f Social Dram a IV , also had a hut built for her at Line’s * form ’. Sakazao, it w ill be recalled, was the brother-in-law and classificatory * grandson ” o f M ukanza Kabinda. H e was rather older than Sandombu. Sakazao was much poorer both in cash and in
PLATE HI M u k a n z a V il l a g e
in
S ep t e m b e r 1951
The chota (men’s forum) is in the centre of the village. Kasonda’s Kimberleybrick house is under construction (see p. 138).
!«■■*• iiwiiuttrii |—"*T r~ *
Tm in-*‘*‘~nriiir
O
137
o
kind than Sandom bu and Kasonda. His health was uncertain. In his you n g m anhood he had been considered som ething o f a * m ischief-m aker * (chipapokt). His favourite practice was to travel to a part o f the D istrict w here he was little know n, offer to sell a gun to a villager at a very cheap price, receive partpaym ent fo r it in advance, and set o ff hom e, prom ising that he w ould return soon w ith the gun. H e never returned. H e had been cured o f this practice b y his village kin w h o g rew tired o f paying damages to Sakazao’s victim s, m ost o f w hom trackedhim dow n in the end. His personality was a curious m ixture o f naïveté and shrewdness, but it was universally regarded as lovable. Sakazao was not respected as M ukanza Kabinda was respected, but he was loved. W hen his youthful exploits were recounted in the chota, people w ould grin and shake their heads, clucking their tongues and repeating his name. W h en he made his pretendedly grum py sallies in response to teasing everyone w ould be delighted. It was thought b y m ost m em bers o f both village lineages that Sakazao w ould make an excellent successor to M ukanza, since he was unaggressive, hated no one, and, though som ew hat inactive, was hospitable to strangers. M ost im portant o f all, no one had ever accused him o f practising sorcery. B u t i f he became headman it was generally recognized that the real m anagem ent o f affairs w ould devolve upon a younger and m ore capable man. Sakazao w ould m ake an admirable figurehead, but a poor representative o f the interests, jo in t and several, o f his villagers. Y e t in his o w n w ay, and during the struggle for pow er described in this chapter, Sakazao was astute. Sakazao played on the suspicions o f bus classificatory * father \ N tololu (II,D z) o f Shika V illage, that his relatives w ere bewitch ing him , in order to get N tololu to com e and live uxorilocally in Line's farm . N tololu, perhaps influenced b y Sakazao, built a hut there. N tololu ’s w ife Kalusa (I,G i9) was the sister o f N yam uw ang’a (I,G i7 ), and both these w om en, although about the same age as Sakazao (LH9), w ere his classificatory mothers o f m alabu lineage. N yam uw ang’a had a hut built fo r her beside Kalusa’s. Just previously, Kalusa’s and N tololu ’s daughter Ikubi (I,H i4) had undergone her puberty ritual at M ukanza Village. N tololu and some boys from M ukanza built a hut for Ikubi at Line’s farm , w here she was to aw ait her husband’s return from the Belgian C on go. Also at Line’s farm w ere a wom an o f
o
M atrilin eal Succession
©•
e
©
© © © 0
©
© © ©
© © ©
© ©
©
© © © €l
©
© © ©
P '
p
b O
O &
b o
<•4 D 0;
138
Schism and C ontinuity in an African Society
m a l a b x j lineage and her husband, a road labourer in Sandombu’s gang. Thus, at the end o f the dry season in September 1951, it was already possible to observe in the spatial structure o f Mukanza Village visible signs o f gradually approaching conflict for author ity. Behind Sakazao’s hut w ere five huts belonging to members o f malabu matrilineage, form ing, as it were, a nucleus o f social pow er. A hundred yards from the village behind the circle occupied by his generation mates o f nyachintang ’ a lineage was Sandombu’s great house, occupied by himself, tw o w ives and a grandson o f his senior w ife, and her son’s small hut. In Mukanza Village itself Kasonda had built a large Kim berley-brick house and a smaller one for his recently married third w ife.
O o
SOCIAL D RAM A III
o
Kasonda is Accused of Bewitching his Uncle Kanyombu (my own observations)
o
:o / o o
() G o
O C; o o Q o
o
Before proceeding any further with this account I must mention my own impact and influence on the life o f the village. A t die beginning o f 1951 1employed Kasonda as a cook and general henchman. After a time I made a permanent camp in Chief Ikelenge’s area, returning to Mukanza Village on intermittent visits to watch the progress o f events. In June o f that year Kanyombu (I,Fp), Mukanza’s younger brother and a leading doctor in several cults, died after a protracted illness. Kasonda told me that the cause o f his death was probably Sandombu’s sorcery. He had received a letter from Sakazao asking him to return for the funeral ritual. Since I was about to pay a protracted visit to some bush villages at that time I was unable to release Kasonda whose services I needed on tour. In July Kasonda received a verbal message that his classificatory mother’s brother, Kajata (I,Fio), an extremely old man o f malabu lineage who had recently returned to his matrilineal kin after about fifty years’ absence, had also died; and the message urged him to come to the funeral ritual. It happened that I had planned to pass by Mukanza Village on m y way to the capital village o f a Kosa chief and I took Kasonda with me. When we reached Mukanza we found the entire village in an uproar. The day following Kajata’s death Mukanza himself had been taken seriously ill. Sakazao and Ikubi (I,Hi4), his young female relative in Line’s farm, had been ill for about a week. A social drama was, in fact, in progress. This time the breach in regular relations between villagers had not
t____________ _ — J
140
Schism and C ontinuity in an A frican Society
been caused by a dispute but by the natural calamity o f what I took to be an epidemic o f malaria. For among Ndembu -uiducky—events occurring in the natural order .are thought... to..originate. in_J>reaches of-tbe moral order. More explicitly, in -the. co n textofvillage life, these misfortunes onginate either-in_..tlie .maligmty o f kin. w ith evil powers against other kin or in .the, p_uiutive ,actioii_o£ancestorL.spirits against their living kin who have fbrgotten..thermor_transgressedJdnship "norms. The epidemic just mentioned was thought to have been caused by one or the other o f these agencies. I arrived at the phase o f crisis while the people were still arguing about the origin o f their troubles and about the best means to banish them. Kasonda was received w ith hostile looks, and it soon became clear that he was regarded as the sorcerer who had bewitched Kanyombu and Kajata, and who was now threatening Mukanza’s life. Presently, Nyamukola (I,Hxo), Mukanza’s senior wife, and Nyakalusa, Sakazao’s junior wife and slave, began openly to accuse Kasonda o f having gone to Kalene Mission, where the missionaries have created a sanctuary for people banished from their homes in Angola as witches and sorcerers, in order to procure medicine to kill his three senior relatives o f the first ascendant generation and his classificatory sister’s son Sakazao (I,H9).X Manyosa (I,Gi3), Kasonda*s uterine sister, and Chawutong’i (I,G iz), his matrilineal parallel-cousin, strenuously defended him, as did his senior wife, Mangaleshi (II,Ez), but they were in the minority. Kasonda had already begun, with the help o f his cross-cousin Gideon (II,El), Mukanza’s son by a deceased wire, to build his Kimberley-brick house, and Nyamukola shouted that this was clear proof that he was planning to succeed Mukanza. W e went in to see Mukanza 12 who was lying on his bed, obviously in a condition o f great fear and distress. Mukanza’s two wives followed. An extraordinary scene took place beside what looked like becoming a death-bed. Kasonda became very angry and said: * Alas, Africans are very, very stupid people. Kajata was an old man. God (Nzambi) has taken him. But i f God has taken Kanyombu, it is obvious that Kanyombu’s spirit {mukiski) is now troubling the village. Women are m ighty fools and w ill tattle about anything. How could I have gone to Kalene Hill when I was working for Mr. Turner all 1 Sakazao was grouped with the senior in the linking o f alternate generations. 2 Under normal circumstances Kasonda would not have entered Mukanza’s and Nyamukola’s hut; since Mukanza was his mother s brother and Nyamu kola rated as his mother-in-law, to be avoided, since she was the mother o f his divorced wife, Koniya (I,Jr). B u t as m y servant, the dominant role he chose to play, he ignored these customary rules.
M atrilin eal Succession
141
the tíme ? W hy should I kill Kajata ? For meat ? I had plenty o f beef, goat’s flesh and duiker meat at Mr. Turner’s place. I do not eat human beings even i f other people in this village do.’ A t this point I intervened and offered to take Mukanza without delay to Kalene Hospital. The old man accepted gratefully, not, I suspect, because he had much faith in European medicine, but because he longed w ith all his heart to get away from the situation in which sorcery was thought to be active. Once away from his relatives he would be safer. Others demurred and we all went into the chota to discuss the matter fully. Sandombu was away at work at the time. Mukanza’s two adult sons, Line (LH iy), Kasonda’s tw o grown nephews, and the headman o f a nearby village whose father had come from Mukanza Village, were present. The women sat outside the chota and took a lively share in the discussion. But first I dosed the sick people with quinine and offered to take them all to hospital. Manyosa (I,Gl3) said that her mother’* brother, Kanyombu, had in his lifetime drunk mwiyanawu medicine, (that is, medicine which would enable him to come back as a ghost (musalu) and kill members o f the lineage o f the sorcerer who had bewitched him). She implied by this that since so many members o f malabu lineage were sick the sorcerer or witch had belonged to that lineage. Nyamukola, w ife o f Mukanza, had now calmed down somewhat with the prospect o f accompanying her husband to Kalene Hospital before her, and interjected that it might well be that Kanyombu was haunting the village, not because o f revenge medicine, but because he had given Mukanza detailed instructions about how his funeral stretcher should be made and medicated, and Mukanza had omitted to carry them out. Perhaps he was angry at this neglect. If Mukanza were taken away from the village and given European medicine he might recover. Yes, he should go. Sakazao’s wife then interposed that Sakazao had been dreaming about Kanyombu all the time he had been ill. Yes, Kanyombu*s ghost was to blame for the trouble, not Kasonda. Kasonda spoke up at this point, saying that M r. Turnea: had once given Mukanza some money to pay Fisher’s lorry driver to transport Kanyombu to Kalene Hospital when he first became ill, but that Mukanza had divided the money among his wives and children and not sent Kanyombu at all. This was another reason w hy Kanyombu was angry. It was not Kanyombu’s ghost (musalu)* his evil power, that was aedve ; Kanyombu, in spite o f what Manyosa had said, had never drunk mwiyanawu for he had never observed the taboos o f mwiyanawu, such as eating only one side o f an animal or chicken.
o o o o o. o
142
Schism and Continuity in an African Society
O
But Kanyombu's spirit (mukishi)yhis moral power, was angry because his relatives had scorned his wishes and neglected to look after him properly. Kanyombu had died because he had not been taken to hospital. If, on the other hand, Mr. Turner took Mukanza to hospital he would recover. As Kasonda made these points, most o f the men clapped their assent, and cried ‘ Eyo-o \ ‘ that's it The upshot was that it was agreed that Mukanza should go to Kalene Hospital, but Sakazao and Ikubi (I,Hi4) refused with the excuse that m y medicine would soon make them recover. Mukanza and Sakazao both recovered, but the young woman Ikubi died.
C)
Analysis
o
c O
o
O
i
G
;
O
!
o
I
e ! o !| o o O
O
G O G G
l)
G
;
In this situation misfortune was interpreted .hy._variaus_p.eople in .terms .ofla...bxeach o f several-different, moral, norms,„and thus it hrQUght..tCL..the_.s.urfaee..a~number.x)£..difierent latent conflicts within the village. This tim e the lines QCconillct.were not clearcut.. I f there, is .a.real .dispu.te...b.etween„Hving_vjhagejnianbers, created b ya breach ..of rules, j.udiciaL.action_can_follo-w.— W hen serous illness occurs it is believed ,to„ be. .caused_by_one._of a number o f mystical beings o r .forces.. The_operatioruo£aIL.those beings and forces is associated w ith disturbances...o£- the_m oral order. Therefore__every^laio_wi]LitenxQ£_the.pa£ient’s.b.ehayiour in- the recent past is brought under review ,.jbi....o.rder_.ta..survey possible sources o f disturbance, ...and., to._locate_that-which is reiponsibie.^, THs..is._.4on^_in._aE:j!3.$.^_j^hin_myL.,^pjerience even_before.resort is made_tQ^
M atrïlineal Succession
143
he-the.p.unitive action,.o f guardians o fkin sh ip m orality ; in. the second ewent». the-m alevole3PLce..Q£„witches,._sorcerers _or .ghQs.ts„has_heen_at w o rk. T h ere-atecertain . exceptions ..ta this rule__It is . thought that sometimes, in. extreme_cases,._theL.killing_by-sorcery. o£.a_notorious sorcerer, or w itch is justifiable.....Butnwhen a^sorcerer-is-believed toJh-ave beenJkiHed_bysorcery, ^-.cetiain^dium.is..&till-.£eltJto be attaghftil m *b<- «layer, ?1thougfr it* *h« case iS-Seldom invoked. and his.identification. isdeft to„gossip.-and xumour^ . O n the_otHerJiand.,.certain..mamfestations..o£ancestor.spirits.express unjustifiable m alevolence. ..to.war.ds_ th eir- living.....kirn. ....When a spirit is th ough troh avecausecL unjustifiedseriousillnessor grave misfortune,, it is not execrated but placated,-althougbuelements o fe x o re ism m a y h e.foun d .in -the ritual. W hether misfortune. _or illness .is..ascribed.. to...the..unrighteouS“ ness..of the. victim^ or to the..unrighte.ousness-o£..a-Corpareal or spiritual enem y, depends on a num ber..o£factors :_. the-degree or kind, o f m isfortim e_;the.sex.age or status of the victim in varying social-situations. ;. and. the_-Current_leveJ_Q£..morale._of_his. local g ro u p ro fld n .... h i m ost cases, hath theories in vari.QUsforrns are put_forward. Som etim es. sufficient consensus.. is.ohtrinecLahout the particular cause to avoid recourse ,to -a.-d ivin er,-w h ich is costly... and today risky as w ell. For instance, w hen a certain pregnant w om an becam e seriously ill, it was agreed by her kin and husband that the Wubwangu or tw in ritual should be per form ed for her, since both her m other and m other’s m other had been mothers o f twins, and it was considered likely that her grand m other’s spirit was afflicting her reproductive processes. W hen the w om an adm itted that she had forgotten to make an offering o f first fruits to her grandm other, her kin regarded it as estab lished that the latter’s spirit was offended and had afflicted her in this w ay. Hut it is m ore...us.ual_for-two-m ajor ...theories to develop and the diviner in effect has.to choose.betw een them. This tendency for tw o .theories to .develop sometimes .produces unexpected-consequences. W hen I was in M w inilim ga, a song was current throughout C h ie f M w ininyilam ba’s area, the gist o f w h ich was that w hen a headman died in one village, the senior m en o f tw o village matrilineages each w ent to a diffèrent diviner and each obtained from his ow n diviner the name o f his rival as the sorcerer w h o had bew itched the headman. T he song
144
Schism and Continuity in an A frican Society
related how they returned to the village at the same tim e and each prom ptly accused the other o f the headman’s death. A dis pute arose and each led his kin group aw ay to form a new village, so that the name o f the old village * died \ There is a droll irony in the flat account o f the events w hich is irresistible. N dem bu, in ia ct, often express scepucism.abnut.divinationsLmadeL.fbii-others, but__se]dom abQut_thos_e _saught._o_ut_hy. themseives. The situation in M ukanza V illage brought out into the open not o n ly an im portant cleavage in the social structure, but also interpersonal grudges. W hen N yam ukola accused Kasonda o f bewitching her husband she was not o n ly attacking him out o f jealousy because his im proved financial position as m y cook gave him additional qualifications for headmanship, but also because Kasonda had once been the husband o f her favourite daughter K oniya and had treated her so badly that she had left him after a m onth o f married life. A nd when M anyosa (I,G i3) said that Kanyom bu’s ghost was attacking members o f m alabu lineage because K anyom bu had possessed vengeance m edicine, she was not m erely defending her brother from N yam ukola but also revealing the hidden conflict that existed between nyachin tang ’ a (I,D i ) and malabu (I,D z ) lineages ; fo r i f Kanyom bu’s vengeance medicine w ere w orking, this could only mean that someone o f that lineage had killed K anyom bu b y m ystical means. Later in fact the m yth became established am ong members o f nyachintang ’ a lineage that N yatioli (I,H n ), sister o f N yam u kola, and divorced w ife o f Kanyom bu, had bewitched the old man. T h ey had lived a married life o f continual domestic bickering and N yatioli had developed a grudge against her form er husband. W om en are believed to possess familiars o f a kind different from , and m ore dangerous than, those ow ned by men. T h ey are called tuyebela, tushipa or andumba, and have the variable form s o f little men w ith reversed feet, hyenas, jackals, owls or small rodents. T hey are believed to be inherited m atrilinealiy and to make demands on their owners to kill junior relatives. T h ey m ay be refused three times but after that, how ever hard the ow ner (nkaka) m ay plead, they k ill their chosen victim . W hen the owner o f tuyebela dies, they seek out one o f her close kin, it m ay be a daughter or uterine sister, and attach themselves to her whether she desires them or not. Their victim s are the husbands or jun ior matrilineal kin o f their owners.
M atrilineal Succession
145
[f a husband is killed his ghost becomes w hat is know n as a kahwehwi, and leads the band o f tuyebela. He is m ore pow erful than they, and m ay on ly be gainsaid once. Tuyebela are excep tionally active i f their ow ner has a grudge (chitela) against some one, w hen they are liable to take instant action against him w ith out asking their ow ner. N yatioli had been suspected for some time o f having tuyebela, as indeed had her classincatory m other, the notorious w itch N yam uw ang’a (I,G i7), also o f malabu lineage, o f w hom m ore anon. N o doubt, one o f N yam ukola’s (I,H io) reasons for pinning the accusation o f w itchcraft on Kasonda w as her fear that she herself, com ing from a lineage the w om en o f w hich w ere suspected o f w itchcraft, m ight be accused i f M ukanza w ere to die. His niece M anyosa’s reference to K anyom bu’s vengeance m edicine was a fairly direct hint to that effect. Kasonda’s later handling o f the situation, after his first outburst o f anger, was extrem ely diplom atic. First o f all he cleared him self b y attributing the sickness in the village to K anyom bu’s spirit. N ext, he cleared K anyom bu from the im putation that he had used vengeance m edicine, a practice bordering on sor cery. I f this argum ent w ere pressed too hard he realized that it w ould be tantam ount to a public recognition o f the cleavage existing betw een the tw o village m inor m atrilineages. B u t i f he could gain general acceptance o f the vie w that K anyom bu’s m oral spirit and not his vengeful ghost was troubling the village, he w ould have produced a form ula that w ould unite the village again. A ll members, including Sakazao, had erred in neglecting the interests and ignoring the last wishes o f the old m an. That was w h y he had afflicted so m any people w ith illness, as a rebuke. Furtherm ore, since I had given them m oney to spend on Kan yom bu, hut M ukanza had spent it on his ow n fam ily, Kasonda subtly im plied that M ukanza was under an obligation to make restitution to m e, and through m e, to Kasonda him self, w h o had urged on them that K anyom bu be sent to hospital before he died. This M ukanza could do in tw o w ays, first b y absolving Kasonda from blam e, and secondly b y goin g to Kalene Hospital him self for treatm ent. I f M ukanza did this he w ould show publicly that he accepted Kasonda’s point o f v ie w and did not believe that he was bew itching him . In other w ords, Kasonda com pletely reversed the case against him b y posing as the w ould-be
•0
o o o
o o
\.•
o o V.
O C
o o o
0
o c
o c o 0 o o
Q 0
o O '
146
Schism and Continuity in an A frican Society
benefactor o f K anyom bu and by im plying that M ukanza and the other elders o f the village had w ronged the old man. He omitted to m ention the death o f the other old man, Kajata (I,F io ; see p. 138), for ghosts (nyisalu) are believed to have pow er to kill, though spirits (akishi) are rarely thought to do so. Later, he told m e that it had been accepted, in village gossip, that N yam uw an ga (I,G iy) had killed Kajata and I have no doubt that Kasonda did not strive officiously to kill this im putation. Later, N yam uw ang’a was accused publicly by Sakazao and N tololu (II,D2) o f having killed Ikubi (I,H i4), so that it became natural to believe that she had killed Kajata as w ell. Kasonda’s m ajor political triumph was to have created a general opinion that everyone in the village had wronged Kanyom bu and that the main burden o f guilt was M ukanza’s. H e had kept the tw o lineages together, and strengthened the view that he and not M ukanza was the real peacemaker, the most im portant o f an N dem bu headman's roles. B ut w h y was Sandombu at no tim e accused o f sorcery in this situation, except privately, to me, by Kasonda w h o was intensely jealous o f him ? As always, where human beings are concerned, there is no simple and all-sufficient answer. O ne reason m ay have been that for several years now Sandombu had lived quietly in his Public W orks Departm ent camp and had secured w o rk for village members, thus enabling them to pay their tax and clothe their families and themselves w ithout having to leave their homes. T he scars o f the old conflict between Sandombu and the rest had healed. B ut a m ore cogent reason, I think, was the sudden increm ent o f members o f malabu lineage to the village population, creating a num erical imbalance in favour o f malabu people over nyach intang ’ a people, whose lineage had always held the leadership. This led to tension between the tw o groups, tension w hich was openly expressed in M anyosa’s reference to Kanyom bu’s vengeance medicine, w hich form ed a com ponent o f N yam ukola’s accusation against Kasonda, and w hich later gave rise to the n yach intang ’ a m yth that N yatioli (I,H n ), a malabu wom an, had killed Kanyom bu b y w itchcraft. The fact that M ukanza him self was so ready to lend an ear to Kasonda’s arguments, w hich im plied that nyachintang ’ a people should perform a ritual to placate Kanyom bu’s spirit and thus consolidate the lineage, suggests that he him self was reluctant to believe
M atrilineal Succession
147
that Kasonda, his uterine nephew, o f nyachintang ’a lineage, was the sorcerer. Sakazao, too, M ukanza’s grandson, brotherin-law and friend, was in agreement w ith Kasonda*s view , for he also was partially responsible for not carrying ou t Kanyom bu’-s last wishes, and the b elief that Kanyom bu’s spirit was punishing the whole village probably seemed to him a means o f reuniting a settlement o f w h ich he hoped at some tim e in the future to be headman. Kasonda’s remark that ‘ w om en are m ighty fools * refers to the fact that N yam ukola, Nyakalusa and M anyosa attributed the com m on m isfortune to personal animosities ; i f their view s had been accepted, the village w ould alm ost certainly have split into its com ponent lineages. Another im portant factor w as m y role in..the..„situation_as a stranger....Q nceM ukanzajw a^ in.m y.vanettehe,w as„m y.responsibility a n d n o lo n ger th a to f thevillagejm em hers— Frankenberg 1 has_poin.ted.out, in an,analy$is o_£„a .Yillage_comimniity„in N orth W ales, how the responsibility.fox.makingjdecisions, ahout-matters on..which_ the villagejm em bers. interlinked b y lon g propinquity. kinship, re lig io n ,.. etc.,...were...divided^am ong themselves^. was often thrust, upon. strangers-or. village.m em bers-w ho. w ere struc turally outside., the. particular ..situation .of controversy ,_Sirnilarly Jn_this...Mdemb.u_viUage..o£.Mukanza, j£._the..hfi.a.dm.an had died-arK alene .Hospital, according.to N dem bunotions-his death w ould have lain at m y door, -though-.as I.. w as ^JEum.pean no action x o u ld . he., taken.against m e. _Sjnce I had taken the .first step in asking M ukanza.to com e ..with me.,._ICas.Qnda_conld_.not be held responsible for his fate. As. it happened_Mukanza made a speedy recovery and bo th ..Kasonda.-.an.d_L..in_..cons.e
14S
Schism and C ontinuity in an A frican Society
village. W hen meat was divided M ukanza gave a large share o f the carcase to Sakazao w ho took it to his ow n hut and sub divided it am ong his lineage kin. O n marriage and divorce malabu and n yach intang ’a people dealt w ith the bride-wealth matters o f their respective kin separately. W ith the progress o f structural tame the village cell was developing tw o independent nuclei. T he next social drama is concerned w ith divisions w ithin malabu lineage itself and shows h ow such divisions w ere ex ploited by Sandom bu to build up his ow n follow ing. SO CIAL DRAM A IV The Expulsion of the Witch Myamuwang a (my own observations) Shordy after the above events, the young bride Ikubi (I,Hi4) died, I believe o f malaria, in Line's farm. Ntololu (II,Dz), her father, prompdy accused Nyamuwang’a (I,Gi7), her classificatory mother, o f having killed her by means o f witchcraft-familiars (tuyehela). The usual post hoc ergo propter hoc arguments were invoked. It was said that Nyamuwang’a had asked Ikubi for some meat that she had cooked shortly before her illness. Ikubi had said that it was for her parents and that she had not enough to give away. Nyamuwang’a had become very angry with her and had threatened her in a roundabout way. Nyamuwang’a defended herself by saying that Ikubi’s death was probably due to Kanyombu’s vengeance medicine, not an argument likely to commend itself to Sakazao (I,H9) and his lineage kin, since it implied that Kanyombu’s killer had come from their ranks. Gideon (II,Ei), son o f Mukanza by a dead wife from Shika Village,1 and uterine sister’s son o f Ntololu, father o f Ikubi, then lost his temper and started to beat up Nyamuwang’a. Sakazao stopped him, but told Nyamuwang’a to leave the village, and stay with her sister in Chimbila Village in Chief Ikelenge’s area ; not because she was a witch but because she was a trouble-maker. Line (I,Hi7) endorsed this, saying that he would have no witches in his farm. A ll formed a united front against the old woman. She collected her belongings and fled to Chimbila Village. Later Sandombu said she had been accused without proof and offered her a hut site in his farm. Nyamuwang’a, who had met her 1 See Appendix II for skeleton genealogy o f Shika Village lineage.
O © © O © © O ® © O D O D O 9 O O -3 O ® '3 Q
o n.Dj. NYATUNGEJl
o "'1 D ;'y
^ O ^ -T\ O ^ O ^
150
Schism and Continuity in an A frican Society
daughter Ikayi (I,Hi2) in Chimbila, came back with Ikayi in 1952 and Sandombu built them a small hut beside bis house. Nyam uwangVs hut in Line’s farm had been burnt and pulled down while she was in Chimbila. Ntololu returned, with his wife Kalusa (I,Gi9) and their rexnaining children, to his own village o f Shika, saying that there was less witchcraft there than in Mukanza. Analysis This was not the first tim e b y any means that N yam uw ang’a had been accused o f w itchcraft, hi the summer o f 1950 she had been accused b y M ukanza V illage o f killin g her uterine sister, Shim ili (I,G i8), at Chim bila V illage, although Chim bila people blamed Shim ili’s other sister N yam alita (I,G2o) for her death. It was also said that N yam uw ang’a Bad killed her ow n children by means o f tuyebela. Kasonda denied this, saying that N yam u w ang’a first acquiredher tuyebela from her m other-in-law , a slave o f Shika V illage. W hen I put it to Kasonda that the generally accepted belief was that tuyebela w ere m atrilineally inherited, he adm itted that that was usually the case, but said it had not hap pened thus w ith N yam uw ang’a. This inconsistency is under standable in terms o f the social structure. Kasonda, for instance, had no hesitation in telling me privately at other times that w om en o f maxabu lineage possessed inherited tuyebela, when conflict arose between the tw o lineages w ithin M ukanza Village. B ut since Kasonda had acted as representative o f M ukanza V illage in the discussion o f death-paym ent (tnpepi) w hich fol low ed Shim ili’s death at Chim bila V illage, he had obviously tried to whitewash his ow n village m atrilineal kin in a dispute w ith members o f another village, by claim ing that N yam uw ang’a tuyebela, i f she had any, w ere acquired, not inherited. H e had either com e to believe this him self o r else wished to clear the name o f M ukanza V illage in m y eyes. H ow , then, had N yam uw ang’a com e to acquire her unenviable reputation ? In the first place, she was o n ly h a lf Ndem bu, or rather h a lf K aw iku, since her father had been a Lw ena ; and she had spent m any years in A ngola at her father’s village. Persons ofLm ixed tribal origin, or w ho have been reared, am on goth er trihes>arefrequeatly.regar
M atrilineal Succession
151
T h e y d o n o t * quite^b.elotig ’ to^tKe locaL society^ and -a& such make useful scapegoats for m isfortune. Jbx th e secQnd _place_she, like/Sandom buT.had been a prodigiously hard w orker in her gardens, knew where. to_plant.her crops, so that.she-would. obtain a higher yield than other people, jand often-w entjon,w orking in the h eat o f the day. w hen others had retired—to_gossip..iti. the kitchens. ,.And so. the meyitahle.questiQn w as.posed...fooiruyhat som ebsdid she .draw h erxn ergy. a n d w h y w ereh en xrops-better than, anyone else’s ? O bviously ..she.had supem aturaLpowers w h iclrgave her outstanding stxengthLandluck^.. .^ h en .sh e. was a young w om an she also had the., reputation.jo£-being,.something o f a_joymphomanaac. ...Mdembu J n .conversation strongly, condem njsxcessive sexual desire and associate, it .with._prop.ensities to sorcery and w itchcraft...L ike .w itchcraft,-desitc-^tuvumbi) is ‘ hot..!, and in the boys* circum cision .ritual,jm ysticaLprotection. in jthe form o f medicines is used against both w itchcraft and for nication. A gain, like w itchcraft, sexual desire causes social disturbance—-rrmrital inhdeHty,^crimes o f. passion, jm d.argum ents over jh e repaym ent ofh rid e-w ealth . In the w ords o f the most often sung N dem bu hunting song, * the buffalo in the bush, the w om an in the village, are the death o f man \ T he-tw o beliefs coincide in the notion o f the kahwehwi,. the g h o s to f a w itc h ’s husband,, .k ille d . b y . her.sorcery,_X hc kahwehwi -is~thought to be an incubus w ho continues to have intercourse w ith his w id ow , and tO cgive..her., inordinate, pleasure.— -On. these—three counts, thexi7-h er _foreignness, her. diligence,, and her forrner_nyniphomania, N yam uw ang’a came to be reckoned a w itch as she grew older. A series o f deaths in her lineage w ere attributed to her w itchcraft : those o f Shim ili (I,G i8 ), Kajata (l,F io) and Ikubi (I,H i 4).
O ld w oraen are .o£ten..considered .to. be.w itches. LfLthey are widowsjor-divorceesH keiSfyam uwang’a^wkhjiO-Sonsjorhrothers, they_have no one to obtain m eat, fo r. them, and have to xely on the^iimcertaiii - generosity.of. rem oter. kin__T h ey..are .constandy grambling._ab.out^lack-.Qf itieat, „and. in. .a..society_with. a lively belief. in_the necrophagous_practices-.of—witches,_this- is highly suspicious, behaviour. . It is. beHe-Ved that wit.ches.coliect. together in_the bush_with their_fomiUars^.and. divide_andLeat-.the..bodies o f their, wictim s. T h ey substitute in the m eanwhile a simulacrum o f the deceased, fabricated from his faeces and bod y dirt, in his
152
Schism and Continuity in an A frican Society
|
place in the village ; and this is mourned over and buried b y his relatives. It is, I think, significant in terms o f this belief that the j alleged cause o f N yam uw angV s grudge against Ikubi was the I latter’s refusal to give her m eat.1 j P itc h e s usually strike,, at...their... close. jnatemaL.kin,_on..whom they; depend, and.w ho therefbre are .those, w ho neglect.-them by notJuM lling their obligations. Itis usuahy_Qld.wiomen.who are negletted because they tend to. be an economic_burden_to the com m unity. Thus allowances are seldom _mad.eTar_old_wonien detected as witches by any meansL_of-divination__In...the_past, it is said, they w ere burnt to death. Today,...they_are often driven, from their, villages. Accordjuag_to_my observations, old womenwho..hy.ewith.closekin.s.uch_as.son.sLQrbx-athersrjLlthough they. m ay be suspected of.witchcraft,.„are...never_treated_ in this way... It is .the class x>f_unfortunateddLwomeii_who are w ithout im m ediate, male kin, but .who. are. classificatoxy_ m atrilineal kin o f the village headman, that acquire .the. role o f scapegoats for misfortune. T hey consolidate the rest o f the .village .or lineage membership against them. H ow ever, in the case o f N yam uWang’a, it w ould appear that Sakazao m ade a mistake in banish ing her. For unknow n to him at the tim e, N yam uw anga’s daughter Ikayi (I,H i2), w ith her ow n small daughter, had just returned from the C ongo, having been aw ay from M w inilunga for five years. She was staying at Chim bila Village. Her husband had recently received a long sentence o f imprisonment for manslaughter after a figh t w ith Ikayi’s current lover. I f Sakazao and Line had retained N yam uw ang’a they w ould have acquired tw o further wom en o f their m atrilineage w ho w ould have borne children for that group. B ut it often happens .that after a death has occurred,...intense animosity, is atoused, a n d if this is. focused-on_a_putative_witch, a reasoned consideration.o f the resultmgsociaLsituarLan.Js_swept a w a y b y anger and hatred against her. T he prim e movers in this m atter appear to have been Ikubi’s father N toloiu (II,D2) and her cross-cousin Gideon (II,E i), both from Shika Village lineage. T here is a long tradition o f interm arriage between Mukanza and Shika villages.2 Kasonda’s senior w ife (II,E2) 1 Cf. how quarrels in preceding episodes arose over the division o f meat. 8 See Appendix II.
" fS
©
153
G
came from Shika, Mukaixza’s first w ife (II,D3) w as a Shika wom an, Kasonda’s father (II,D i) cam e from Shika, and so on. For this reason, since marriage is liable to in vo lve disputes between the kin o f spouses, considerable tension as w e ll as am ity existed between the tw o villages. N yam uw an gV s form er hus band, W illiam M atorokoshi (11,D 7), cam e from Shika, and she had stayed w ith him there fo r several years. W illiam had recently divorced her because he said that he used to dream every night o f * animals * (tunyama), m eaning that he feared that he was being bew itched b y tuyebela. H e said that these belonged to his w ife, and sent her back to M ukanza V illage. In addition, she had quarrelled continually w ith members o f the Shika lineage, refusing to cook fo r the men in the ckota, and squabbling w ith the w om en. W h en Ikubi (I,H i 4) died all the accum ulated dislike o f N yam uw ang’a felt b y Shika fo lk, and their suspicions o f her w itchcraft, broke loose and received expression in the beating given her b y G ideon (II,E i). Sakazao, furious at the disruption o f the fo llo w in g he was beginning to build up, ordered her out o f Line Farm. She had already left her h u t near the main village circle to build next to Ikubi’s m other, Kalusa (I.G ip ), her uterine sister, because she had quarrelled w ith Gideon and M anyosa (I,G i3 ), her neighbours. G ideon was n o w vio len tly opposed to her return. In any case Sakazao had ordered her to g o and live in Chim bila V illage i f the headman w ould allow her to do so. W h en Sandom bu heard that N yam uw an gV s daughter had joined her m other in Chim bila he im m ediately saw this oppor tunity both to add to the strength o f his ow n farm , and to ensure its continuity through tim e b y inviting them to stay w ith him. A t last he w ou ld be in a position to acquire the nucleus o f a m atrilineal group. Three generations o f m atrilineal kinswom en w ould be added to his follow in g, even i f one o f them was a w itch, another a prostitute, and the third a child in arms. In allow ing N yam uw ang’a to reside at his farm Sandombu was openly defying the people o f M ukanza V illage, and behaving like the headman o f a rival village. H e was acting as though he intended to split M ukanza V illage. N yam uw ang’a eagerly accepted Sandombu* s offer, for she was persona non grata at Chim bila on account o f her reputation, and
O
M atrilineal Succession
€1
© © 0
0 Q O 0
a © © ©
© e o
©
G © © 1f K; C? m c
© © G
®
G ©
o C) o
o o
D O
o c Q
.o o o c 0
G O G
O O
o
o
o c 0
c
o €=
o o G
o o
154
Schism and Continuity in an African Society
besides she had left her large gardens behind at M ukanza. She and Ikayi (I,H i2) came and lived at Sandombu’s farm where, at the tim e I left M ukanza V illage, Ikayi was m aking m oney by selling her favours to passing lorry-drivers, African, traders and other w ell-do-do Africans, although not to indigent villagers. She was reputed to give N yam uw ang’a,and Sandombu a share o f her earnings, and became a lively source o f scandal am ong the respectable older w om en o f M ukanza Village. A bou t this time too Sandombu added to his follow ing a young lad named Chikim hu (I,J6), daughter’s son o f Ikubi’s m other Kaiusa (fC ip ). Chikim bu’s m other had gone w ith her husband to the latter’s place o f em ploym ent in the Congo urban area. Chikim bu was too w ild and w ayw ard to be taken w ith her to her new husband’s place, she said, and N yam uw ang’a offered to look after him. Soon he too acquired a reputation for sorcery in spite o f his tender years : m y w ood and water carrier, a tall youth from a nearby village, gave notice because Chikim bu, h a lf his size, had threatened to kill him as the result o f a quarrel over a game o f hopscotch. Chikim bu was the problem child o f the local Mission out-school and his teachers could do nothing to discipline him. I have given this thum bnail sketch o f Chikim bu to show w hat kind o f a follow in g Sandombu had collected around him to further his unsatisfied ambitions. A t this tim e his farm con sisted o f his senior w ife’s children b y other men and a number o f rem ote m atriiineal kin o f malabu lineage w ho w ere social outcasts and misfits. B u t for this very reason they could be guaranteed to remain lo yal to him and w ork hard for him accord ing to their varied abilities. Chikim bu, for exam ple, was always running errands and doing odd jobs for Sandombu, although ne w ould do nothing for anyone else. N o t unnaturally, the people o f M ukanza V illage referred to Sandombu’s form w ith a selfrighteous shudder as ‘ that village o f w itchcraft * (mukala wawuloji). B u t Sandombu continued to visit the chota every day, to take part in discussions on such important matters as bride-wealth, the holding o f rituals, death-payments, cases o f petty theft, slander, etc., and to state his opinions forcefully and authori tatively. It was clear that at some time or other he w ould make another bid fo r the headmanship. B y the m iddle o f 1953 it became obvious that the headman
M atrilin eal Succession
155
Mukanza. Kabinda was becom ing very old and infirm . H e spent m uch o f his tim e dozing in his kitchen, and although he still w ent into the bush to hunt, he seldom killed anything. The question o f succession to headmanship was once m ore beginning to disturb the village. B y this tim e I had made m y permanent camp ju st on the outskirts o f the village, n o t far from M ukanza’s ow n huts. I was, therefore, in a position to observe the trend o f events far m ore closely than before. Kasonda, still m y cook, was n ow m aking plans to found a form o f his ow n . H e said that w hen M ukanza died the village w ould break up, som e going w ith him , others goin g to Sandombu, and others staying w ith Sakazao. Shortly after I arrived M ukanza publicly announced in the chota that he had appointed Sakazao as his mulopu o r suc cessor. This was done in the presence o f N sang’anyi, headman o f the oldest K aw iku village, and o f N swanakudya, headman o f an im portant neighbouring village belonging to the lineage o f C h ie f Kanongesha. Sandombu was aw ay at the tim e on a visit to Ekelenge Area. W h en he returned and heard the news, his first action was form ally to succeed to the name o f K ahali by asking N sang’anyi’s permission to use it, a name w hich, it w ill be remem bered, was the ancient title o f the M ukanza headmanship. N sang’anyi, after dem urring at first, agreed in the end to give permission. M ukanza publicly approved Sandombu’s step, and said sm ilingly in the chotat w hen both Sandombu and I were present, * N o w there are two villages, M ukanza and N swanakahali.1 Sakazao w ill succeed in M ukanza V illage, and perhaps 2 someone or other w ill succeed in Nswanakahalid Sandombu said nothing but shortly rose and returned to his village. Trouble was patently brew ing. A w eek later Sandombu invited m e to drink w ith him in his house. H e was broodingly drunk w hen I arrived. Soon he asked abruptly, * W h at is the nam e o f this village in the Govern ment tax-register ? * I said that it was w ritten down, as ‘ K ahaliM ukanza \ * Yes,* he said, * in 1947 Bw ana Heath [the D istrict Com m issioner] cam e to the village to collect tax and Kahali * Literally, * Heir-of-Kahali \ * The use o f this * perhaps * (kw iji) was a powerfully compressed taunt which contained the notion that Sandombu's followers were ail witches or sorcerers who might bewitch one another and was an oblique reference to Sandombu*s sterility.
156
Schism and C ontinuity in an A frican Society
Chandenda, m y uncle, told him to change the name in his booh from M ukanza to Kahali, fo r the old name from lon g ago is Kahali ; but M ukanza Kabinda and Sakazao told him that the village had been founded b y M ukanza fCandulu (I,F6) and was know n b y all as M ukanza. So Bw ana Heath was very clever and w rote dow n Kahali-M ukanza. B ut he was w rong. Y o u have w ritten our history and you kn ow that the true nam e o f the village is Kahali. Please g o to the Bom a and tell the D .C . to change it to Kahali.* I replied that although it was true that the village used to be Kahali, the m ajority o f people today thought that the village should be called M ukanza. A t this point Kasonda cam e in, although he had not been specifically invited. H e quickly picked up the thread o f the discussion and sa id : * A long tim e ago the Lenje people had a village in the place where B roken H ill n o w stands. W h o remembers the name o f that village n o w ? People on ly m ention “ B roken Hill**, “ Broken H uT *, “ B roken H ill’*— even Africans. N o w a long tim e ago there was a village called Kahali. That village divided and a new village was started b y M ukanza Kandulu. M any people came and visited M ukanza V illage. For m any years people have o n ly been speaking o f M ukanza. It is like Broken H ill/ Sandom bu angrily retorted that he had ju st visited C h ie f Ikelenge’s area and that there people still thought o f M ukanza V illage as Kahali. Kasonda replied, * Ikelenge Area is another place, not this place/ T he conversation turned to other matters, and after a w h ile I left. Sandombu.went-om drinking..heavilyand-iruthe...evening another social drama began w h icK revealed that th e jealousy .over succession had n ot been eradicated but had beenJfestering.beneath the. show o f outw ard harm ony. Sandombu was w ell aware that M ukanza, Kasonda and Sakazao w ere attem pting to seal him o ff from any possibility o f succeeding, b y asserting that his farm n ow constituted a separate village, and that M ukanza V illage was not continuous w ith Kahali V illage, w hich had, in their reckon ing, becom e extinct. Perhaps he felt that he had made a mistake in taking the name o f Nswanakahali, and that M ukanza and Kasonda w ere chuckling over their exploitation o f that mistake. W hatever the cause the fact was that b y nightfall he was in a prodigious temper, and threw over the restraints he had imposed upon him self fo r years.
a
M atrilineal Succession
157
SO CIAL D R AM A V Sandombu Slanders and is Slandered (my own observations) W hile Sandombu was drinking great excitement was aroused in the village by the arrival o f a masked ikishi dancer from a Boys* Circumcision Camp some miles away. A t a certain point in die seclusion period o f the boys, the dancers emerge from the camp and visit ndgnbouring villages, accompanied b y ritual officials from the camp, to obtain gifts in cash or kind from the villagers or to extract various kinds o f forfeits from them. The whole village was in a condition o f social excitement, with drumming, dancing, singing, and drinking. This reached a high pitch when m y kitchen accidentally caught fire. Zuliyana (IJ2), Sandombu’s junior w ife and daughter o f Mukanza and Nyamukola, went to join the other women who had assembled to sing and dap around the ikishi dancer. She came back in a temper because she did not possess a new dress as some o f the other women did, and began to harangue Sandombu on this theme. She was assisted by Katiki, the senior wife, and both accused Sandombu o f giving away large quantities o f millet beer to ingratiate himself with people instead o f selling some to buy them new cloths to be made up into dresses by Kasonda, who had by this time bought a sewingmachine. They said that apart from organizing the ckenda collective tree-felling party, Sandombu had done little in the w ay o f cultivat ing the millet, while they had weeded it, harvested it, thrashed and winnowed it, ground it and made it into beer. Certainly he was owner o f the seed, but they were entitled to some reward for their labours. Zuliyana then said that Sandombu just wanted to make himself a big man ; that was w hy he had taken the name o f Kahali. He wanted to become headman o f the whole village. In view o f the previous arguments o f the day this was an unfortunate remark and Sandombu staggered towards Zuliyana to beat her. But he was extremely drunk and she easily evaded mm and ran o ff to her mother and father for protection. Sandombu did not beat Katiki, who seemed to possess some sort o f influence over him, and she persuaded him to sleep o ff his rage. But when he awoke later that night, his wrath had not abated and he rushed to Mukanza’s hut. Meanwhile Zuliyana had told her father that Sandombu had been boasting to many people that he was die elder (mukutumpi) o f the village while Mukanza was his junior (kansi, literally, * child *), and that the name * Kahali * was a * heavier *. name than 4Mukanza ’. Mukanza had also been drinking and a
n f>
o O
o o (.? O
o G O G O
V
-^•ANDERSON MULUMBl
C f\
./
o o © {; ï
o o c o © ■ o ■■O c © © © G
G C■ ) 0
<3
9__9
M atrilineal Succession
159
major dispute developed between them. Sandombu accused Zuliyana o f trying to bewitch him with tuyebela which she had been given by her mother Nyamukola. Zuliyana began to deny this energetically from the kitchen where she was sitting with her mother, but both men told her to keep quiet. Mukanza said : * Sandombu, you have slandered (ku-tamukita) m y wife, and I w ill give you a big case. Also you have been heard by others to have threatened me. W hy do you want to kill me ? Have I not given you m y daughter ? You have a very bad liver to speak as you do to me. You must take your wife back again, treat her well, and buy her a new cloth. I f you want to speak in this w ay you must stay away from the chota o f Mukanza. Sandombu refused to take back Zuliyana, and repeated that she and her mother wanted to bewitch him. He went back to his house and for a long time his voice could be heard roaring out abuse against the people o f Mukanza, while those in the chota who were not immediately involved shook with suppressed laughter, especially the little boys. Zuliyana slept that night in the hut o f Line’s wife, Maria (I,Hi9), the latter’s husband being away. N ext day Sandombu came to the chota and demanded die return o f his wife. She came back with him but Sandombu was told to return the day after to explain his conduct at a meeting o f village eople. He did so and persisted in accusing his wife and mother-iniw o f witchcraft. In such intra-village meetings in Mukanza the summing up (ku-sompa) was usually left to a blind member o f malabu lineage, Chayangoma (I,H i5), who had lost his sight as the result o f an industrial accident in a sawmill in Bulawayo. He had neither wife nor child, and used to stress that he felt free to say what he liked as he was a ‘ dead person *, a mufu. Since he was ‘ already dead ’ no one had any motive for bewitching him, and he could speak impartially. In his summary o f this case he said that grave charges had been made both by Sandombu and Mukanza, involving witchcraft. In the eyes o f the Boma witchcraft accusations should be severely punished. Therefore he thought it wise i f the affair were kept within the village and not discussed by the local * law-men *, the mahakuv or vicinage elders with a high reputation for advocating and judging cases, since news might leak out to the Boma and their relatives would be arrested. Sak&zao said that Sandombu was drunk at the time and could not be held responsible for his words. I f he apologized and paid 1os. to Nyamukola and a new cloth to Zuliyana nothing more need be said. But Sandombu refused to pay. He said that he had been wronged by his wife, who had first caused trouble in the house by starting an argument and had then borne tales to her parents against her husband to cause trouble in the village. He had accepted her back and yet they still wanted him to pay. As for the accusations o f witchcraft,
E
16 0
Schism and Continuity in an African Society
both parties had made them in anger/ He would pay nothing to Nyamukola, for he had been greatly provoked by both women. It so happened that Sandombu himselfat this very time was bringing an action for slander against the headman o f a neighbouring village, Chibwakata (cf. p. 150). A t a beer-drink at which I was present the day following the Zuliyana affair, headman. Chibwakata had accused Sandombu o f having accepted 10s. from Kasamba, Chibwakata’s classificatory son by a slave mother, to kill the headman by sorcery. Sandombu at first thought that Chibwakata, who was his own classi ficatory grandfather, was joking with him, and laughed heartily. But Chibwakata, who was drunk, became very angry and said that Anderson Mulumbi, another classificatory son and Kasamba’s mother*s sister’s son, had been witness o f the transaction. Sandombu promptly charged Chibwakata and Anderson Mulumbi with slander (mutamu) and asked Chibwakata to settle with him on the names o f mahaku who should discuss the case. Thus Sandombu was in the odd position o f being defendant in one slander case and plaintiff in another at the same time, both cases involving accusations o f sorcery and witchcraft. The Kasamba case was settled first. The mahaku were Nswanakudya, headman o f a village o f the Kanongesha lineage, headman N g ’ombi, and a Kawiku elder. Mukanza was invited, but he refused to discuss any case in which Sandombu was involved. Gideon (II,E i), son o f Mukanza, acted as Sandombu’s witness (chinsahu). Sandombu described how on the previous Sunday a quarrel had broken out at a beer drink between Kasamba and Anderson because the former had obtained the job o f D .O .’s cook for which both had been compet ing, and Anderson had pushed Kasamba into the fire. Kasamba had attacked Anderson and chased him away. N ow Anderson was trying to get his revenge by causing bad feeling between Chibwakata and Mukanza’s people. I f he said he had seen Kasamba and Sandombu conspiring to bewitch Chibwakata he had lied. Chibwakata, who had made the accusation, and Anderson should pay damages to Kasamba and himself. He demanded £ 5 from them. After much argument the court said that Chibwakata ought to pay Sandombu 30s., and Kasamba the same amount. Anderson had to pay each o f the plaintiffs 10s. In this case Sandombu identified himself with Mukanza Village and with the exception o f Mukanza and Nyamukola all the Mukanza people were pleased with, the decision. They had old scores to pay o ff against both Anderson and Chibwakata. A t a deeper level, too, the level o f the Ndembu-Kawiku cleavage, they were pleased, for Chibwakata was the descendant o f the Ndembu warleader who had first attacked the autochthonous Kawiku. In addition, as w e have said, he had been reared by Chokwe, the hated slave-
M atrilin eal Succession
161
raiders o f the nineteenth century, and in certain circumstances was identified with the Chokwe. N ext day the Zuliyana case was settled, still at the intra-village level. Chayangoma (I,Hi5) was again asked to arbitrate. Sakazao was suggested, but turned down by all parties who claimed that he had no inherent faculty (chisemwa) for settling disputes. Besides, he was own brother o f the co-plaintiff. Chayangoma said that Sandombu should pay Nyamukola 10s. for defaming her character, since the previous day he had obtained damages from Chibwakata for slander against himself. In both cases there had been witnesses and the fact o f slander had been clearly established. Sandombu paid up, and people joked with him afterwards, saying that after all he had made a profit from his two cases. Peace was restored for a time in the village. Analysis It-w ill be recalled that I divided the. social -drama ..into four m ainphases.: _breach,..crisis,„operation o f xedressive-mechanisms, and-either reintegration ofodto-SO cial-gm up-or.sooaLrec^n|tie|i^^e o f irreparable schism. -. .Im plicit .in. .the- n o tio n -o f xeintcgrad«c^ f^ ^ ^ the__CQncept._o£.^oqial. eauiH hrium ... This, concept ■1 view that a social system is made up o f interrelated i persons and groups, whose interests are som ehow ^.maintained in balance ; ..and further, ..that when_disturbance.occurs^ read justm ents. are. m ade w hich, have the effectofrestorin g. the balance. B ut it is necessary to rem em ber that after disturbance. has_occurred and readjustments have, been made, there m ay have, taken place profound m odifications in the internal relations. o£.the_.group. T h e new equilibrium is seldom a replica o f the old. T h e interests o f certain persons and groups m ay .ha.ve_..gained__at., the. expense ofithose o f others. Certain relations betw een persons and groups mayJ 3ave-mcreasedinintensity^hileo.thers.m ayihaye_.dimuushed. Others.again m ay have, been com pletely, rupm red w h fieiiew rela tionships have com e into being. A social .system_is.in_dynamic m ovem ent-through space and. tim e,-in som e-w ay. analogous to an organic system in that it exhibits grow th and^decay,. m fact the .process o f metabolism. In one aspect,., the..spcid drama is a process w h ich reveals reaHgnments .o f social „relations .at critical points o f structural m aturation or_deeay.in..another,_it_m ay be regarded as a trial o f .strength betw een . conflicting-interests in which_persons or groups try_.to_. manipulate., -to.. their.._own
362
Schism and C ontinuity in an A frican Society
advantage the actually existing network o f social, relations,-both structural and contingent, within the_system...Thus the social drama, m ay represent either the natural, inherent .developm ent o f a given, social system thro-Ugh -space-tiine-at a distinct phase, at a_critical point o f maturation, .or. the. dehberate.--attemptSL.by some .o f its m em bers -to._accelerate.-or. retard. that_dexelapment. It m a y be . either an index or a vehicIeL o f -change. In most case^Jboth aspects are.-present. Thus in Social Dram a III, the crisis w hich follow ed the death o f Kanyom bu (IJFp) indicated that there had been a change in the internal residential structure o f M ukanza V illage, in that malabu lineage had increased in membership and strength. A t the same time, through N yam ukola (I,H io) and Nyakaiusa, slave-w ife o f the senior elder o f MAtABU lineage, an attem pt was made b y that lineage to im pair the claims o f Kasonda, then the most prom inent m em ber o f nyachintang ’a lineage, to succeed, b y fastening on him a charge o f sorcery. Kasonda’s new role as a man o f relative w ealth and influence residing outside the village w ith Europeans was also expressed in the charge ; in the course o f the drama he used this new role to rebut the accusations, to m obilize n yach in tan g ’a lineage, and to further his ow n interest in the succession. la this social drama it may also be noted that the^socio-spatial system form ed b y M ukanza Village, was .b y n.o_means_the-same systein as that revealed in Social Dram a I at-the-tim e-of the death o f K ahali Chandenda. Form er members h ad go n e, n ew mem bers had com e into it. The conflict between Sandombu and the. rest was in abeyance ; a new cleavage existed between the form erly allied lineages o f malabu and. n yach intang ’ a . O ld members had form ed new.and.broken o ff old. attachments_T h e basic principles governing residential affiliation remained,— m atriIiny,jthe unity o f m ale matrilineal kin, virilocal marriage,, the relariom Jb-Ctween genealogical genfirari_ons.,_the.jamty„.o£jLiterine siblings, etc.__ .But the .relations.. h.etweeiLp.er&om-Jmought into daily, contact b y these principles_had_altered. A nd in. .the course o fjh n e the status o f genealogic.afi^.lh^ed-..p.ers.ons h^d changed asym m etrically. M ukanza was now._a_ headman.. ancLN yam ukola a headman’s w ife. Sandom bu-.was..no-Jonger_..a ,rival clairiiant for office but M ukanza’s jun ior, .under-_his._authority. Kasonda w as n ow old enough to be a„po.werful„claimmtLin.£uture, and as the result o f external factors, had increased his internal
M atrilineal Succession
163
influence, in _the_ system . Sakazao w as older and had a large foUowing w ithin. the villa g e.,.. As b<^ye
164
Schism and Continuity in an A frican Society
B ut apart from these considerations, Sandombu had in other respects im proved his position in the years that intervened be tw een Social Dramas II and V . H e was now a farm head with a follow ing. H e had obtained w ork for m any M ukanza men. H e had been liberal in offering beer to the villagers and to many others in the vicinage. There w ere even some w ho, like M anyosa’s husband Chikasa,1 and his m other’s sister’s daughter’s daughter Bibiana (I,H4) in Ikelenge Area (just on the point o f divorce and m other o f four children), w ould support his future claims to village headmanship. Even i f he did not succeed in M ukanza V illage, he m ight yet be headman o f a large farm and resuscitate the title o f Kahali. It is this heightened importance o f Sandombu that was re sponsible for the w a y the case was played dow n and kept within the village, for the small am ount o f the fine im posed, and for the fact that Kasonda kept in the background instead o f openly givin g M ukanza his support. A conciliatory tone was adopted towards Sandombu throughout, no attem pt was m ade to retain his w ife, and after the case was over it was jo k ed about. A lthough Sandombu paid damages to N yam ukola he did not lose face thereby. In the altered climate o f opinion, as he had just scored a victo ry over Chibw akata and Anderson M ulum bi, it was in Sandombu’s interest to make a gesture o f goodw ill towards the village. O n balance he had considerably im proved his position after these tw o cases, and M ukanza’s refusal to have anything to do w ith a case involving Sandombu had not added to the old man’s popularity w hen Sandombu had been disputing w ith a m ajor enem y o f M ukanza V illage. Nevertheless, b y die end o f Social Dram a V , there is reason to predict that M ukanza V illage w ill split up along certain lines after the headman’s death. Kasonda told m e shortly before I left that whereas he had once thought o f supporting Sakazao’s claim to headmanship and remaining w ith the latter, he, Kasonda, had becom e unpopular w ith the people o f maxabu lineage and they w ould continually quarrel w ith him . H e w ould start a farm near a form er site o f M ukanza V illage, about a m ile from 1 Manyosa supported her brother Kasonda, but her husband, Chikasa, was very friendly with Sandombu. These two, Manyosa and Chikasa, had married late in life, after each had reared families with previous spouses ; and they quarrelled incessantly over almost every possible issue.
PLATE IV A Sy m p t o m o f S o c i a l C o n f l i c t Headman Mukanza chats with his wife Nyamukola during an interval in an Ihamba ritual. An apprentice-doctor from Kafumbu Village (see Social Drama VI) is applying a cupping-horn to her back by suction. This ritual, to remove the tooth of a punitive dead hunter from the patient’s body by 'catching’ it in the cuppinghorn, immediately preceded the Cftihamba ritual described in Chapter Ten., Nyamukola, principal patient in both rituals, occupied a social position under severe strain at this time. Note the forked branch with a piece o f ant-hill at the base set up as a shrine to the hunter’s spirit. Nyamukola sits on a duiker skin. Traces of pounded herbal medicine adhere to her skin.
n
o
o o o o o o o G O O O C:
P e
o o o
o
o o o o o 0 o O 0
c c fe
M atrilin eal Succession
165
the m otor road, w ith M anyosa’s (I,G i3) children, w ith Chaw utong’i (I,G i2) and some o f her children, and w ith G ideon (II,E i), at once his m atrilateral and patrilateral cross-cousin, and his children. H e him self had tw o w ives and five children, so that he w ould have a fairly large fo llow in g. M anyosa’s husband m ight g o and live w ith Sandom bu but there was a strong possi bility that M anyosa w ould divorce him and com e and stay w ith Kasonda. A gain , the husband m ight accom pany his w ife to Kasonda’s form . Sakazao w ould be left w ith m ost o f his lineage kin. Some members o f both lineages m ight build at Sandombu’s form. Sandombu, how ever, told m e that i f Sakazao succeeded, many w ould leave him and live at Sandom bu’s form , for Sakazao had a poor head fo r * speaking cases \ was idle and w ithout w ealth, and no-one really respected him . Sakazao said to m e that i f he succeeded, the village w ou ld not break up, because unlike both Sandombu and Kasonda, he had never been accused o f sorcery and no-one feared him in this respect, and because some o f M ukanza’s children belonged to m alabu lineage and others to the lineage o f his kinsman Line’s w ife (I,H i9). H e thought that Kasonda, M anyosa and C haw utong’i w ould rem ain w ith him, and i f they stayed their children w ou ld stay also. O f these appraisals that o f Kasonda’s is perhaps the most real istic. Sakazao, for instance, cherished the illusion that Kasonda w ould remain w ith him , whereas I was aw are o f Kasonda’s plans to m ove, perhaps even before the death o f M ukanza. T he main link between nyach in tan g ’a and m alabu lineages is the mar riage between M ukanza and N yam ukola. Since they have already becom e genealogically so distant, once that link is snapped there w ill be little to hold them together. W hen M ukanza dies, Kasonda w ill becom e the senior m ale o f chinbng ’a (I,JS4) lineage, the num erically predom inant segment o f n yach intang ’a lineage. I f he founds a farm m ost o f its members w ill fo llo w him and not Sandombu. Basically then the division w ill probably follow the lines o f lineage segm entation. Sandom bu’s lineage o f NYAKAPAKATA (I,E i), w ith Bibiana as its forem ost fem ale, w ill live separately from the other sub-lineage o f nyachintang ’a lineage, that o f chineng ’ a . m alabu lineage w ill probably remain in the site o f the present M ukanza V illage, w hich w ill becom e. Sakazao Village.
0
M atrilin eal Succession
o
167
From beneath the m anifold and transitory shifts and realign ments o f social relations from year to year in die village, w e can see the fundamental cleavages between maternal descent groups emerge in the lon g run. M arriage and alternate generation alli ance are means o f delaying the final residential split between lineages, but as the w edge o f tim e drives deeper into their division, so these lateral and supplem entary linkages wear thin and break. B u t at the same tim e this division into com ponent lineages, in another aspect, represents a division betw een genealogical enerations. M ost o f the leading members o f n yach intang ’a neage belong to the senior adjacent generation (G) to the leading members o f m alabu lineage. N yam uw axig’a o f malabu lineage belongs to the senior generation (G) but she lives w ith Sandombu o f her ow n generation, but o f a different lineagesegment. Chinem a (I,G zi), Sakazao’s classificatory m other’s brother, also o f m alabu lineage, w ho is w orkin g atK itw e (a m ining township on the Rhodesian Copperbelt), has reported that when he returns he w ill not live at Sakazao’s village, fo r Sakazao is his ju n ior. Sandombu claims that Chinem a w ill stay w ith him. M ukanza’s eldest son b y N yam ukola, Zachariah (IJ3), a m em ber o f m alabu lineage and sister’s son o f Sakazao, has said that he w ill go w ith Kasonda, his cross-cousin, i.e. in his ow n genealogical generation, reckoning through his father’s side, and in the alternate generation, through his m other’s side. C onflict between lineages, and. hetwcen_adjacent^gexiealogical generations throw s into high relief the role of-aUiance between alternate genealogical generations _as_a_m em sofm tegratm g the residential unit. In a sense__the._village of^ M ukanza was held togeihesLby-the_grandparenfegrandchil
0 0 ©
o © 0 o
f
O ©
©■
©
© ©
©
© ©
© © .
©
©
© ©
© ©
© © m
"O
o n
o a o o e o c. o
G
G O
a o
G
e o o 0 0
© G
O O 0 '\ y o G
G G €)
168
Schism and C ontinuity in an A frican Society
afflicted by the spirit o f N yam ukang’a (I,E5), a daughter o f M alabu ; Sandombu was the principal male organizer ; and Sakazao’s senior w ife was the principal female organizer. The ritual adepts w ere drawn from all the villages o f the vicinage. B ut since this ritual appeared to have the additional social func tion o f reconciling conflicts that had arisen between Mukanza V illage and other villages w ith which it was connected b y ties o f kinship, affinity and spatial contiguity, the ritual cannot be discussed until a careful consideration o f the sources o f inter village disputes has been made. In the next tw o chapters I present material indicating under w hat conditions fission o f villages actually takes place, and what kinds o f social grouping are involved.
CH APTER VI VILLAG E FISSIO N , S L A V E R Y A N D S O C IA L C H A N G E fission ’ I mean a situation in w hich a group or B Y_.lvillage section w ith in . the .total membership o f a „village_de£aches itselfand builds a separate settlement. JT_a^umber_ofYadiyiduals leave_a village and severally affiliate themselves to. .diiferentJieaclm erLthisisnot fission__Sii33iladLy,if.aften.the„deatfi.oLa^headman or as_the result o f some disaster,.a„viUage >hceaks. up.,com pleteiy and its., members scatter in attachment to.many^cHfferentv,headmen, this is fragm entation,.not-fission. L d efin e.fission am ong the N d em buas the division, of. a tilla g e .com m unity,.along lines o f structural, cleavage so th at one section. .maintains ^continuity, usually sym bolized b y the retention o fits name. withJthe_ojiginal undivided village ; and.the other section .or sections ^tiamed, after theirdeading elders, seek to establish themselves as independent villages. The lines o f im m inent fracture in M ukanza V illage demarcated, as w e have seen, three segments o f a matrilineage, nyachintang ’ a (I,D i ) and m atabu (I,D2) lineages o f equal span, and within, n yach intang ’ a lineage, n yakapakata (I,E i ) and chineng ’ a lineages (I,E4), also o f equal span. In the case o f nyachintang ’ a and malabu lineages, genealogical distance from a com m on ancestress had led to their social differentiation as corporate units w ith certain independent social, econom ic and ritual functions. Breach o f spatial relations o f interdependence had been postponed b y m arital and alternate generation ties. Betw een n yakapakata and chineng ’ a lineages, the possibility o f social severance had been accelerated b y the ambitions o f Sandombu, the leading elder o f the form er, and b y social and econom ic changes brought about b y the introduction o f cash econom y w hich facilitated the em ergence o f increasingly sm aller settlements inhabited b y ever narrow er ranges o f kin. B u t in both instances fission seemed about to take place dom inantly along the lines o f lineage cleavage. In m any other villages fo r w hich I have records fission took place between segments o f a local matrilineage. Nevertheless, fission does not_always._oxeven. dom inantly take place between..s.egxnents...o£.an. _eq.ual_arder o f segmentation. V ery frequently the dissident ,group., .is. a uterine 169
ly o
Schism and C ontinuity in an A frican Society
sibling group,Jtheir..cbiid.ren, aiicl their ,mother,_i£ she .is. stilLalive. The seceding section m ay not consist o f m atrilineal .kin .o f ..the headman at all, but m ay be descendants o f a female slave, {ndungu) o f ...a village member. These., under the. pax. bdtannica,Jtizvc experienced an im provem ent in status and wish to-em ancipate themselves in fact and. in .sp ace from , their._former_.owiiers. Again, it may. happen -that a .village, is composite. in_memhership, some o f its ...inhabitants being, matrilmeafly. descended-.from a sister and some from a w ifb o f the first headman. T he m atrilineal descendants o f th e w ife m ay-split. o if from_those o f the sister, and build, separately. O r—the sister’s—descendants—may m ove. B ut in nearly ,all cases o f fission foe...prmciple-.Qfjnaternd de&cmt_is ..crucial,..and..maternalj^ ,a lm o s t alw ayilform -jhe.core o f a n e w . village. Ndernbu and Tallensi Social Organization In order to bring out the essential character o f N dem bu village fission it is instructive to com pare briefly the main features o f Ndem bu and Tallensi social organization.1 Fortes w rites : Tale society is built up round the lineage system. It is no exag geration to say that every sociological problem presented by the Tallensi hinges on the lineage system. It is the skeleton o f their social structure, the bony framework which shapes their body p o litic; it guides their economic life and moulds their ritual ideas and values. [The Tale lineage is] a striedy unilineal, agnatic descent group, [and] eight to ten ascendant generations are usually reckoned between contemporary minimal lineages and the founding ancestor o f the maximal lineage o f which they are part (p. 31). [All Tale lineages] are hierarchically organized between the limits o f the minimum span, i.e. the minimal lineage, on the one hand, and the maximum span, i.e. the maximal lineage, on the other. Thus every minimal lineage is a segment o f a more inclusive lineage defined by reference to a common grandfather, and. this, in turn, is a segment o f a still wider lineage defined by reference to a common great grandfather ; and so on, until the limit is reached— the maximal lineage, defined by reference to the remotest agnatic ancestor o f the group.. . . W ithin a lineage o f whatever span each grade o f segmenta1 Fortes, M., T h e Dynam ics o f C lanship among the T allensi (1945), p. 30.
a Village Fission, Slavery and Social Change
171
don is foncdonaDy significant. Bach segment has its focus o f unity, : and an index o f its corporate identity, in the ancestor by reference to i whom it is differentiated from other segments o f the same order in the hierarchically organized set o f lineages. Sacrifices to the shrine : o f this ancestor require the presence o f representatives o f every segment o f the next lower order ; and this rule applies to all corporate acdon o f a ceremonial or jural kind o f any lineage---A lineage o f any span emerges in any o f its activities as a system o f aliquot parts, not as a mere collection o f individuals o f common ancestry (p. 31). Tallensi do n ot live in discrete com pact villages but in hom e steads, containing on average some 12 to 14 people, scattered apparently indiscrim inately, som e d ose together, others farther apart, over the country, in irregular blocks w h ich m ake up * settlements There is a distinct though sometimes narrow territorial separation betw een the outerm ost homesteads o f one settlement and those o f another. A ccording to Professor Fortes, stability and continuity are essential characteristics o f a settlement. They are implied in the native concept o f a teng (settlement) as both a definite locality and a fixed community. The ancient settlements have attained a very high degree o f stability and continuity. . . . Precise local orientation is essential in the economic activities and soda! relationships o f the Tallensi (p. 157). * A settlem ent is a m iniature o f the w h ole society, and reveals all the basic principles o f the social structure9 (p. 154). A settlement is a locality defined in relation to a corporate unit o f social structure, a lineage, d an , o r inter-connected group o f dans. T ale society, then, is built up round the hierarchically organized lineage system , and lineages o f considerable depth are anchored to specific localities. Each segm ent o f a single lineage o f w ide span tends to form a coherent residential cluster, situated on land which has been owned by its members for several generations, and in close proximity to the graves o f their ancestors. The lineage as a whole tends to form a local aggregate, though a somewhat looser one than any o f its segments (p. 197). A ccording to Professor Fortes, * neighbourhood ties are ipso
a o
a
0
o 0
0
a ■© © & o ■ © © a © © © © © © o © © © © © # © © ©
172
Schism and Continuity in an A frican Society
facto lineage ties, and therefore econom ic, religious, jural and moral ties 7 (p. 211). Lineage proxim ity is correlated w ith local proxim ity. This intense local fixation o f the lineage is no doubt congruent w ith 4the rigorous precision w ith w hich its members fall into a set o f hierarchically articulated segments, o f hom ologous form but all exactly differentiated both genealogically and functionally ’ (p. 19s).
AmQng^Mdem.b.u_thereJs_naolose_mrrelatiQn.hetweemJineal andJocal. ties.... Local, stability.and. continuity is-typical. o£_Tale, and_sp.atial_mobility is typical o f Ndembu-residentiaJLoxganization. Lineages are anchored to specific tracts-.of.-land-, among-, the Tallensi, and have deep genealogies ; and -the. ..fission., o f one homestead results in no m ore .than the estabMshment-ofLan adja cent homestead. A m ong Ndem bu an articulated lineage has little depth (see Tables XIII, X IV . and X V , p.. 8.3), and-its.^fFective structure is confined within_a_-smgle_village. Fission is often accompanied b y violent internecine conflict o f the kind des cribed in Chapters Four and Five, and results in the seceding group putting m any miles between itself and the parent village. A n N dem bu vicinage, w hich m ay perhaps be com pared w ith a Tale 4 settlement \ is not a locality defined in relation to a cor porate unit o f social structure, ‘ a lineage, clan, or interconnected group o f clans \ but is extrem ely heterogeneous in character, so that in certain vicinages none o f the com ponent villages m ay be m atrilineally interlinked. Cer.tain__v.Ulages^ .often. ..dispersed throughout., the entire Ndembu..region,^may. claim relationship through putative maternal descent, but they do_not.trace..descent to ajnamed com m on. ancestress and caim ot-place.one.another in a hierarchically organized maximal. lineage_with_functionally signi£Ic4 n t .grades _of segmentation. Such..a.. .*.lineage.!.,.is_not 4a system o f aliquot parts \ but * a mere collection o f .[small groups o f] individuals o f com m on ancestry \ The contrast between Tale and Ndem bu lineage organization is strikingly displayed in the ancestor cult. A m ong the Tallensi ancestors have permanent shrines and graves, w hich are attached to specific localities, w hich are usually surrounded b y the residen tial and farm land held b y a maximal lineage, and w hich therefore serve as foci o f the unity o f the lineage as a w hole and o f its segments : hence, as Professor Fortes succinctly expresses it, 4 the
V illage F issio n , Slavery and So cial C h an ge
173
ancestor cult is the calculus o f the lineage system.’ B u t am ong the Ndem bu, ancestor shrines consist o f quickset muyombu saplings, planted to placate the spirits o f deceased village members, w ho are still remembered as livin g persons b y their kin. These spirits m ust be placated because they have returned to afflict individual relatives w ith misfortunes, as the result o f neglect * to m ention their names * and pour them out beer when praying to other spirits, or because conflict has arisen between village members. Such shrines are abandoned w hen the village moves to a new site and it is very lik ely that n ew muyombu saplings w ill never be planted for the same spirits again. R em ote ancestors are seldom worshipped except in a few im portant and infrequent rituals, such as the boys’ circum cision ritual {Mukcmda) or the great ritual o f Chihamba ; and in both these rituals the unity o f all N dem bu, even o f all Lunda, is emphasized rather than the unity o f the particular village w hich sponsors the ritual. Graves are never used as shrines b y N dem bu and are feared and avoided rather than tended and used as sites o f sacrifice. For the Tallensi the ancestor cult is related to the land, to agriculture, and to permanent residence on the land o f w ell-defined corporate lineages. Fi5r..N dem bu, the ancestor. cult_is_ass.acia.ted.„with_the busludts-dangers-and _hlessings._with. the transience o f settlement, withjthe. hazards o f life , and .w ith the m obile human- group itself rather.J th a n .itssp e cific habitation. It hasl been_.suggeste.d in earlier.xhapters that the...high, m obility of-Ndemb.ur aticLthe small size o f settlements and their frangible character,...are. congruent w ith the high value set on hunting^.not-oniy..as..a_m eans o f providing an appetizing and nourishing...faad,_hut._also_ as an index_of m asculinity and as a historically, validated.m eans o f acquiring prestige. T he size and m obility of-settlem ents seem to be m ore closely related to the local availability o f gam e and sylvan .resources than to the prod uctivity. o f xhe soil. N dem bu supernatural.beHefk regarding thed ead w arkjtow ard s the prom otion o f m obility rather„.than .tpwards the creation o f deep attachments to the sites w here dead ancestors He buried. In die .past, when a person died, he o r she. was. buried. under the floor o f Jthe hut, and the village changed its site for fear o f the ghost. Beliefs regarding the dead am ong. Ndem bu, then, accelerated~the already h igh rate .o f m obility. O n the other hand, the ancestor cult o f the Tallensi tends to emphasize and validate the
174
Schism and Continuity in an African Society
intense attachment o f lineages to localities. W orship o f the patrilineal ancestors and worship o f the Earth are the deepest values o f Tale religion. A lthough in some respects they are opposed, since the ancestor cult tends to emphasize the divisions between lineage and clan segments w hile the. Earth cult tends to stress the com m on interests o f the widest Tale com m unity, the tw o cults are at the same tim e com plem entary, since the cultivations o f individuals and o f particular lineages are, as Professor Fortes puts it, * hut parcels o f land cut from the limitless earth *. Tallensi are attached to the soil and, given the existence o f patriliny as their dominant principle o f social organization, local and lineal groupings tend to coincide, and to be interlinked w ithin a com m on fram ew ork o f overlapping clanship and lineage ties that embrace the entire Ndem bu m ove over t . _________ game, and given m atriliny as their dominant m ode o f social organization, lineal kin tend to scatter across the region, local lineages tend to be sm all and unstable, and no deep interlocking local lineages and clans develop. T alc ritual expresses the unity or corporate groups from the unity o f minor matrilineages to the unity o f all the Tallensi, w hile in the great all-inclusive rituals the basic units in the com position o f ritual assemblies are patrilineages. In N dem bu ritual, as w e shall see, the basic ritual units are cult associations that cross-cut the relatively transient local and lineal affiliations. The dominant links between, participants in Tale ritual are links o f commonness and corporate identity ; in N dem bu ritual they are links o f likeness and tem porary association, lateral rather than lineal. W ith this general background o f comparison in mind, let us now consider the different modes o f fission in the residential structure o f these tw o societies. Am ong Tale fission does not represent a com plete break between the groups involved. Rather, it represents a process o f regular branching-offby means o f w hich the form ation o f new local units advances and projects the social structure through time. T he new group does not emancipate itself from the lineage ; it becomes a branch o f that lineage. A cleavage occurs between tw o hitherto nascent seg ments o f a m inor lineage inhabiting one homestead ; and w hile this m ay give rise to the establishment o f a new homestead, ties o f ju ral and ritual co-operation between the tw o groups
o
Village Fission, Slavery and Social Change
175
prevent radical breach and contain the cleavage w ithin the fram ework o f the m axim al lineage. But-am ong N dem bu, the cleavage_of a . village^-.W-hether or not along lineage lines, results in the. form ation o f a com pletely newjunit. A hiatus is made in the genealogical ,structure,of the original m atrilineage in_such a w a y that, after the^passage o f a single generation, no member o f either village distinctly, „recalls the precise.-point.of fission. A fter tw o or three generations„all that is left is a vague sense o f com m on maternal...descent_-Yet it often happens that, w hile the name o f the com m on ancestress and the relationship o f the founding ancestor o f the new village to members o f the old village are forgotten, members...of both groups continue to apply to one another kinship ,terms .according to-the.nom enclature, w hich expresses separation or. equivalence in their'genealogical generations. Thus the present Rhodesian C h ief Kanongesha N dem bi calls headman N yaluhana his ‘ older brother ’ (yaya), headman N swanakudya ms * older brother *, headman Shika his ‘ younger brother ’ (;mwanyika), and headman Kajing’a his 4 m other’s brother * (mandumi), although the exact lineal relationship between these m en can no longer be traced and all inhabit diffèrent villages w id ely separated in space. T he facts o f matrilineal kinship and the relevant genealogical generation category are recognized, but a m ore precise identification cannot be made. This system o f nom enclature m ust be distinguished from perpetual kinship since the terms are applied between indi viduals and do not express the relationship between political positions in the social structure. The generation principle governs the authority and marital structure o f a settlement. It is im portant for members o f a society w hich possesses a high rate o f individual m obility, in w hich persons are frequently changing their village affiliation, to kn ow in w hat relation they stand vis-à-vis the headman o f a new village in terms o f genealogical generation. O nce this relationship has been established a new m em ber o f a village know s im m ediately w hom to respect and from w hom to le vy respect, he know s w hich w om en he can m arry and w hich he m ay not, he know s in w hat arc o f the village circle he m ay build his n ew hut, and so forth. T h e stress laid b y N dem bu on genea logical generation as a principle o f local organization and as a means o f facilitating inter-village m obility, m ay be regarded as
o o
o 0 (■ '} o
o
o o ©
o ©
o o o o © o 0 0
e
G G G
G O .O
©
O
C-
c c c
C" ( ' C; (" (' c
( (.; ( :
C .( "■
{ (. (
.
( <■ ( C 1. r
t, t. i
c
! . .
lu c:.
176
Schism and Continuity in an African Society
an index o f the brittleness o f lineage structure, it m ay also be regarded as a means o f organizing local units whose membership contains a high proportion o f seminal children o f the male m atrilineal core, as w ell as a number o f cognatic kin. and slaves. In Chapter E ight I w ill discuss conflicts that arise in villages on the basis o f their m arkedly bilateral com position, and the role o f genealogical generation organization as a mechanism for con taining such conflicts w ithin bounds. T he point I w ish to stress in this chapter is that. although at theJével o f lineage organization structural amnesia, w ith ..regard to ...thé precise genealogical links , between villages, separated by fissionlof a .localJiiieage speedily .occjuxs^_m.emor.y,is .retained o f common, m aternal descent and .o f the .affiliation b y .-genealogical generation between, the detached, groups. Separation has been effected between tw o sub-groups.._within _one-s.ys.tem_oX_social relations, that o f the village, but several .sets o f ties w hich form erly interlinked these groups as members o f a single-com m unity con tinue to operate between them. Each group. is_now a unit w ithin a w ider system o f social relations,., and.w ithin-that w ider system each maintains w ith the other-ties—o f. a._moce. durable nature than m erely those o f jo in t .membership._Q£.a-_chiefdom, spatial propinquity, or affinity. After, the feelings„o£.anim osity associated w ith the. initial ...brea.c.h-ha.ve._died_ has a special claim on the hospitality. o£ the.,0 ther,,_ the. .members o f both „exchange long visits, and each.. may..setve_.in_tum as the base.o£the other’s hunting expeditions. It is thought appropriate i f cross-cousin and grandparent-grandchild (cf. pages 80, 246) marriage takes place between them .1 It-is_an...advantage to merhbers o f a . highly.-nio.bile „society. .tO-.he- ahle_to_jab.tain, hos pitality, from distant matrilineal..kin in ..remote..areas,„through w hich they m ay wish to pass-.on.theiii-travels. Eission .extends the geographical range o f each village!s ties_o£.durable kinship, ancLeach component, village in .a vicinage, -througbu.the frequent fission, o f -villag.es,. is. .interlinked—by. ..ties—of-com m om maternal descent..with-_many . viUages._in_other—vicinages. A m ong the Ndernbu, therefore, the lineage_principle_Qp.er.ates_.to mesh to gether in a single system distant and.discretejvullag-es-and vicinages 1 See, for example, Appendix III for marriages Village and its offshoots.
between
Nsang’anyi
V illage Fission , Slavery and Social C hange
177
and-to.p.CEmit.of..awidaraiig.a,Q£iiijcliyidual^iabiIity:— Am ong the Taliensi, the lincagc-principlc. operates.. to...flx„ atidxonsolidatc lineaLkirt in specific compact blocks o f land. But it Tmust not be thought that Ndem bu consciously and purposefully utilize the fission and dispersal o f villages to prom ote the closcErjcohesioii o f the w ider social-system,.,the...tribe, or. tribal section. - Independently o f their wishes. in..the..matter-it_may w ell be that:.centrifugal tendencies dependent on .their .ecological system-Constantly break up villages and change...the.. composition o f vicinages. It is m ore likely that .the cohesion. .of,the. wider system is maintained despite these tendencies, and.that Ndembu, in fact, make a virtue o f necessity. .Fission is usually...regarded by the people themselves as something deplorable and .they, take every possible measure to avert it. InJtheir conscious behaviour and judgments Ndem bu act and. spea.k_as _though_thc breakrup o f a village or its departure from a vicinage..was a.calamity, and the initiators o f fission are always condemned, n o t only by memb.ers^of the village from which they led away a-section but by members o f other vilIag.es_ inL_thc .vicinage. Again, if one asks the headman o f a dissident, settlement, w h y he-broke. away from the original village, he w ill most likely answer, that it was an unfortunate necessity. It was unfortunate ,that, the village should have split, he w ill say, but the headman .or on e.o f his close relatives was a quarrelsome person or-a sorcerer,-or-theheadman foiled to.Jbe impartial in his judgments, favouring his ow nprim ary kin in disputes w ith the seceding group, and so ..forth... ..Both groups, those w ho go and those w h o remain, w ill, in fact, stress the value o f village stability and continuity, the original group emphasizing the wickedness or fo lly o f the secessionists, and the latter the exceptional circumstances which compelled_them to leave*, But after a. lapse o f time, and especially after the deaths o f the leading _
178
Schism and Continuity in an A frican Society
tha_trthe_headman o f the seceding, group.was an eldec .with many close-kin and had little chance o f succeeding to office.in_an estab lished village. H o w this process operates..may_hest..be. illustrated by. a social, drama. The first o f these is based on accounts given to m e by members o f all three o f the groups involved. These groups are Mukanza V illage and its offshoots Kafumbu and Yim hwendi Villages. I use material from Mukanza once more on account o f its plentiful detail, because it provides an interesting comparison w ith the situations analysed in the tw o previous chapters, and because it illustrates tw o m ajor types o f village fission. I hope also that readers have begun to be acquainted with some o f its inhabitants. SO CIAL D R A M A V I Yimbwendi and Kafum bu Secede from M u kan za (compiled fro m informants)
In the period 1924-28 Mukanza Village passed through a crisis which almost resulted in its disintegration. A t that time there were three major factions in the village. The cleavage between nyachintang’a and malabu lineages was then only incipient, and both these lineages were united as nyachipendi (I,C2) lineage and as such were opposed to NYACHULA lineage (LD3, see Appendix I), o f the same genealogical depth and o f equivalent span The common ancestress o f all three lineages was Chipendipendi (I, A i), mother’s mother o f Kabonzu(I,Ci), the founder o f the village, but two inter vening ancestresses between Chipendipendi and Nyachula had become forgotten, nyachula lineage had returned to Mukanza Village .after many years’ residence in Chibwika area. In the village also was a large group o f hereditary slaves o f Kahali Webala (LE3), descended from a woman, Mpeza (I,F i), who had many daughters by two brothers o f nyachintang’ a lineage. The eldest daughter was married to her cross-cousin Kafumbu (I,G i), also a slave belonging to nyachintang’a lineage. Another daughter was married to Yimbwendi (I,F i i ), senior elder o f nyachula lineage. The village was a large one but full o f constant bickering. One o f the main reasons for this state o f tension as given me by the present inhabitants o f Mukanza, was that at this time there was no recognized headman. Kahali Chandenda (I,F5) had not yet settled permanently in Mukanza but spent his time alternating between his kin in Ikelenge Area and Mukanza Village. Mukanza Kabinda (I,F8) had inherited the name ‘ Mukanza ’ from his deceased elder brother Mukanza
11
o n
o o ©
o O
o
n
o o
o o o ©
©
o o o o o o © o €J
o
©
o o © © €1 €1
18 0
fV, /"
\ ■
( r*■ • c (
i : £
4\ i. ■
. f ■
S chism and C o n tin u ity in an A frican S ociety
Kandulu (I,F6) as an indication to all that he hoped to succeed to the headmanship. Yimbwendi (I,F i i ) and Kafumbu (I,G i) formed an alliance with their respective followings, and Yimbwendi strongly urged his own claims to office. When Kandulu died in 1919, Kafumbu had been suspected o f killing him by shooting his ilomba, in the same way as Sandombu had later been suspected o f shooting Kahali Chandenda’s ilomba (see Social Drama I, p. 95). Kasonda told me that Kafumbu had called in a sorcerer from Angola, named Mapachi, to assist him in this deed, and that Mukanza Kabinda had consulted a ‘ great diviner ’ with a divining-basket who had named Kafumbu as the sorcerer. I suspect that this tale is a myth retrospectively attributing evil qualities to Kafumbu in view o f his subsequent de fection from Mukanza Village. Although he was a slave (ndungu), Kafumbu was a powerful man in Mukanza Village. He was a famous hunter, a skilled orator, and had a large following in the children o f Mpeza (I,F i). He claimed that slavery (wudungu) had been abolished by the European Administration and that Kahali Chandenda and Mukanza Kabinda no longer had any rights over the services and allegiance o f their former slaves. He was constantly urging Yimb wendi (I,F i i ), whose two sisters had many adult children, to succeed to the headmanship. According to Mukanza Kabinda he knew that if Yimbwendi formally succeeded Mukanza Kandulu, the people o f nyachintang’a and malabu lineages would leave the village and found a new settlement, and the group o f slaves would be rid o f their masters. In n y a c h in t a n g ’ a lineage at this period there were a number o f adult men who, on the whole, supported the claims o f Kahali Chandenda (I,F5) and desired his permanent return to the village. These included the famous hunter Lupinda (I,G7), whose spirit Sandombu was later alleged to have invoked to bewitch Nyamwaha ; Tapalu (I,G8), later to become a wealthy African trader in Livingstone ; and Kachambili (I,G9), Sandombu’s brother who fled to Broken Hill in 1928 to escape from the sorcery o f Sandombu, according to some accounts, after a quarrel, and who has remained there ever since. Lupinda did not want to succeed, for he preferred the wild unanchored life o f a hunter. In addition, he was feared as a sorcerer by the villagers. In 1928 the simmering tensions in the village were brought to a head when Kahali Chandenda decided to take up permanent residence in Mukanza Village. This decision coincided with the decision o f a special gathering o f village elders to change the site o f the village to the southern edge o f the Kawiku Plain where the hunters o f the village would be nearer to their major hunting ground. When Kahali announced his intention, Yimbwendi said that he could no
V illage F ission , Slav ery and So cial C hange
181
longer remain in Mukanza Village as it was 4full o f troubles \ * people were not living well together*, and so on. (In fact the change o f site gave him a good opportunity to start a village o f his own.) He was now, he said, a man with many children and sisters’ children. He was an important elder (mukulumpi). He had been the headman o f a settlement before coming to Mukanza. The time had come to divide off (ku-baluka). He pointed to the fact that Kanyombu, younger brother o f Mukanza, had married his (Yimbwendi’s) own sister’s daughter Nyamalata (I,G23), who was Kanyombu's classificatory sister’s daughter, without performing the ritual to cleanse an incestuous union (kudisola), as an indication that the lineages o f nyachintang’a and nyachula were no longer closely related. He then asked Kafumbu to come with him as they had both married into the one lineage, that o f mpeza (I,Fx). Kafumbu agreed and said that he was weary o f the quarrelling in Mukanza Village. Kanyombu and Lupinda became very angry at this and said that, according to Lunda custom, Kafumbu and Mpeza’s descendants belonged to them and had no right to dispose o f their own future. Kafumbu said that a new law introduced by the British South Africa Company had *killed * the old custom and slavery had ceased to be. After some hot exchanges, fighting broke out in the village in the course o f winch Kachambili (i,Gs>) and Yimbwendi’s uterine nephew Samunuma (I,Ga2) were badly beaten up. Yimbwendi and Kafumbu declared that the fight had clearly demonstrated that there could be no peace between the followers o f Kahali and Mukanza, and the rest, and led their followers away to the far side o f the Kakula river about seven miles from Mukanza Village, and nearer to the Boma. Possibly Kafumbu, who had chosen this site, thought he would be safer near the Boma i f the Mukanza people tried to begin a feud. Not ail the followers o f Yimbwendi and Kafumbu accompanied them. Nyawunyumbi (I,G25), uterine niece o f Yimbwendi, remained at Mukanza Village, for the reason, as she told me, that Yimbwendi was a sorcerer, ‘ with a bad liver ’, and she had three daughters and a son whom she wished to protect from sorcery. Mpeza’s son, Kayineha (I,G<5), also remained. He was married to Kasonda’s sister Mwendiana (I,Gi4), daughter o f Nyamwaha (I,F7), was personally friendly with Mukanza Kabinda, and in fact played a major part in bringing the fighting to an end. Nyamalata (I,G23) remained with her husband Kanyombu (I,F9). Line (I,H i 7), on the other hand, went with Kafumbu, although he belonged to Malabo, because he was the cross-cousin o f Yimbwendi, and one o f Kafumbu’s sons was his personal friend. Later he married Nyawunyumbi’s daughter (I,Hi9) and returned with her to Mukanza. Thus, although three major groups were clearly involved in the cleavage, several individuals O
182
Schism and C ontinuity in an African Society
deserted from their own kinship factions and changed their allegiance. Such individuals, although acting apparently in accordance with their personal desires and interests, have an important role in reuniting the initially hostile factions at a later date. Each assumes an intercalary role in a wider social system, the units o f which are villages and vicinages. After the departure o f Yimbwendi and Kafumbu the District Commissioner o f that period wanted to combine Mukanza Village with Chibwakata and Nswanakudya Villages since the village had lost more than half its population. But an appeal was sent by Kahali (I,F5) to Ikelenge Area, to Sandombu and others living there to return and save the independence o f the village. A number o f them, in cluding Sandombu, and their wives returned and the dreaded amal gamation was avoided. In 1929, after one year, Kafumbu and Yimbwendi fell out over the headmanship o f the new; village ; there was a fight between their fol lowers, and Yimbwendi left Kafumbu with his children and sisters' children to settle in Chief Chibwika’s area, far to the south o f the District on the edge o f Mbwela country, where Kabonzu had once made a village about a century before, and where Yimbwendi had once had his own village, although in a different vicinage. When he died a few years ago his village split up once more after a quarrel between Samunuma (I,G22), son o f his oldest sister, and Makumela (I,G24), son o f a younger sister. Makumela, who had been nominated mulopu or heir by Yimbwendi, was a younger man than Samunuma, who hived off and made a farm o f his own with his children and sisters’ children. When Kafumbu died, his deceased sister’s son (I,Hi) claimed the headmanship but he was rejected by the over whelming majority o f villagers who chose Kafumbu’s son, Samlozang’a (I,H2), as headman. Samlozang’a was the senior elder of m peza lineage to which most o f the villagers belonged ; he had another claim in that he had 13 children, some o f them adults. Samlozang’a was supported by Mukanza himself, by this time reconciled with the village through nis friend Kayineha (I,G6). Kayineha, although mother’s uterine brother o f Samlozang’a (I,H2), did not succeed because he was working at the Copperbelt when Kafumbu died. Kafumbu’s uterine nephew (I,Hi) made a farm about half a mile from Kafumbu Village where he lives alone with his wives and children. His farm is called derisively by Mukanza and Kafumbu people * Kafumbu Kamu ’, meaning, * the solitary Kafumbu ’, while Samlozang’a’s village is called ‘ Kafumbu Kevulu ’, * Kafumbu o f many people *.
Village Fission, Slavery and Social Change
183
Commentary and Analysis W hen a long-established village splxts, it often, happens that the... seceding sections are .highly, unstable. T h e y ...lack-...longestablished bonds o f discrete .unity.-_and._they. often.break.. into small setdements- o f uterine siblings and their families,..each under the leadership., o f the., oldest , brother._.Gnce__the_strang bonds which hold together the members _of a traditional, village .„have been „broken,. the powerful .individualism o f the. JbJdemhu ethos, nurtured-by. the. selfrreliant.liie.of. the. hunter,, asserts,, itself for a time and-the.dissident group -breaks.-up-into. its. basic-units, the remnants , o f what were once .matricentric-Jfamilies... But the original village ...slowly .and painfully., .builds—up. _its_jnembership again. N dem bu take, great pride, in .belonging_tOL.a. village with a famous historical name, and although- such -villages are, like others, liable to fission, it is extremely__rare_.for„__them to perish utterly. Villages founded b y . the. twelve .headmen who, according to tradition, originally accompanied-Kanongesha, are still inexistence,.most..o£_them large.and-thri-ving,-as-indeed are villages., founded b y the .early chiefs-for-their-close relatives. The ancient K aw iku villages too, such as Nsang’anyi, Mukanza, Kasai and Nyachiu, have persisted for many years in spite o f frequent fission. The.-headmen o f. such ..villages, enjoy greater respect than the headmen, o f newly-form ed villages and farms, ana i f there is a gathering o f local headmen at a. ritual or at the hearing o f a case they are always given beer or food before the newer headmen. A few words must be said about the w ay in. which this irre parable breach o f intra-village relations has been recognized b y the parties concerned and how new social relations have been established between the villages. The people o f Mukanza Village say that both Yim bw endi and Kafumbu were sorcerers, and every death that subsequently occurred in their villages was laid at their door as p ro o f o f this statement. Kafum bu’s slave status, and that o f the mpeza lineage is also mentioned, and blame is attributed to the Europeans w ho caused skvery to be abolished, thus allow ing upstarts like Kafumbu to break away from their rightful owners (ankaka). T h ey also allege that Yim bw endi was a quarrelsome man, not the kind to inspire respect or confidence in his judgment. Yim bwendi and
184
Schism and Continuity in an A frican Society
Kafumbu people, on the other hand, assert that Mukanza Village contained many sorcerers and witches, mentioning Kahali, Lupinda, N yam uw anga and Sandombuin this connection. Mukanza Kabinda was never considered a devotee o f the black art, even by his enemies, but the dissident group say that at the time o f the split he was a poor speaker in village lawsuits and was not suffici ently firm in his decisions to give a village good guidance. Kahali Chandenda was not acceptable to them as headman on account o f his roaming propensities (ku-kimboka, * to move about from one place to another’), his irritability, and his suspected sorcery. B ut although there appears to have taken place a complete rupture o f social relations between the leading actors in this social drama, Mukanza Kabinda, Kanyom bu, Kafumbu and Yim bwendi, this did not occur between their several groups o f followers, many o f whom were interlinked b y complex ties o f kinship and affinity. Thus Sakazao’s senior wife, Nyaluwema (I,G4), was the youngest daughter o f Mpeza (I,Fi) by Chisela M alwa (I,Fa), brother o f the fourth headman, and she was also sister o f Kafumbu’s wife Katendi (I,Gz), so that Sakazao, accord ing to Ndembu kinship nomenclature, called Kafumbu ‘ older brother ’ (yayrt). Chinema (I,Gai), at that time headman o f the village in Ikelenge Area abandoned by Kahali Chandenda, was the husband o f Nyaluwema’s and Katendi’s other sister by the same father, Kawila (LG3), and, it w ill be remembered, Yim bw endi was the husband o f their other sister, Luseng’a Mulusa (I,G5), by a different father. Eventually die children o f all these men, w ho called one another ‘ older and younger brother would form the residential core o f the new unit founded by Kafumbu. This is probably the reason w h y Sakazao took no part in the fight which led to the secession o f Kafumbu and Yim bwendi from Mukanza. Kayineha (I,G6) o f the mpeza lineage was, it w ill be recalled, married to M ukanza’s sister’s daughter, and was Mukanza’s close friend, in spite o f being his classificatory son-in-law (muku, a term em ployed not only b y a woman’s father, but also by her mother’s brother, towards her husband). Ifciis._.nat. frequently.^enough stressed by anthropologists that village kin,., whatever, their genealogical relationship, .tend to develop, newuelationships o f friendship or hostility based on their constant interaction and on temperamental affinity or..animositywhich.jn..jthe_rcgular flow
f.Gi. K A F U M B U [,G2. KATE N D I— ¡.G ì- K A W IL A ---I,C4- N Y A L U W E M A -
l© t.Fi. M PEZA !
jO I.GJ. LUSENGA M ALUSA
V
LA I.Gó. KAYINEHA—— — i O
I.F?. N Y A M W A H A —■— O I.G14 - M W E N D IA N A -----» — ——
----------- — - ■■■&i ,G í 6. N y a m
............... -=A I.F11 . YIM B W E N D I-
a
Su
■¿\ I.G2I. CHINEM A
n c a
n
I.Hÿ. S A K A Z A O -
186
Schism and Continuity in an A frican Society
o f dh^..so-daL.process..gi¥e rise.to..behaviQur .which cancels out in many^situations the institutionalized^^ by kinships rules. T ims' Mukanza and K.ayiiieha._shQuId .have par ticipated.in a relationship which respect on the latter in his dealings w ith .the .fbrrner^ gave the former peremptory authority, over. the. lat.ter,„,..B„utin practice the two men. were, friends, visiting Angola together, where they obtained a new variety o f cassava which gave rich yields and which they introduced into Mwinilunga. They- slept-side by sideJh-temporary camps when custom decreed that they should never .:oeeupy.-die_.same._dwelling. In this case the fact that Mukanza and Kayineha were more or less contemporaries, and were brought up as adolescents in the same village, probably helped to mitigate the rigours o f genealogical and affinal regula tions. Kayineha did not leave Mukanza with the others but remained until the death o f his wife several years later. He finally departed from Mukanza on the best o f terms, and his friendship w ith that village probably had much to do with the rc-animation o f tics between the two settlements. Line (I,Hi7), on the other hand, went with Kafumbu and Yim bwendi and eventually married the daughter o f N yawunyum bi (I,G25), Yinibw endi’s sister, a woman called Maria (I,Hi9) with whom he eventually returned to Mukanza. N ya w u n yu m b i1 has re mained in Mukanza Village, and after a few years married the headman o f a nearby village, Mbimbi, son o f a male member o f malabu lineage in Mukanza and her classificatory cross-cousin. The children o f N yawunyum bi played an important role in inter linking the various sections within Mukanza Village, as well as in the rc-cstabJishmcnt o f relations w ith Yim bw endi’s new village. In addition to die daughter Maria (I,Hi9) w ho married Line o f malabu lineage, her oldest daughter, Seliya (I,H i 8), married Mukanza himself, o f nyachintang ’a lineage ; and her son Kenson (I,Hao) was brought up in Mukanza Village where he shared a bachelor hut with Pearson (I,Js), son o f Nyatioli and sister’s son o f Sakazao o f malabu lineage, and with Daudson (I,H7), son o f Manyosa o f nyachintang ’a lineage. Kenson is now affianced to his classificatory cross-cousin Mwendiana (II,Fi), daughter o f Kasonda o f malabu lineage. In 1953 he
1 Cf.
p . 18 1.
T 5!
V illage F ission » Slavery and S o cial C hange ;
• ;
r f
187
paid several visits to Yim bwendi village to buy meat from the hunters there for resale in the vicinage o f M ukanza. Mukanza and Kafumbu people regularly attended the others* rituals; but Yim bw endi people, although sometimes notified o f such events as the funerals o f important members o f these two villages, never attended, and vice versa. I have heard Mukanza wom en wailing in the traditional fashion w hen /news was brought to them o f the death o f an elder in Yim bw endi Village, but no representative was sent to the funeral ritual. O n the other hand, when Pearson o f M ukanza V illage was killed in a motor accident in 1954, many people came from Kafumbu Village for the interment and mourning rituals. Again, I have known large parties o f Mukanza people to g o to Kafumbu, in the vicinage o f w hich Headman Nyaluhana is the principal village headman (mwenimbu), when Kafum bu boys were initiated at the boys’ circumcision ritual at Nyaluhana, and when a Kafumbu girl’s puberty ritual was celebrated. I also attended the Chihamba ritual in 1954 at w hich a final reconciliation o f the two villages tacitly took place, and at w h ich I, an honorary member o f Mukanza Village, was given the adept-name o f Samlozang a Ndumba, also possessed b y the present headman o f Kafumbu. T he possession b y tw o persons o f die same name makes them majinda> * namesakes a mutual relationship o f mild jo kin g and mutual assistance w ith „goods and services. In conclusion, jh erefo re,.it ..might. be.,said„jdiat. aftexL.die. fission o f Mukanza Village, new relations offco rd iality,. expressed in the maintenance and resum priom of. connubial-and, ritual ties, were formed between Mukanza and Kafijmbu,^each o f which became points o f contact between two. discrete vicinages. But, on the other hand, w hile there has been a. marked, reduction, o f open animosity between M ukanza Village and„Yim bwendi, the great spatial distance set. between them, fu lly reflects- the con tinuing., tension in their reciprocal relationship, and-is perhaps a precaution against actual conflict.
Cl o
0 t . ■) o o
o n
o o
© Q
©
Cl Cl
Slavery and Social Change For the purposes o f this analysis, Social Drama V I m ay be said to have begun w ith the decision to change the site o f the village, an event w hich m ay precipitate breaches o f regular soda! relations which have been formed in a particular place. This
© f-i
188
Schism and Continuity in an A frican Society
decision coincided w ith the permanent return o f Kahali Chandenda to the village as headman, an event which profoundly modified the internal relations o f its members and precipitated a crisis. Yim bw endi and Kafumbu refused to accept Kahali as headman. For Yim bw endi had come to Mukanza Village with a large follow ing about ten years before these events, and he had hoped perhaps to succeed him self; while Kafum bu wished to become independent o f his traditional masters, the members o f n yach in tan g ’ a lineage. Redressive machinery was brought into action, in the form o f a special meeting o f the village’s leading men. B ut the aims o f the leaders o f the different groups were divergent. Yim bwendi and Kafumbu wished to resolve the crisis b y breaking away from Mukanza Village, i.e. b y irre parable breach o f intra-village relations. Kahali, Mukanza Kabinda, Kanyom bu, and Lupinda, the senior men o f nya chintang ’ a lineage, wanted to retain the lineages o f nyachula and mpeza in the village. I f Yim bwendi had confined himself to arguing a case for the secession o f his ow n follow ing there m ight have been no quarrel, at least no overt violence, especially since he and his kin had only been in Mukanza a short time. B u t when he openly urged Kafumbu to go w ith him, and when Kafum bu agreed to go, mentioning the discordant state o f Mukanza Village, they touched o ff anger which culminated in fighting. For at this point one o f the basic norms o f the tradi tional social system was threatened, that w hich governed the relations between master and slave. It was threatened not so much by the demands and pretensions o f individuals, but by major currents o f social change set in train b y British occupation and legislation. Kafumbu, Tike people everywhere and at all times, was merely utilizing new laws and norms to his personal advantage. T w enty years earlier he could not have made such a hid for independence. In the past runaway slaves were hunted b y all free Ndembu, and i f captured were returned to their owners w ho punished them severely. It was in the interest o f all N dem bu to catch escaped slaves, for i f they did not do so the system o f slave-ownership would break down. B ut slavery had been speedily abolished under British Government and noone dared to take, action against slaves w ho had emancipated themselves b y flight. W h at were the main features o f slavery am ong Ndembu
V illage F ission , Slavery and So cial C hange
189
before British occupation ? Village slavery was distinguished from commercial slavery. Village slaves were paid as fines to terminate blood feuds, to settle debts, to compensate for homicide, to pay a chief or senior headman whose poison-oracle had cleared a person accused o f sorcery, and to discharge fines for a number o f other offences. The principal mechanisms through which payments o f slaves were made were :— a ch ief’s c o u r t; a council o f elders or mahaku ; direct negotiation between the parties involved in a dispute ; or the invocation b y the wronged party o f some famous warrior such as Chipeng’e w h o defeated the Chokwe and Lwena slave raiders, to press a claim against a transgressor. T he persons paid as slaves were usually young children. O ften they were already o f slave status; but i f a man had to pay a relative, his choice usually fell on his ow n sister’s child. O n ly as a last resort w ould a man pay a junior sibling and in custom he had no rights over the disposal o f his ow n children in this manner. B ut in practice a man w ould resist the demands o f a brother-in-law w h o wanted to pay his children, the latter’s sister’s children, as compensation for debt or homicide. For example, the mother’s brother o f one o f m y informants went to m y informant’s father about thirty years ago in Angola and tried to take his daughter to p ay her as compensation for a large debt o f long-standing. B ut m y informant’s father became very angry and told his brother-in-law to rem ove him self from the village forthwith, saying, * I f yo u take one o f m y children, one day, perhaps soon, w e shall have a fight.’ His brother-in-law replied, ‘ Y o u have no cause to fight me. A re not these m y sister’s children ? I can do as I like w ith them.* The father retorted, * N o , they are m y children. I f yo u have got into debt because y o u love adultery and have had to pay many fines, perhaps yo u can pay your ow n children.’ His brother-in-law then said : ‘ Y o u are a bad brother-in-law (ishaku). Y o u have no right to keep your children from m e.’ But he was afraid to take the girl, and went away. In an egalitarian society like that o f the Ndem bu, slaves were not markedly exploited, and in fact were regarded more as rela tives than chattels. I f the slave was young and his or her master an elderly person, slave and owner initially stood to one another in the fictitious relationship o f grandchild-grandfather. T he term ttkaka in Lunda means both * grandparent ’ and * owner *.
190
Schism and C ontinuity in an A frican Society
This relationship provided a point o f departure from, which the slave could be assimilated into the kinship structure o f the village. -T h e owner’s sister’s son became his ^mother’s brother ’, the owner’s son his * father and so forth. Female slaves were often taken as concubines b y their owners and a new pattern o f kinship terms with corresponding rights and obligations developed from this relationship. T he children o f female slaves by free Ndem bu remained in their slave status, the children o f male slaves b y free mothers— in the case o f such marriages the slave husband had to pay bride-wealth— were regarded as free Ndembu and could go where they pleased. In other words, status was inherited matrilineally. Slaves were inherited in the same w a y as chattels, passing to the nearest matrilineal kinsman o f the deceased. Slave widows, but not the w idow s o f free Ndembu, could be inherited b y the brothers o f their dead hus bands. Sister’s sons also might inherit the slave widows o f their uterine uncles. In the course o f time, as in M alcanza Village, one finds in effect a large slave-lineage, like that o f mpeza , side by side w ith and owned b y a free lineage. O ften, indeed, the slave-lineage m ay be the larger, since most o f its members may marry within the village while the virilocal marriage o f female members o f the free lineage m ay reduce its effective membership within the village. I have been told o f powerful slave-lineages which, like cuckoos in the nest, ultimately ousted free lineages from the headmanship o f villages. Perhaps Kafum bu had cherished hopes o f doing likewise in Mukanza Village. Slaves had to w ork in the gardens o f their owners, even, in the case o f male slaves, after they had married, an act that normally betokens among free Ndembu the assumption o f adult status and economic independence. I f they became hunters they had to give the whole o f their kill, w ith the exception o f the sacred portions reserved for the hunter, to their owners. I f the owner o f a female slave did not wish to take her as his concubine he received bride-wealth for her from any man w h o wished to marry her, and her children belonged to him and to his heirs. A slave could, however, purchase his independence and that o f his relatives. I have recorded several examples o f this, which suggest that slaves were entitled to accumulate property o f their own. I give one case here b y w ay o f illustration. T he father o f Kasonda o f Mukanza Village, a man named
V illage F issio n , Slavery and Social C hange .
191
Kahumpu (II,D i), was an elder o f Shika Village, and his wife v? Nyamwaha (I,Fy) o f Mtikanza Village was living w ith him viri•locally. O ne day, he caught his w ife in delicto flagrante w ith a ' . man called Kalubinji, classificatory brother o f the headman o f Kasai Village, one o f the ancient K aw iku settlements. Kahumpu attacked Kalubinji and killed him w ith an axe. H e then ran ,; off and hid him self in the bush. N ew s was brought to Headman Kasai o f the murder and he gathered a force o f armed men from V his ow n and neighbouring K aw iku villages and came to the s< periphery o f Shika Village. Headman Nsang’anyi, senior head man o f the K aw iku and son o f a former headman o f Shika Village, came also to arbitrate between the groups. Kasai shouted from the bush outside the village on tw o successive days and demanded and obtained from Headman Shika Ikubi (II,C2), :i: a promise to pay tw o slaves, tw o guns, 10 yards o f cloth and a • conus mollusc shell (itnba), as compensation for his brother’s death. Kahumpu (II,Dx) returned from the bush o n the second day, and after pleading great provocation paid tw o slaves towards the total compensation, a young man o f about fifteen called Sakutoha and his sister, together w ith a gun. Kasai received these. According to Kasonda, the reason w h y a large amount o f compensation was made over to Kasai was that Shika and Kahumpu stood in great fear o f the recently established European, government (the murder took place in 1913) o f the British South Africa Com pany which had made murder a capital offence. Kasai had taken advantage o f the new situation to pitch his demands very high. It was tacitly agreed between the parties involved that no report w ould be made to the B om a i f the Shika people agreed to Kasai’s terms, which were, in fact, a form o f blackmail. But after compensation had been paid, Kahumpu as a precautionary measure fled to Balovale District for m any years. The compensatory slaves, Sakutoha and his sister, were allo cated b y Kasai to his classificatory mother’s brother, Kahona. They had com e originally from Nsang’anyi Village and had been paid as slaves to Kahumpu’s mother’s brother, the late headman o f Shika. B ut although they had thus been returned to a K aw iku village sprung from Nsang’anyi, they still retained their status as slaves. Every little Ndem bu village, as long as it endures, m ay almost be likened to a sovereign and independent
192
Schism and Continuity in an A frican Society
state. It was immaterial whether the slaves were Kawiku or not, or whether they came from Nsang’anyi Village or not. This did not give them freedom. Nsang’anyi and Kasai were independent villages, each had its own business to mind. In the late twenties Sakutoha purchased his freedom from Kahona with a muzzle-loading gun. A few years later he tried to obtain his sister, w ho had borne Kahona several children, by the offer o f another gun. Kahona did not wish to lose his con cubine and children, the major part o f his follow ing in Kasai Village ; he took the gun, feigning that he would release them to Sakutoha in a short while. Sakutoha, w ho had returned to Shika Village, was planning to break away and found a settlement o f his ow n w ith as many o f his siblings as he could muster. But Kahona then claimed that although Sakutoha had purchased the enfranchisement o f his sister she still wished to remain w ith him, and he offered Sakutoha bride-wealth to legalize the marriage. He knew that Sakutoha could not prefer a charge against him in a chief’s court, for the Boma might get to hear o f it, and the British authorities might realize that slavery was still an active institution among Ndembu. This would mean much inter ference by Government in village affairs and possibly imprison ment for both disputants in the case. Sakutoha, w hom I knew well, was a simple and straightforward man, traditional in the extreme, and he accepted the bride-wealth, although condemn ing in round, terms Kahona’s double-dealing. But his grudge against Kahona and his family did not die out, for when Set, Kahona’s scapegrace son by Sakutoha’s sister, w ho had been expelled from the Copperbelt for persistent thieving, was re fused admittance to Kasai Village by the headman in 1953» Sakutoha w ould not allow him to build in his farm near Mukanza Village. Tim... case-history ..shows, the. tenacity w ith which Ndembu clung to the institution o f slavery long after . British .rule had overthrown or greatly modified major-aspects-of-the- traditional political system. It also illustrates how. Ndembu,-both.slave and free, played o ff the new laws against the old .customs.-to further their, ow n interests. It also reveals that slaves, although they might be inherited or transferred like chattels and were compelled to remain in their
V illage F issio n , Slavery and So cial C hange
193
owners’ villages, nevertheless enjoyed certain rights. Sakutoha, for instance, had been able to obtain a muzzle-loading gun o f his own w ith which he had bought his freedom. The status o f slave could not have been very onerous, for Kahona’s slave woman refused to leave him to jo in her brother, even though Sakutoha had obtained her liberty and guaranteed her an important position as the potential founder o f a village matrilineage. If_many..Ndembu.continued..torecQgm.ze_sIavery_a&.an.impor“ tant institution in the thirties and. Looked„askance.„Qmihose who had. resort to European-supported laws against.it,, feeling.,against those-who refused to conform to its time-honour.ed_norms.must indeed .have been strong in. the twenties,. -when-Kafrimbu. (I,Gi) made it clear that..he..\w:anted_.to._take__advantage__o£_the new ordinance. W h en Kafumbu and his followers joined, the fight between -Yimbwendi’s and Kahali’s parties, this act-represented a violation, of. the. norms governing relations betwe.en_slav.es and their_owners._ „ I n .the pasJLslaves. were.expected,.to..frght,.for their masters 4/in this squabble the slaves fought against them. For merly,, the sanctions that w ould have been...im.p.osed_against such recalcitrant slaves might have included die .execution o f their,, ringleader and the_suspension. o£..the, rest b y their necks from .thedforks o f tall trees. I f they had imposed these penalties, the people o f Mukanza would have been supportecL b y the other villages in their vicinage ; and C h ie f Kanongesha himself, i f called upon, would have dispatched a_punitive force from.his own vicinage, i f Kafum bu and. his group hadofferedresistance.— But the alien European authority was .known, to possess overwhelm ing force in the form o f armed police to. back np. its__laws and render_nugatory attempts to uphold the norms o f the.traditional political system. In short, the mechanisms which formerly maintained the norms governing the relations o f slave-owners and slaves could no longer operate, although m any Ndembu o f both categories still adhered to those, norms. M oreover new mechanisms had been introduced b y the British w hich applied sanctions against Ndem bu w h o were c a u g h t.attempting to maintain the institution o f slavery, mechanisms o f a legal charac ter designed to enforce conform ity to norms whichjwere .opposed to those o f the indigenous-society. Since there was no. way, therefore, o f redressing a breach o f traditional relations between slave-owners and slaves by traditional . machinery,, schism
194
Schism and Continuity in an African Society
beCwem. mcmbers o£-die. tw.o categories within a. single vi Wage resulted in^-the-fission of-that village. Yim bwendi, on the other hand, was not condemned by the people o f M ukanza £br withdrawing his own lineage from the village so much as for conniving w ith Kafumbu at the secession o f the slave lineage o f mpeza (I,Fi ) from their hereditary owners, the people o f nyachintang ’a lineage. Yim bw endi’s group, after all, had previously split o ff from Kahali Village and could no longer trace precise genealogical relationship w ith either nyachintang ’ a or malabu lineages, although they still remem bered to what genealogical generation they belonged with reference to the founding ancestor Kabonzu (I,C i). The lineage o f nyach ula (I,D3) was sufficiently distinct from the combina tion o f nyachxntang ’ a (I,D i ) and malabu (I,D z ) lineages to constitute a virtually autonomous group. B ut it seems that Yim bwendi had at one time cherished ambitions to become headman o f Mukanza and that he was encouraged in them by Kafumbu, even i f this would have entailed the secession fronvthe village o f most o f the members o f the other tw o lineages. But he had not been successful in winning the support o f any o f die leading men o f the established lineages without which he could not have become headman o f Mukanza Village. He had to content him self w ith securing the temporary allegiance o f Kafumbu. This was an uncertain asset, for Kafumbu was also an ambitious man, w ho wanted to use his relationship with Yim bwendi as a lever w ith which to secure his ow n emancipa tion and that o f mpeza ’ s lineage, his personal follow ing, from nyachintang ’ a lineage. Later, when the tw o men founded their setdement, Kafumbu, whose follow ing at that time out numbered Yim bw endi’s, claimed the headmanship o f the new village and forced Yim bwendi, w ho naturally did not agree to this, to leave the area. W h y were ritual mechanisms not brought into operation to resolve the conflicts In Mukanza Village, as in the case o f Social Drama II in which Sandombu was reintegrated into the Mukanza community ? The answer probably lies in the nature o f the social relationships involved in the respective dramas. San dombu was a close kinsman o f the principal actors in the earlier social drama and an almost life-long member o f the residential group. Kafum bu and his faction were slaves, and breaches o f
Village Fission, Slavery and Social Change
a a
195
0 owner-slave relations w ere traditionally settled b y legal means supported b y the sanction o f organized force. Y im bw en di and his lineage kin w ere indeed matrilinea! relatives o f the M ilitanza ; -group, but on ly distant relatives, so distant in fact that tw o marriages had taken place between members o f the respective groups which w o u ld have been considered incestuous had kin ship nomenclature alone been the criterion. B efore the split, Kanyombu (I,F 9 ), o f n y a c h in ta n g ’ a lineage, had married Nyamalata (1,623), ° f n y a c h u ia lineage, w hom he called ‘ sister’s daughter ’ ; and after the split Line (I,H i7 ), o f malabu ; lineage, had m arried M aria (I,H i9 ), o f n y a c h u la lineage, w hom k; he called * sister * b y one means o f reckoning kinship. Th e sole if common ancestress o f these m en and w om en was Chipendipendi (I,A i), mother’ s m other o f Kabonzu (I,C i), the founder o f the village that was subsequently known as Kahali, then as Mukanza. Chipendipendi was in the fifth genealogical generation before Kanyombu, the sixth before Nyamalata, and the seventh before Line and M aria, so that the relationship between these spouses was extrem ely distant. Yim bw endi and his group had lived separately from the rest o f the village for about tw enty years before they returned after Mukanza Kandulu’s death, so that ties o f spatial propinquity between the tw o sections had lapsed and had not been strongly revived in the short interval o f nine years before the final split. It w auld seerxis, therefore, , and this view can.be.supported by
eviden<^-foom„.QfoeiL..vill^gesr...that.. _ritual.. mechanism sare _only brougham.-tO-remtegrate.a~setdementgroup_o£-rdativeXy_close linealJdn w ith a. genealogical depth, o f about three-generations at m ostbetween founding-ancestorand the. oldest, living-members o f the lineage,— w hat ..I call a 4minor-lineage -__RituaL mechan isms are only brought .into operation,-moreover,_where_ties o f lineage.Jdnship-.are .supported b y . a-long--tradition -o£-joint resi dence be£weeELthepersons-.aiid-Subrliiieage&.concerned. More over, as a rule only a single type o f ritual is utilized primarily for this purpose, the ku-sw anika ijina ritual, in the course o f which a muyombu tree is planted to a matrilinea! ancestor or ancestress o f a village member w ho inherits on that occasion his or her name (ku-swanika , to cause to inherit or succeed to, ijina, a name). M iastjother xituals,. aithough-.they. m ay,Jiave the subordinate^function _-of reintegrating, the -ritual subject's.-village,
a
0 a © ■
0
0 a
G © © © © © © © © -
©
©
© ©
© ; © :
©
;
© ©
!
i"
O ©
© ■ M
0 n
D o o
o-
o o o C)
o c o o O f) Cl
o c
C
O .
G
c
Q 0 Q
1 u ij G t)
Ac
T96
Schism and C o n tin u ity in an A frica n S o ciety
dominantly emphasize the unity o£alLNdem bu,-or- even- o f all Lunda. But the lineage group which is reunited b y the ku~ swanika ijina ritual is the effective local lineage. Marriages between members o f this lineage other than between classificatory grandparents and grandchildren are reckoned to be incestuous unions (chiinahvamalwa or chipikapika), and a ritual (ku-sola, or ku-disola) is performed to cleanse the couple i f they have a common great-grandmother. W ithin closer degrees o f matrilineal kinship than this, in pre-European times the couple might have been permanently ostracized or put to death. In Mukanza Village in 1954, when I left the field, the groups predominantly concerned in ku-su/anika ijina rituals were tend ing to become, not the mayor lineage o f nyachipendi (I,C2), the mother o f Malabu (I,D2) and Nyachintang’a (I,D i) ; but the separate minor lineages descended from these daughters (see M ap 4). The children o f Manyosa’s, Chaw utongYs and Kasonda’s generation, and the grandchildren o f N yam ukola’s, N yatioli’s and Sakazao’ s generation, when they g ro w up, will no longer constitute together an effective local lineage, and ritual mechanisms w ill no longer be able to reintegrate the whole village i f there is a breach o f regular social relations between the tw o sub-lineages. In any case, new social forces which are now rapidly developing w ill almost certainly accelerate fission o f the larger villages into ever smaller residential units. So.cialDramaVI.welLillus.trates,thetype-o£fission_which,tended to occu£_iii_the_first -pexiad_Qf..sociai_change— Traditional ties became more brittle ; traditionaLsocial-machinery-for-reintegrating/ a . disturbed _group....became. .replaced., in.. scv.eral„situations b y the'' legal machinery o f the. ..superoxdinate_aliem_authority ; new -norms regulating behaviour were .introduced from above ; or new norms began to develop on the basis_offnascent social relationships o f a new type and ..came. .into—confhct^with old noons. But thdTirsL.efiect,o£these.imiovations^tended.to-be the loos.emng..jQf„traditionaL-ties^-rather—tham-their—replacemenc. by new„ones. The
V illage Fissio n , Slav ery and So cial C hange
197
the-people ;... and. although, in private conversation... a. certain stigma-" is attached to membership o f such- .a-viUage._by free Ndembu)-the-viUageso£slave-origu^are_treated--ijxmQst^respects in diG-samc-.way.as.viHages-ofifiee.-men....—Eojxner-.slav:es and free l^embM-inteinnarry. freely; andtfie^ ch^ slave w om en, are. regarded- asJxee. :Accretion The role o f n yach u la lineage in the situation illustrates not - only one type o f fission, but also accretion as it operates among , the Ndembu. W h en Yim bw endi returned to M ukanza Village after tw enty years he was utilizing ties o f matrilineal kinship which had n o t yet fallen into complete abeyance. His ow n uterine uncle Saluvaji (I,E6), the father o f Line, moreover, had remained in M ukanza Village after (he first secession o f the n yach u ia lineage from Kahali W ebala’s village, and acted as a living link b y means o f w hich the lineage could be attached to the village once more. In most N dem bu villages one finds a few scattered classificatory matrilineal kinsmen o f the headmen, members o f sub-lineages that have split of£ hut w h o have re mained perhaps on account o f marital ties, as in Saluvaji’s case, with the principal village lineages, perhaps because they have fallen out w ith their o w n primary matrilineal kin. These per sons, for example, N yaw unyum bi (LG25) and her children in Mukanza Village, in practice seem to assume the function o f maintaining connection between the dissident and remaining sections o f a divided village and to provide points o f re-entry for members o f the seceding group. In other words they have important roles in the system o f inter-village and inter-vicinage relations, or from the point o f vie w o f members o f a village in their system o f external relations. Fission is seldom along abso lutely clear-cut lines. Ties other than those o f lineage affiliation may temporarily or even permanently prove stronger than the latter so that each core o f primary matrilineal kin in a village may have a number o f classificatory encrustations. Such per sons must not be regarded simply as social isolates withm a v illa g e ; they have functionally significant roles in the under system. W h en Kayineha (I,G6), for instance, was living at Mukanza Village he held the door open, as it were, for a reconciliation w ith Kafumbu.
198
Schism and C ontinuity in an A frican Society
Lineage or Uterine Sibling Group ? In effect, the groups w hich split o ff from Mukanza Village were groups o f uterine siblings. Kafumbu managed to gain control over tw o o f the daughters o f Mpeza (I,Fi), and these, together w ith their children, formed the core o f his new village. Kayineha should have been the leader o f this group but his attempts to persuade his sisters to remain in Mukanza Village, where he had married, were resisted b y them. Kafum bu had many children b y Katendi (I,G2), and Katendi herself had con siderable influence as the oldest sister and a strong personality in her ow n right over K aw ila (I,G3) w ho also had m any children. Here the powerful bonds between uterine siblings and between members o f a matricentric fam ily were exploited b y the husband o f the oldest sister in his o w n interests. The success achieved by Kafumbu (I,G i) against Kayineha (I,G6) in the struggle to control the all-important fruitful wom en is one case in which a husband defeated a brother. M ore often in such a struggle between brothers-in-law the brother wins ; but the outcome in Ndem bu society is never a foregone conclusion. Undoubtedly the fact that Kafum bu was a cross-cousin o f his w ife and her siblings was o f great assistance to him. Yim bw endi’s follow ing consisted o f his ow n group o f uterine sisters, and their and his ow n children. B ut his youngest sister died ; and her daughter Nyawunyum bi (I,G25) suspecting Yim hwendi o f having killed her mother by sorcery1 to enhance his hunting powers, refused to leave Mukanza w ith him. His uterine uncle Saluvaji (I,E6), w ho also remained behind, was a feeble old man w h o had strong bonds o f marriage and friendship with Mukanza Village and who, besides, would not accept the authority o f his own. sister’s son. B ut Yim bw endi’s other tw o sisters and their children remained loyal to him. Later when Samuntima (I,G22) split o ff from Yim bw endi’s village after Yim bw endi’s death, he led aw ay w ith him his own uterine sibling group, leaving Makumela (I,G24), his maternal parallel-cousin, w ith Makumela*s ow n sibling group. The grpup ^ f uterine siblmgs ..under foe ^leadership o f the oldest or most_capable_Jbxo.foeii-is_the„.basic_Ndembu„ceridenrial_unit and 1 Cf. pp. i8i, l 8<5.
T 3 @ •jv tifa
V illage F ission , Slav ery and So cial (Change
199
the most com m on nucleus o f a new village... -Happy. is the ambitious man .who. has .many .5is.ters and..unambitioiisLy.ounger brothers w ith children o f their ow n. . Sandom.bu.-is. anuexample o f a m am who is not favoured w ith fertile siblings^_ond,who_iias no children o f his ow n. H e must resort tp discreditable-means to scrape, together w hat follow ing he can,, and. is. always regarded w itLsuspicion. This .nuclear..sibling., group ..has always, been important as the pioneer element i n ..foundmg_a_jxew._.village, according to m y genealogical material and according..to..infor mation. 'collected from m y older informants.. B.ut_in._the..past, say the latter, fission was much less..frequent,. mdirL.ad.ditioiL.to his own ,
o o 0
0
©
©
©
©
©
! © ! © Ì © ! © © © ©
I© i ©
I© l @ © © © © ©
200
o
j
1 (") s.\ i. /
Schism and Continuity in an A frican Society
r e lu c t a n c e o f t h e . d e f e a t e d c la im a n t . to _ s .u b in it_ t< X - th n „ a u th Q r ity o f the-successful one,, were frequent causes .offission— .The.ahsen.ce o f a strong, centralized political authority capable, .of. compelling people to remain in a village after .conflict .had occurred..„also favoured village schisms. B ut once a village. had.-b.ecame_established .great efforts were usually made,„as in JMukanzaJ^ji ilage, to p r e v e n t its L .d is m p tio n .„ a n d _ t.Q _ Jc e e p . i t a l i v e .
c'"\
V :
o f'\ V
Q O
j j
O f) (")
| I j i
C 1; *vi ...";
j
| j '
€• • €■ y
Personal Factors Much...depended .„on„the._.personality.._Q£_the.._headman, the supreme agent o f the transition from_uterine...sibling_jallage to matrilineal village, i f the group was.to . remain._undiminished through the period o f instability.which_super.venecL„when the proximal generation .to. the village .founders. _h.egarL_.to str_uggle amQiig-.themselvesfor_p_ower__Lfoimdjthat.siiccessfuLheadmen among Ndembu _were .firni.-hutmnahtrusive_.pers.onaIijtie.s,.jjiiaggressive, and. ready to . share ,whatjwealth...they jnight._acquire withjtheir relatives. Scrupulous, fairness is reckoned -to...begone of_their most essential characteristics. B ut .the..best .notion o f whatJSfdembu consider the ideal personality type, fbr..headmen is afforded b y the follow ing-text I collected-about-Headman N g ’arnbi during a discussion w ith somemembers-of-the vicinage to which both N g ’ombi and Mukanza -Villages belong : Headman N g’ombi, even before he succeeded to headmanship, used to laugh with everyone, elders and children alike. He liked both men and women. If Ms younger brother o f Ms own lineage or o f a senior or junior lineage, or Ms sister*s son, or his father were hungry he would remember to offer them food. If they are very old (adinawevu, literally, *those who have a beard ’) he helps them (wayikwashang a, literally *he helps them all the time ’ : ku-kwasha means * to help in every way— with goods and services, to be solicitous to their every need ’) ; i f they are sick he helps them ; i f they have come from another village he helps them with food, with beer, and gives them a place in wMch to sleep. And even i f the Europeans send their messenger he will help Mm with one thing or another that he may require. If CMef Kanongesha is coming to call a meeting o f headmen he will bring a calabash o f beer or a sheep or cassava meal as tribute. CMef Kanongesha is very pleased and says * You have brought me tribute as a village headman; that is a good thing to do.* He tells him ‘ The one who helps me in our village is a son in. every respect.’ Headman N g’ombi has a good liver, he helps many people,
;■!
-j -J d
■ ~i; i
i \
’r; •'
^ J, ^ y y
V illage Fission , S lav ery and So cial C hange
201
he makes gardens for many people (i.e. to give them hospitality), he digs these gardens, he has strength, he has ability to argue cases, {waheta wuhaku), he does not steal nor tell lies, he does not slander people, he is not a sorcerer. B y w ay o f contrast here is a text describing the personality o f a headman w h o was foared and disliked : He is a bad man, selfish (uraheta chifwa), proud, quarrelsome, given to reviling people, lazy, a liar, without skill in legal argument, a thief, lustful, a slanderer and a sorcerer. He is wqfwaha walu/a, a person who scrounges beer from everyone without making a return. Such a erson who takes without returning is like a mad dog. A man who as self-respect (kavumbi)1 does not do so. He also respects other people, without regarding them as useless.
E
Ndembu_do not admire overweening^_domineering^hea.dbtiien ; men w ith such temperamentS-are condemnecLaiictaithough they not iiifeequendy.succeedtG.office^eldGm..acquirer Jarge.ibllow ing. B om Yim bw endi and Kafum bu appear to have been men o f this type. M ukanza possessed many o f the qualities attributed to N g ’ombi, although he was not such a good arguer o f cases. Kasonda once told me that m any headmen eat privately in their own kitchens, sending w hat is left o f their food to the chota to be eaten b y the dkwachota (the group o f male villagers, regard less o f lineage or generation affiliations). B u t Mukanza, N g ’ombi, Kam aw u and M ulila (other respected headmen in the vicinage) preferred to eat in the chota w ith * their juniors and children \ T h ey were ridiculed for their commensaKty by such headmen as Nswanakudya, Chibwakata and Nsang’anyi, but the people said that this behaviour proved that they were not sorcerers (aloft), since it is believed that sorcerers share their food with their creatures o f sorcery, such as malomba and tutotoji. O nly chiefs are expected to eat alone or with their children, apart from their people. O n account o f Chibwakata’s pnde (iwitiyi), argumentativeness and reputed sorcery, four groups, led respectively b y N g ’om bi’s uterine uncle the founder o f the village, N yakapwipu, Sayifon and M akayi, divided o ff (kubaluka, literally * to split a piece o f w ood *) from Ciiibwakata. 1 C . M . N . "White (personal com m unication) writes that * havumbi means “ respect fo r others, n ot self-respect ** \ M y o w n informants, how ever, gave m e the m eaning ‘ self-respect \ ‘ care to bé thou ght a good man \
202 j | j
Schism and Continuity in an A frican Society
The_good headman.is_the..go.od.£eUo_w,_the man. .who.. * laughs w ith .everyone ’, w ho is hospitable, ..selferespecring,...helpful and democratic. Perhaps the most significant section o f the com mentary on Headman N g ’ombi is that in which he is described as a man w ho would give food not only to * a younger brother * o f his ow n lineage but also to members o f other lineages in his village. His field o f friendly co-activity is not circumscribed by narrow minimal lineage relations ; it extends outwards to include everyone o f the village, regardless o f their precise degree o f relationship to him. The headman in his person should typify and exemplify the most general norms governing social inter action within the village. His generosity and range o f interests should not be confined to his ow n village ; he should offer hospitality to visitors from other villages and should be able to argue cases in other villages w ith freedom from prejudice. Towards the agents and representatives o f external authority, indigenous and alien, he should be courteous and respectful so that his village should be w ell thought o f and he and his people should not incur the rancour o f capricious officialdom. Am ong Ndembu, however,...it.is not. easy..to ..fiad.m en who measure up to these exacting standards o f generosity.,, impartiality, and-unselfishness. For txadhrion. and_training-uphold .and pro duce-among the majority of.m en. a cast..of.personality.in many w ays opposed to the ideal ty p e .o f.headman. This..i$ the ideal personality o f the hunter,_The__professional hunter., (chiyang'a) is a_man o f the.bush who. spends.much-ofhis.life..alone, pitting his-wits, against the.fieetness or-£erocityL.o£axiimals^-fierce himself, boastful, dreaded.as..a_sorcerer, b y .character ..and_ra.ode_jO.£living unable-to-stomach...the. .authority .o f another,. patriarchaLxn his family, life in a matrilineai society,. a. wanderer.who. m ay travel from M wantiyanvwa’s country to Ishinde’s chiefdom in Balovale District in quest o f game, and given to long and frequent drinking bouts round the calabashes, o f beer..madedfeom .the. honey which he fearlessly collects from .the nests.Qf.w ild bees, in.tbe. woodland. A prayer made by hunters at the beginning o f their Wuyanga ritual defines the type o f man that is meant b y * a professional hunter ’ (chiyang a). * W e w ant a man w ho can sleep w ith ten wom en in one day, a great thief o f a hunter.’ * T h ie f’ actually refers to the ‘ theft * o f game from the bush, not to the stealing o f property, but the bias o f adulation for what in ordinary life
V illage F ission , Slav ery and So cial C hange
203
would be illegal acquisition is plainly evident. .Again, a head- man should not be licentious ; in a hunter sexual..potencyr-^with a fine.-.disregard o f its lawful direction—is-reckoned a -strong .. qualification. It is clear that such a personality type is diametric; aily o p p o sed .to that, thought..proper_for_a.headm an. Few Ndembu..succeea-in..achiev3ng..a workable compromise between these ideals. M any professional, hunters are headmen, .but few o f them, are successful headmen ; many successful headmen.w ith , large followings are not professional hunters. B o th Kafumbu and Y im bw endi were hunters, and acted as hunters, but not headmen, were supposed to act. O ne o f the reasons w h y Yim bwendi w ent south into C hibw ika Area was because there . was much gam e there. Although Mukanza was him self a -hunter he was not a great and dedicated hunter, and spent most o f his time in and around his village. 1 Because fe w men possess or develop the personality ideally 'required for headmen new settlements o fte n .fail to b e c o m e ¡ established. T he spine o f continuity w hich maintains. aJlongestablished village is provided .by its. famous historical.name or by the mere, fact that it has persisted fromJong.-aga~and~has.-thus acquired-the prestige o f longevity..- Where...these-footers are absent, w hen the founder o£ a_village dies, fissionjojften..occurs between uterine sibling groups sprung from different, mothers. -Thus Yim bw endi Village split into the groups led b y Makumela and b y Samunuma, and Kafum bu into those o f Samlozang’a and Kafumbu K am u. T he _.name o f the founder becomes extinct in many „.cases, and new discontinuous settlements., come into existence.
r\
V . .■
o
o C H A P T E R VH ( .: ( .) (.'/ O
o i. o o
o
ë '■ £ .
o
4' : 4 • £
■
I .. M..S
4-
V A R IETIE S O F V ILLA G E FISSIO N
N order to demonstrate that the tendency for groups o f uterine é siblings to split ofF under the oldest or ablest brother is no unique or exceptional phenomenon, I present Tables XVII, 4 X V III and X 3X compiled from data obtained from village genealogies collected in four Government chiefdoms and from discussion w ith the informants w ho supplied the genealogical material. Table X V II shows the relationship o f the leader o f the seceding group to the headman, whether Hassificatory or descriptive, and Table X VIII shows the span o f the maternal descent group w hich formed the nucleus o f his new settlement. I restate m y use o f terms : * uterine sibling grouj^ * refers to the • children o f a single mother ; * minimal lineage^ to the matri linea! descendants o f a single grandmother ; * minor lineage * to the matrilinea! descendants o f a single great-grandmother ; and ‘ major lineage * to the matrilinea! descendants o f a single greatgreat-grandmother. I found no lineages o f greater span than the * minor lineage * as nuclei o f new settlements. Lastly, I give in Table X IX the relationship o f the leader o f the seceding group to the senior wom an o f his new settlement, in order to bring out the feet that there is often a struggle between brothers-in-law for the control over a wom an and her children, and sometimes over her younger sisters as well, as in the case o f the rivalry between K afem bu (I,Gi) and Kayineha (I,G6) described in the previous chapter. Table XVXI shows that 18 out o f 33, or roughly h a lf o f the leaders o f seceding groups were sisters* sons, primary or classificatory, o f the headm en o f the original villages. O f the 11 classifica» tory sisters’ sons, 8 belonged to the same minor lineage as the headmen but to different minimal lineages within it, 2 belonged to the same major lineage as the headman but to different minor lineages within it, while one, Kafumbu, was more distantly related to the headman, KahaJi Chandenda. It has already been pointed out that the secession o f sisters’ sons, both primary and classificatory, is often associated w ith the tendency towards adelphic suc cession in Ndem bu villages (Chapter Three, p. 87). O nly one
I
204
o I)
V arieties o f V illage Fission
205
iS fe- uterine brother seceded from his older brother’s village, and in SMhl this case he made a small farm w ith his tw o sisters and his and their VILLAGE FISSION TABLE XVII R ela t io n sh ip s B
etw een
L e a d er s
o f S e c e d in g G r o u p s H eadm en
Classifie.
Relationship
Primary
z S ..................... YB . . . . S ..................... Cross-cousin . mB . . . . Not known .
7
it
18
2 2 i
4 I I I
5 3 3
1
P r ev io u s
Total
2 2
Total number o f seceding groups ' t.'ty í'
a n d t h e ir
• 33
V ' ' j ’■--
TABLE XIX TABLE XVIII
R e l a t io n s h ip o f L e a d e r o f S e c e d in g G r o u p t o S en io r W o m a n o f S e c e d in g G ro up
L in e a g e S p a n o f N u c l e a r S e c e d in g G r o u p s Category o f nuclear group
Î 'S-;
„¡SÇ .V \ •
Number of groups
Uterine sibling group . . Minimal lineage Minor lineage . Total
.
.
.
Relationship
OB YB S H Class. B
21 XI
i
Total .
33
Number
. . . .
. . . .
8 ii 6
.
.
33
5 3
legend
'/ :
, ^ c
>3
O B Older brother Y B Younger brother S Son H Husband
Class. B z$ mB
Classificatory brother Sister’s son Mother’s brother
children about a hundred yards from his original village. O f the 4 classificatory brothers w h o hived off, one belonged to the same major, but to a different m inor lineage as the headman ; one
206
Schism and Continuity in an A frican Society
to the same minor, but to a different minimal lineage ; and two to different uterine families in the same minimal lineage. In both cases where classificatory mothers’ brothers split off* they belonged to the same major lineage but to different minor lineages within it. Abundant confirmation is offered b y Table X VIH for the view that the dominant nuclear unit o f a new settlement is the uterine sibling group. In 21 out o f 33 new settlements the nuclear group consisted o f uterine siblings ; in 11, o f a minimal lineage ; and in only one case o f a minor lineage. Secession o f a minor lineage occurred in one o f the oldest villages in the N dem bu region, M wanta waLuunda in Nyakaseya area, the headman o f which used to have the office o f supervising the caravan which took the tribute that was irregularly despatched from C h ie f Kanongesha to M wantiyanvwa. The classificatory sister’s son o f the headman built a new settlement quite close to that o f M wanta waLuunda Village and continued to pay tax through the latter. The uterine uncle o f this leader o f the seceding group had previously been headman.1 In die majority o f cases in w hich a primary sister’s son founded a new settlement this was built at a considerable distance from the original village. W h y this should be so I do not know , unless it is because conflicts between close kin o f proximal generations when they occur are liable to be more violent than between distant kin, ow ing to the greater intensity o f their previous interaction. This theme o f the spatial dispersion o f close lineal kin after fission in a village pervades Ndembu society just as the theme o f lineal ramification in a single locality pervades Tallensi society. I f w e examine the spatial arrangement o f almost any vicinage in the Ndem bu region w e shall find few adjacent villages indeed that are linked b y close lineal ties. Close lineal kin do not seem to make good neighbours i f their leaders belong to adjacent genealogical generations. O n the other hand, when uterine brothers or parallel cousins o f the same minimal lineage divide, the new settlement m ay often be built adjoining the old. Thus, when Makayi, younger brother o f Headman Chibwakata, founded a farm in 1952, he built about a hundred yards from his brother’s village ; when Sakanya, parallel cousin o f Headman 1 Cf. the case o f Sandombu in Chapter Four.
Varieties o f V illage Fission
207
Machamba, made a new settlement he built about fifty yards from Machamba Village, and so on. W h en ties o f genealogical generation affiliation are combined w ith lineal connections they „ seem to act against the breaking o f links o f spatial propinquity, r W hen, on the other hand, conflict breaks out between close : lineal kin o f adjacent, and hence competitive, genealogical generations, ties o f spatial propinquity are abruptly and irremedi ably severed ; and the seceding group moves out o f the vicinage. Indeed, frequently it moves out o f the senior headman’s area to v which the original village belongs, an area which m ay include several vicinages. In seven out o f eight cases where fission took place between groups headed respectively b y primary mothers’ brodiers and sisters’ sons, I found that the seceding group had settled more than six miles from the original village. O n the other hand, where classificatory mothers’ brothers and sisters’ ^ sons had separated from one another, I found that in eleven cases : the new group had settled within six miles o f the original village, and in only one case more than six miles from it— although in five instances the new settlement was in a different vicinage from the old. In the m ajority o f cases (22-33), the leader o f the seceding group was the oldest brother or classificatory brother in a group o f uterine siblings or in a minimal lineage. In six cases he was the son o f the senior wom an in his following. This again illustrates the tenacity o f the mother-son bond in Ndembu society. I f the mother o f a founder o f such a settlement is alive he w ill take his mother w ith him. She w ill become the apical ancestress o f the village lineage i f the village becomes firmly established. In a sense, she is a living index o f the matrilineal character o f the N dem bu village, and she is her son’s embodied hope that his sibling village w ill become a m inor local matrilineage. In order to bring out more fully the structural implications o f fission in Ndem bu society I give some examples o f different types o f village cleavage below. (a) Fission in Shika Village 1 Shika Village, about twelve miles from M ukanza Village today, is a village whose nuclear lineage belongs to the chiefly matrilineal 1 Appendix II should be consulted for genealogical information.
208
Schism and C ontinuity in an A frican Society
group o f Kanongesha. In the latter half o f the nineteenth century: its founder Kanjimana (II,Bi), uterine nephew o f Kanongesha Kajing’a Mpata Yibam ba (II,Ai), left the area o f Kanongesha near the Lovw a R iver in Angola, on the occasion o f his uncle’s death. He moved because he feared the jealousy o f the new chief, the son o f Kanongesha Kajing’a ; 1 and he settled near the present site o f Shika Village b y the Lunga R iver w ith his brother and three sisters, W ayanda (II,Bz), Nyam bunji (II,B3) and Chilum bu (II,B4), his wives, and the children o f the uterine sibling group. After his death he was succeeded by his uterine nephew Ikubi Chingongu (II,C2). The older brother o f Ikubi, Chanza (II,Cx), resentful that his younger brother had been preferred to him as headman, founded a new village with his children, one o f his sisters and her children, and his classificatory sister N yakaw ong’a (II,C3), daughter o f Nyam bunji (II,B3), and her children. N yakaw ong’a’s son Nswanakudya (II,Dio) became one o f the first African Messengers o f the British South Africa Com pany’s Administration. Nswanakudya returned to the new Boma near the Lunga, but after a short while left the service o f the Administration and rejoined his kin at Shika Village. B y 1914, five years after his return, he had become the leader o f a strong uterine sibling group and had hopes o f succeeding Ikubi (II,C2), now an old man. He was older than Chipwepu (II,D4), uterine nephew o f Ikubi and his designated mulopu (deputy) ; and had a larger follow ing than Chipwepu. B ut in this year his sister N yawatwa (II,D ii) committed incest w ith her parallel second cousin Samuheha (II,D6), uterine nephew o f Ikubi. The old headman drove the pair violently from his village, which at that time was quite near the present site o f Mukanza Village. The incestuous couple lived for several years in the bush, shunning the local villages. Eventually Ikubi forgave them and the ritual o f ku-disola was performed to cleanse them, for they belonged to the same minor lineage w ith a com m on great-grandmother. B ut he said that they must live in another village lest his people ‘ should feel ashamed ’ by their presence. Nswanakudya (II,Dio) m ay well have seen this expulsion o f his sister as a blow to his hopes o f succeeding, for she had four children already by two 1 Succession to the Kanongesha chieftainship was patrilineal until 1890. Cf. p. 321.
Varieties o f V illage Fission
209
'-.previous husbands, a pow erful pressure group to support his claim for headmanship. In the early I920*s he broke away from Shika V illage and invited his sister and her husband, to build w ith him. He started his new v illa g e 1 w ith his m other and his own uterine sibling group, including his older brother M atoka, two younger brothers, and tw o sisters, one o f w h om had been married to the late headman o f the nearby village o f Chibw akata. Nswanakudya’s ow n father had been headman o f Chibw akata Village, and m any ties united and still unite the tw o villages, both descended from the first Lunda invaders. Later Nswana- kudya was joined b y his m other’s m atrilineal parallel cousin 7 Nyakashilishi (II,C4), daughter o f Chilum hu (II,B4), daughter t o f the apical ancestress o f Shika V illage lineage, nyachibamba ;.(II,A2). Nyakashilishi had recently been divorced b y her husband and had elected to g o to Nswanakudya rather than to ’ Shika, her village o f origin. Perhaps this was because her mother was closer to N swanakudya’s grandm other in sibling order than, to Headman Shika’s grandm other from w hom all the contemporary members o f Shika w ere descended. O ne often finds on consulting N dem bu V illage genealogies that the matri lineal descendants o f adjacent siblings o f similar age tend to live together, rather than w ith the descendants o f a m uch younger or older sister. In this case w e find a m arriage, reckoned incestuous, between members o f a village lineage, as the prelude and catalyst o f fission. Had there been any able senior m en am ong the children o f Nyam bunji (II,B3) it is possible that one o f them m ight have succeeded Ikubi, and the division o f Shika V illage m ight have been averted. B u t the senior lineage o f w a y a n d a retained the headmanship and rendered virtu ally certain the secession o f Nswanakudya (II,D io) w h o belonged to a ju n io r lineage. It is difficult to say whether the group w hich follow ed Nswanakudya was a m inim al lineage or a uterine sibling group to w hich w ere attached Nyakashilishi (II,C4) and her daughter. Certainly w ith reference to the com plem entary and opposed lineage o f w a y a n d a , the split occurred between that lineage and those o f nyambunji (1133) and chilumbu (II3 4 ), her sisters ; but in practice, the 1 Nswanakudya told me that when he left Shika Village there were about thirty huts there.
2 io
Schism and C ontinuity in an A frican Society
seceding group consisted o f Nswanakudya, his m other, and his siblings. Finally, it should be noted that the breach o f residential con tinuity w hich occurred w hen Nswanakudya and his uterine sibling group left Shika V illage to go w ith Chanza (II,C l) prob ab ly helped to disqualify him for headmanship in the eyes o f those w ho remained, the close lineal kin o f headman Ikubi (II,C2) and his uterine nephew Chipw epu (II,D4). M embers o f groups w hich have once broken aw ay from villages, like those o f N swanakudya in Shika and Yim bw endi in M ukanza, i f they return are less likely to succeed to office than are persons who remain. Residential continuity, difficult to maintain, is prized in proportion to that difficulty. (b) Fission in Nsang anyi Village andmits Offshoots 1 N sang’anyi is the senior K aw iku village in M winilunga D istrict, and was already in existence when Lunda invaders came from M w antiyanvw a. T he headman claims also to have come from ‘ Luunda the homeland o f all the peoples o f Lunda stock, at a period long before the arrival o f Kanongesha’s Ndem bu. Informants say that all the K aw iku villages and farms in the D istrict, tw enty-six altogether, hived o ff originally from N sang’anyi. In spite o f periodical loss o f membership Nsang’anyi V illage did not becom e extinct, but after each secession slow ly but surely expanded once m ore. The follow in g account attempts to trace briefly the history o f fission in N sang’anyi and its recent offshoots in the tw enty-five years between 1929 and 1954. The present headman (III,D4) succeeded to office in 1924. His predecessor had belonged to a different m atrilineage within the village, and they knew o f no com m on ancestress from whom both could trace matrilineal descent. The present headman called his predecessor * older brother w hich im plies that they were m atrilineal kin. N either lineage was reckoned b y the people o f N sang’anyi village to have been o f slave origin. Perhaps the lineage o f the previous headman, that o f n yan k u k a (III,B i ), may be regarded as a group analogous to w hat Professor Fortes calls among the Tallensi ‘ an attached lineage ’, linked to the authentic lineage, that o f nyachim a (III,A i ), by actual or putative cognatic 1 See Appendix HI,
V arieties o f V illage Fission
211
ties, and in the course o f tim e treated for ail practical purposes as matrilineaUy connected. I have collected another village ' genealogy, that o f Chikang’a, w hich shows that the founder’s son succeeded him , and was succeeded in his turn b y the founder’s sister’s son o r sister’s daughter’s son. T he uterine siblings o f the founder’s son rem ained in the village and his sisters’ children regarded it as their prim ary local group. Thus in a single village . were found tw o distinct matrilineages, one descended from the senior w ife and the other from the m other or sister o f the founding ancestor. In the second or third descending generation, members 1 o f the tw o groups began to call one another * brother ’ and ‘ sister * instead o f * cross-cousin *, since they w ere genealogical generation mates prim arily affiliated to the same village. Ties o f co-residence and com m on genealogical generation affiliation v took precedence over lineal tics w h ich m ight have differentiated i; the tw o groups.1 T he same thing m ay have happened in ; Nsang’anyi V illage, w hich is o f considerable antiquity and enjoys great prestige. These factors m ay have induced both lineages to remain in the same village. W hen the present N sang’anyi succeeded, tw o m en in particular were disappointed in their claims fo r office. O ne was his classificatory ‘ brother M atem pa (III,D3), w h o was son o f the new headman’s m other’s older uterine sister, and w ho was an able jud ge and advocate and one o f Senior Headman M ukang’ala’s councillors. T he other was N sang’anyi’s classificatory sister’s son, M w anaw uta Ikatu (III,E3), older than the headman, and leader o f a large, uterine sibling group w hich included five sisters. M wanawuta Ikatu was the son o f the older brother o f Senior Headman M ukang’ala. In 1929 the D istrict Com m issioner made a road along the C on go B order connecting the Bom a w ith Solw ezi B om a. This road passed beside N sang’anyi V illage and, nearer the Bom a, came w ithin a m ile o f Senior Head man M ukang’ala’s capital village. M w anaw uta Ikatu (III,E3) 1 Yet, as we will see, Kamawu’s brother, a member o f n y a c h im a lineage, who called Makumela (III,F5), o f k y a n k u k a lineage, his ‘ older brother \ married Makumela’s sister. This marriage was not regarded as incestuous. Thus in certain social situations, such as succession to headmanship, the two village lineages regarded one another as lineally related ; in others, such as marriage, distinctive matrilinea! origin was stressed. See Chapter Fight for a full discussion o f the dual character o f Ndembu village structure.
o o p 'O n o o f>
C o
o G o o o C) o o o o C; ■ 0 o o o c
G
G
■V m
z iz
Schism and Continuity in an fA rican Society
inform ed N sang’anyi (III,D4) that he wanted to ‘ split off* and make a village o f his ow n near his ‘ father * M nkang ala and closer to the Bom a.1 He said that he had m any siblings w ith their children, and that i f he lived nearer the B om a and Hose to Senior Headman M ukang’ala, w ho, in 1930, had been elevated to the status o f a Governm ent Sub-Chief, there m ight be an opportunity fo r his jun ior kin to find paid w ork. N sang’anyi tried to persuade him to remain, hut he left the village and made a new settlem ent near the Lwakela R iver bridge on the Govern m ent road, a m ile from M ukang’ala’s village and fifteen miles from M w inilunga Boma. In 1940 M atem pa (HI,D3) also broke aw ay from N sang’anyi and settled on the northern m argin o f the K aw iku Plain. W ith him w ent his o w n uterine sibling group and all the members o f n yan ku ka lineage. T he last tw o headmen o f the village had been members o f that lineage. T h ey could not expect to supply a third successor, but they objected to the appointm ent o f N sang’anyi on personal grounds. In 1930, w hen the policy o f Indirect R u le was first introduced, Nsang’anyi had applied to the Bom a fo r recognition as a * Govern m ent C h ie f’. T he D istrict Com m issioner had called in C h ief Kanongesha and Senior Headman M ukang’ala for consultation ; and N sangianyi had adm itted in their presence that form erly he had paid tribute to Kanongesha through M ukang’ala, ‘ son ’ b y perpetual kinship, and representative, o f the Senior Chief. Kanongesha claim ed that although an early N sang’anyi had resisted the efforts o f the war-leader (Kambanji) o f the first Kanongesha, founder o f Chibwakata V illage, to subdue his * M bw ela * follow ing, the second Kanongesha had sent his son M ukang’ala Kabanda against Nsang’anyi. M ukang’ala had defeated N sang’anyi and com pelled him to pay tribute, half o f w hich was taken by M ukang’ala for his ow n use and h a lf o f w hich w as sent to Kanongesha. In addition N sang’anyi had to give M ukang’ala a tributary w ife ('ntombu). In compensation N sang’anyi was allotted the ritual office o f Chivwxhanharm or * em blem -purifier * to M ukang’ala. T h e incum bent o f this office was entrusted w ith the medicines o f M ukang’ala’s senior 1 Before Mwanawuta left Nsang’anyi there were more than thirty huts in Nsang’anyi Village.
V arieties o f V illage Fission
213
headmanship and perform ed an im portant role in his installation ritual. B ut the acceptance o f this office excluded N sang’anyi from political authority over any village except his ow n, in the indigenous political structure. O ther K aw iku villages recognized Mukang’ala, n o t N sang’anyi, as their senior headm an. O n hearing this evidence, the D istrict Com m issioner turned down Nsang’anyi’s claim for Governm ent recognition and appointed Mukang’ala as Su b-C h ief o f the area in w h ich m ost o f the K aw iku villages w ere situated. It is lik e ly that the possibility o f Nsang’anyi’s appointm ent as a S u b-C h ief had been instrumental in retarding M w anaw uta’s secession from his village, since he m ight have hoped fo r an appointm ent in the N ative A uthority bureaucracy i f he had remained. T h e rejection o f N sang’anyi’s claim was probably an additional factor in his decision to build near M ukang’ala’s village w here some o f his follow ers w ere later in fact appointed as coim ciliors in the N ative A u th ority Court. Matempa (IH,D3) also, w h o became one o f M ukang’ala’s councillors before he broke aw ay, probably le ft N sang’anyi because he regarded his status as councillor as incom patible w ith his junior status w ithin the village. For in the village he was subordinate to a headman w hose claim for an im portant position in die new political order had been perem ptorily dismissed b y its authorities. T he nuclear group o f M w anaw uta’s village w as a minimal lineage descended from his m other’s m other, Nyachintang*a (m ,C x). B u t in his village w ere three groups o f uterine siblings, potential nuclei o f new settlements. There was his o w n uterine sibling group, the children o f his m other, N yalubenji (IH ,D i). There was also the sibling group led b y his parallel cousin Chipoya (III,E5), the children o f his m other’s younger sister, N yam ahandu (HI,Da). Finally, there w ere the five children o f his w ife Nyam pupa (HI,Ea), b y a previous husband. N yam pupa and her brother Kakunda (IH,Ex) had fled originally from Angola as the result o f a quarrel the nature o f w h ich I was unable to ascertain, and had been given sanctuary b y N sang’anyi. Kakunda w ent w ith his sister w hen M w anaw uta (III,E3) left N sang’anyi, and neither he nor N yam pupa had any other hom e than M wana wuta’s village. M wanawuta, already quite an old man when he left N san ganyi, had becom e m ore or less senile b y 1947. In that year
2 X4
Schism and Continuity in an A frican Society
John. Kam banji (III,E6), his m other's sister’s son, returned from the Copperbelt w ith some cash savings. H e began to throw his w eight about in the village and quarrelled w ith Kanem a (III,F2) and Six (iU,F3), trwo m iddle-aged uterine nephews o f Mwanawuta, w ho had for the last ten years or so undertaken the practical running o f village affairs. H e also quarrelled w ith Sondash (m ,F i), the oldest son o f Nyam pupa (III, Ea) w ho b y this time had becom e the leader o f a thriving uterine sibling group. Kakunda (III,E i), the brother-in-law o f M w anawuta, was an old man now , alm ost as decrepit as M w anaw uta (III,£3) himself. In fact, in the village the management o f affairs had fallen into the hands o f the jun ior adjacent generation to the village founders. In 1947, Sondash (III,Fi) told M wanawuta and Kanem a (III,Fa) that he was now an elder (mukuiumpi) w ith m any siblings, children and sister’s children, that he found it im possible to live in the same village as the obstreperous John Kam banji (III,E6), and that he proposed to found a new settlement about h a lf a mile away. Little resistance was offered to his departure fo r he was not their matrilixieal kinsman, nor was he M w anaw uta’s ow n child. Shortly afterwards, John Kambanji, whose older brother Chipoya (IH,Es) had died several years before, left M wanawuta Village w ith his three w ives, six children, and tw o sisters’ children, and made w hat he called a ‘ farm * quite close to M wanawuta. M w anawuta’s ow n younger brother Katoka (III,E4), an old man w ho had quarrelled w ith Kanem a (III,F2), went w ith John Kam banji. Later, a Lwena Public W orks Departm ent road-labourer from a neighbouring cam p married one o f John’s daughters and built a hut at his farm . B u t by 1954 John had lost nearly all his follow ing except his w ives and children. His sister’s daughter married o u t; his sister’s son left him for fear o f his irascibility ; his m atrilineal parallel cousin Katoka died ; and the P .W .D . labourer also died. It was hinted in. the villages o f the vicinage that John (III,E6) or one o f his w ives practised sorcery or witchcraft. T he fact that John was a blood-brother o f Sandombu o f M ukanza V illage, the notorious sorcerer (see Chapters Four and Five), probably aid not w eigh in his favour either. In 1954 Six (lII,F3) also left M wanawuta and founded a small form, consisting o f a single K im berley-brick house occupied by his elem entary fam ily. This last type o f fission is now greatly
1 Cf. p. 36.
G G
OG
on the increase. In the pedicle area a rash o f such farms has appeared in the last few years. Six has becom e a cash-crop grow er.1 H e cultivates rice as w ell as taking in land fo r cassava over and above his subsistence needs. H e sells to the Bom a ■ ' cassava m eal w hich is m ainly used to feed road-maintenance labourers in a large camp in the vicinity. Such petty com m odity agriculturalists are tending to break aw ay from their matrilineal kin in order to avoid the latter’s claims on earnings. John, too, although his new settlement was originally an incipient village by traditional structural criteria, is n ow the head o f a * farm * and relies on the cultivation o f cash crops and sale o f surplus subsistence crops fo r his incom e. Since 1947, then, three groups have seceded from M wanawuta ■ Village. L ittle conflict accom panied these w ithdraw als, for it was generally recognized that the headman was extrem ely old and infirm and unable to perform his role effectively. Before the secession o f Sondash (III,Fi) M w anaw uta V illage contained about seventy inhabitants occupying some thirty huts. It has ; been previously pointed out that today, w hen the population o f a long-established village has reached about fifty inhabitants, it tends to exhibit signs o f im m inent fission. W e have seen that the social unit w hich m ost frequently form s the nucleus o f a new settlement tends to be a uterine sibling fam ily, ranging from about ten to fifteen persons. A village like M w anaw uta, which ; contains three or m ore sibling fam ilies, some o f w h ich m ay be linked b y a clever leader into a single m inim al lineage, is more conspicuous for its divisions than for its unity. O n ly the joint operation o f a num ber o f factors m aking for cohesion (such as the capability o f the headman, the historical fam e o f the village, or the existence o f a num ber o f supplem entary ties o f kinship, affinity, generation affiliation and friendship, w hich cut across lineal attachm ent) can hold together such a congeries o f virtually independent groupings. M w anawuta V illage possessed few o f these countervailing tendencies : the headman had little personal authority, the village was but recently established, it was situated about sixteen miles from the nearest European centre, and its com ponent sections w ere interlinked b y few o f the ties mentioned. Sondash (III,Fi) and his siblings w ere step-children o f the
3 G G O G G
215
@ ® © @ © @ © O O 8 © O O © © O G S § © 0
Varieties o f Village Fission
2x6
Schism and Continuity in an African Society
headman and unrelated by consanguinity to the other village members ; one marriage only had taken place between members o f his and the headman’s matrilineal kin. John Kam banji (III,E6) had spent m any years on the Copperbclt and his experience and interests w ere different and opposed to those o f the high ly con servative old headman. He was in the senior genealogical generation to Sondash and as the result o f some early quarrel bitterly disliked him . Six (III,F3), although he was the uterine nephew o f the old man, had also been a labour m igrant and saw his future in terms o f personal advancement in the new cash econom y. In fact Kanem a (IIX,F2) was the only senior kinsman o f M wanawuta (III,E3) w ho had remained w ith him at the time I left the field, and he had hopes o f succeeding to the headmanship o f the depleted village, according to informants w ho knew the situation. T he story o f the foundation, rise and decline o f M wanawuta V illage is typical o f the fate o f m any Ndem bu villages in the past tw enty years as the cash econom y has in creasingly penetrated the traditional social system and corroded first the links o f classificatory, and then o f prim ary, matrilineal kinship. T he subsequent career o f M atempa V illage parallels in many respects that o f M w anaw uta. A fter Matempa’s death his uterine nephew Nswanam atem pa (III,E8) succeeded him and for a time the village maintained its unity and increased in membership. B ut in 1952, w hen Governm ent delim ited a new Forest Reserve, w hich included the K aw iku Plain and most o f the hush adjoining it, Nswanamatempa’s people, along w ith N sang’anyi, C hibwakata, and other villages, w ere m oved into the vicinage to which M ukanza, Nswanakudya and N g ’om bi belonged. T he leaders o f tw o sections w ithin the village took this opportunity o f splitting o ff from Nswanamatempa V illage w ith their follow ings. The senior elder o f nyanktjka lineage was M akum ela (III,F5), a man w ith eight livin g children, and he founded a farm , in reality an incipient traditional village in social structure, w ith tw o younger brothers, his m other’s sister’s son, their children, and a number o f his sister’s children. His sister was the w ife o f the younger brother o f K am awu (III,F4), leader o f the other seceding group, and she lived virilocally w ith her husband. K am aw u was the sister’s daughter’s son o f Matempa (III,D3). His follow ing basically consisted o f the sister o f his deceased m other and his
V arieties o f V illage Fission
,F ’
2x7
mother’s brother, his older and younger uterine brothers, his sister, and his and their children. K am aw u had becom e headman o f this group, in spite o f his lo w seniority, on account o f his w idely recognized qualities as a man o f legal skill, as a hospitable and conciliatory person, as one possessed, in fact, o f all the virtues regarded as desirable in a headman (see p . 200). His mother’s brother Biscuit (III,E7) had returned as an old .man after an absence o f m any years at B ulaw ayo, and had taken up residence w ith his sister at K am aw u’s settlement. A s a w idow er, he found it advantageous that the old w id ow was w illing to co ok fo r him . H e was a sad-eyed hum orist w ithout ambition, his on ly requirement a steady supply o f beer. M akumela built his farm facing Nswanam atem pa on the other side o f the m otor road, and K am aw u established him self a hundred yards from M akum ela on the same side o f the road. A ll three headmen, Nswanamatempa (III,B8), M akum ela (IH.Fs) and Kam awu (m ,F4), had asked M ukanza Kabinda’s perm ission to build in that locality, fo r M ukanza was generally reckoned to be mwenimbu, the headman longest established in that part o f the vicinage. T here w ere several abandoned village and garden sites and three separate graveyards o f the Kahali-M ukanza village lineage near the three new settlements, and the ownership o f these gave M ukanza his righ t to be reckoned mwenimbu. O ver the w hole vicinage N sw anakudya (II,D io) was recognized as pos sessing m oral authority since he belonged to the chiefly lineage, but Nswanakudya had no say in the allocation o f land that had once been occupied b y M ukanza V illage. Nswanam atem pa (IH,E8) had the nam e o f being one o f the four most notorious sorcerers in the vicinage. H e was thought to have caused the deaths o f his tw o brothers, o f a sister, and o f Kam awu’s m other’s brother and m other’s sister, b y his ‘ medi cine *. This, alleged Kasonda m y henchm an, w as the * real ’ reason w h y M akum ela (HI,F5) and K am aw u (HX,F4) split o ff from his village. K am aw u and M akum ela both told me, how ever, that all three groups rem ained friendly and recognized Nswanamatempa (III,E8) as their superior. In favour o f their autonom y they argued that * the big village was finished n o w V 1 M eaning that the type o f large village containing a large group o f m atrilirteal kin had been supplanted by the sm all farm inhabited by close kin.
218
Schism and Continuity in an A frican Society
that they w ere elders (akulumpi), and that each had a large follow ing o f his ow n close relatives. It is probable that both rationalizations for secession were actually advanced b y them , the form er privately, the latter in public ; but that really the prime cause o f their secession was am bition to becom e leaders o f local groups, am bition w hich had been given its opportunity to succeed b y m odem developments. Thus in the course o f tw enty-five years (1929-54) N sang’anyi V illage has given rise to tw o villages b y prim ary fission, M wanaw uta and M atem pa ; w hich in their turn have throw n o ff five further settlements, Sondash, John Kam banji, Six, M akum ela and Kam awu. T he history o f this little group illustrates a num ber o f points previously made. T he decline in political im portance o f Nsang’anyi V illage under the British Adm inistration seems to have weakened the bonds that held this form erly large and hetero geneous village together. B ut m any o f the headman’s character traits w ere not such as to commend him to Ndem bu. The leading elders o f his village blamed him for not obtaining the good graces o f the Bom a, pointing out that another senior headman o f M bw ela origin, Sailunga, had been persuasive enough to secure his ow n appointm ent b y the D istrict Com m issioner as * Para m ount C h ief* o f the Lunda-Kosa to the east o f the Lunga R iver. T hey said that i f Nsang*anyi had been firm er and m ore eloquent he m ight have becom e a Sub-Chief, and all his relatives w ould have benefited. H e was garrulous but incom petent in discussing village cases. H e was also m uch addicted to drinking. W hen I knew him he was regarded as a com ic, but rather pitiable, old man. He used to have his xylophone and slit gong, emblems o f a departed prestige, played at night in order to rem ind his departed kin that * he was hungry and thirsty * and that it was their duty to help him . Although he was often neglected in secular contexts, Nsang’anyi was still, however, accorded con ventional respect at rituals in K aw iku villages, w here he was given beer and food before other K aw iku headman. H e continued to preside over the installation rituals o f incumbents o f the Mukang*ala senior headmanship, and to receive presents for it. B ut his glo ry had clearly gone, leaving only decrepitude. The cum ulative pressures o f social and cultural change are
V*:> o
o @
© 0 ©• 1 J © ©
c
220
Schism and C ontinuity in an African Society
im portant factors behind these successive cleavages. W e have already seen h ow in M ukanza V illage m uch o f Kasonda’s and Sandombu s behaviour is explicable in terms o f response to new incentives, and h o w the abolition o f slavery b y the British gave Kafiim bu the opportunity to secede from M ukanza Village. Here again w e find that on the one hand direct political factors such as the refusal o f the Governm ent to recognize N sang’anyi as a Sub-Chief, and, on the other hand, the pervasive influence o f the cash econom y, have led to the gradual reduction o f a large village into smaller and ever smaller components, m ost o f which w ere hardly m ore than groups o f uterine siblings, and in one case, a single elem entary fam ily. In. Chapter T w o it was sug gested that in the latter part o f the nineteenth century some N dem bu villages w ere quite large and others small,1 both types representing different responses to the slave trade, and, later, to slave raids. The large villages w ere often com posite, containing several unrelated matrilineages o f relatively w ide span w ho had settled together both for m utual protection against slave-raiders and in order to m ake successful raids on other villages. T h e small villages m ay have represented another w a y o f escaping the raiders— b y breaking up ana scattering before them into the deep bush. In the present century, under British rule, w ith the abolition o f the slave trade and o f internal slavery, w ith the slow addition o f a cash to a subsistence econom y, and w ith a rapid post-war increase in the rate o f labour m igration, the large villages have tended to break up b y successive stages o f fission into the smallest possible matrilineal units, uterine sibling families. B ut change in the structure and stability o f residence has b y no means everyw here and invariably taken place at the same rate and in the same w ay. In a single area one m ay find traditional matrilineal villages side b y side w ith m odem fam ily forms. N ext to diem one m ay find settlements w hich m ay he regarded as transitional types, consisting o f depleted uterine sibling groups. Tip-service is still paid b y the m ajority o f N dem bu (despite the opinions, quoted above,8 o f M akum ela and K am aw u), to the value o f livin g in a large village o f matrilineal kin. T he members o f some villages, for instance o f M ukanza V illage, until very recently have deeply respected this value. B u t w ith the breakdown o f the 1 Cf. pp. 43 - 4 -
3 See p. 217.
V arieties o f V illage Fission
221
traditional authority o f the old m en o f the senior genealogical generation in a village, and the em ancipation o f the younger men from this authority, the norm s governing behaviour between generations have becom e enfeebled : younger m en o f w ealth and am bition no longer feel constrained to obey the w ill o f the headman and his generation mates, but assert their financial and political independence b y seceding and founding farms. T o some extent the values inherent m m atxiliny as a principle govern ing the com position o f residential groups still p ersist: uterine ablings, that is, members o f the same genealogical generation, between w hom relations exist o f approxim ate equality, still tend to build together. B u t today, especially in the northern pedicle, w ith its grow in g com m itm ent to a cash econom y, even the links between siblings are becom ing m ore tenuous, and individuals are founding farm s w ith their ow n elem entary fam ilies. Nevertheless, although m odem developm ents have undoubtedly accelerated the rate o f fission and perhaps reduced the span o f the seceding m atrilineal group, these developm ents have only recently, and in particular w ell-defined areas, produced the spectacular atom istic effects described above. In the past, as to day, the N dem bu lineage seems to have been shallow b y com parison w ith the lineage am ong such peoples as the Tallensi and the A sh an ti; and fission seems to nave taken place w ithin the m inor or m inim al lineage rather than dichotom ously between structurally balanced and opposed m ajor segments o f a village lineage. V illages appear to have been small and spatially m obile in the past also, as Livingstone observed. I f older informants are to be believed, N dem bu villages, before the epoch o f slave trading and slave-raiding, w ere hardly m ore than small hunting camps ; so that w ith regard to m agnitude, i f not to social struc ture, m odem * forms * resem ble ancient villages. B u t the social core o f the * form * is, as w e have seen, an elem entary fam ily, w hile the core o f an ancient village was probably a group o f uterine siblings. In the exam ples o f village fission given above and from the tables constructed on the basis o f genealogical data, the uterine sibling group appears as the basic unit o f secession. B ut the seceding groups are not necessarily segments o f a local m atrilineage, since die group m ay be variously attached to its leader.
222
Schism and C ontinuity in an A frican Society
KLafumbu’s (I,G i) follow ers w ere a group o f uterine siblings and their children, related to him b y ties o f m arriage and patrilateral cross-cousinship. T he group led aw ay b y Sondash (III,Fi) consisted o f stepchildren o f the village headman. Makumela’s (m ,F5) follow in g w ere not m atrilineal kin o f Nswanamatempa (HI,E8), although M akum ela and Kam awu (IU,F4) called one another respectively * younger * and ‘ older brother \ Appendix in, in fact, shows that K a m a W s uterine brother had married M akum ela’s sister. This marriage was not regarded as incestuous, since the people o f K am aw u said, * W e have a different matrilineal ancestress (nkakuluta) from M akum ela.’ Sometimes, it is true, segments o f equal span broke aw ay from one another. If, for instance, in M ukanza V illage, malabu lineage and nyachintang ’a lineage build in separate settlements, this w ill represent fission between tw o m inor lineages. A t a low er level o f segmentation the secession o f M atem pa (IH,D3) from N sang’anyi V illage and o f John Kambarqi (UI.E6) from M wanawuta, w ere divisions between tw o uterine sibling groups belonging to the same m inim al lineage, divisions between the children o f tw o sisters. T he most com m on unit o f secession is indeed the uterine sibling group led b y a prim ary o r classificatory sister’s son o f the village headman. M w anawuta’s secession from N sang’anyi V illage is an exam ple o f this kind o f fission, as is Kam awu’s from Nswanamatempa’s V illage. Ndem bu society, unlike T ale society, is not, in fact, * built up round the lineage system A fu lly developed ram ifying lineage system requires as a prim e condition for its grow th an ecological system in w hich lim ited access to resources is characteristic. Seen in structural terms, it also requires that the m ode o f postm arital residence should be consistent w ith the m ode o f descent. A system o f patrilineages, i f it is to be ‘ the skeleton o f the social structure ’, requires virilocal marriage ; a system o f m atrilineagcs requires uxorilocal m arriage. T he patricentric fam ily in the form er is w hat Professor Fortes calls * the grow ing tip * o f the lineage system, and like the lineage is attached to a locality. In m atrilineal uxorilocal societies the local m atricentric fam ily is such a grow ing tip. I f fem ale members o f a matrilineage remain in one settlem ent and im port their spouses, each wom an becomes the source and nucleus o f a local lineage w hich m ay attain considerable depth in a society where settlement is anchored
f
V arieties o f V illage Fission
22,3
to lim ited resources. In the next chapter I propose to discuss the structural im plications o f virilocal m arriage w ithin and between settlements in N dem bu society in considerable detail. ■ But some o f the consequences o f this antithesis between mode . o f descent and m ode o f post-m arital residence have already : become apparent in the last three chapters. In m any o f the : simpler societies the tie between m other and children seems to be the closest bond o f kinship. In societies w ith extended kinship systems, there is a tendency fo r members o f the m atricentric fam ily to gravitate together in co-residence how ever far apart ■ they m ay have been throw n b y other social tendencies, such as virilocal m arriage am ong m any W est-Central Bantu peoples. It is perhaps a tacit recognition o f the strength o f this bond that has given rise in m any strongly patrilineal societies to institutions directed against divorce,1 since the patrilineage o f the father has a pow erful interest in retaining his children to replenish its local membership, and there is a danger that w here the m other goes her children w ill g o also. In N dem bu society, virilocal mar riage, coupled w ith and opposed to m atrilineal descent, has the effect o f weakening the local developm ent o f deep lineages ; and at the same tim e it makes every fertile w om an w ho is livin g w ith her husband the potential starting-point o f a totally new settlem ent o f prim ary m atrilineal kin. I f for some reason a w om an is unable to return to her ow n m atrilineal kin, it is likely that when they mature her children w ill start a settlement o f their ow n. Thus w hen slaves became emancipated, the children o f slave wom en am ong N dem bu tended to form their ow n villages. The m other o f Sondash could n ot return to A ngola, and her many adult children, whose leader had been unable to succeed to headmanship in his stepfather’s village, split o ff from M wanawuta Village and made their ow n settlem ent. A gain, due to virilocal marriage, a w om an’s children are frequently brought up in their father’s village, am ong his m atrilineal kin, and n ot in close everyday contact w ith their o w n m atrilineal kin. Since they do not interact intensively w ith the latter, they do not develop strong sentiments o f m utual interdependence w ith them. Each 1 See G luckm an’s argum ent on this p o in t in * K in ship and M arriage am ong the L o zi o f N orth ern R hodesia and d ie Z u lu o f N atal \ in African Systems ofKinship and Marriage, ed. A . R . R a d cliffe-B ro w n and C . D . Forde (1950).
r>
o
O n
o o o o
c o G G ij O
o o o 0
o 0 .0 C; O o G Q C)
224
Schism and C ontinuity in an A frican Society
uterine sibling group tends to regard itself as autonomous. In their father’s village uterine siblings are outside the matrilineal descent group w hich holds office. T h ey do not interact in the affairs o f daily life w ith their ow n matrilineal kin from whom they m ay be geographically separated by a considerable distance. Even when they return to the matrilineal kin o f their mother, either in a com pact block after the divorce or w idow hood o f the latter, or singly and severally in later adult life, the ties inter linking them am ong themselves are o f a m ore durable and pow erful character than those attaching them to m ore distant matrilineal k in -e v e n to the children o f a m other’s uterine sister.1 A number o f com pensatory structural devices exist in the social system to prevent the secession o f uterine sibling groups. Local exogam y, w hich is largely responsible for the initial cohesion and autonom y o f such groups, later weakens their cohesion b y dis persing their fem ale members, as they grow up, am ong the villages o f their respective husbands. T he principle o f classificatory adelphic co-residence, w hich associates male matrilineal kin o f the same genealogical generation, establishes bonds o f co-operation between men belonging to different uterine sibling groups and opposes them to members o f adjacent genealogical generations. T he children, and sisters’ and brothers’ children, o f the head o f a group o f uterine siblings are classed as members o f a generation opposed to that o f their parents. T he uterine sibling group begets the means o f its ow n potential dissolution, since wom en m ay leave their brothers to accom pany seceding sons. I have stressed the importance o f the headman’s reputa tion for justice and generosity as a means o f holding together a settlement full o f potentially disruptive tensions between its com ponent uterine sibling groups and adjacent generations. The astute arrangement o f marriages between his ow n and his sisters’ children, and between members o f his ow n and the ju n ior alter nate generation in a village, is another means b y w hich a headman endeavours to maintain the continuity and integration o f his follow ing.
k) C) u
g ':
1 Unless they have been reared in the same village as the children o f a mother’s sister. If members o f a minimal matrilineage have been reared together they tend to form a potential unit o f secession, e.g. the children o f Manyosa and Chawutong'i in Mukanza Village.
V arieties o f V illage Fission
225
Thus the unity o f a village at any one tim e tends to depend, on a precarious balance between different categories o f conflicting alignments. Persons united as members o f a single uterine sibling group are divided b y virilocal m arriage, and united w ith members o f other uterine sibling groups b y membership in a com m on genealogical generation. M ale members o f such a sibling group are opposed to their o w n children w h o belong to a different sibling group, belong potentially to a different village, and belong actually to a different genealogical generation which '' *r ’ 1 1 * * ary and classificatory siblings and Fem ale members o f a uterine sibling group are opposed through m arriage to their brothers. N orm ally, i f their marriages are to rem ain in being they must live virilocally. Thus they set a lim it to the ambitions oF their brothers, w ho m ay wish to form a n ew village w ith their help, and in any case require the support o f their children to further intra-village am bitions. T h ey are opposed to their brothers* children if they live in the same village ; for their brothers tend to favour ow n children w ith food and attention rather than sisters* children, w ho w ill ultim ately oust brothers* children from the village. Sisters are united w ith their brothers as members o f the same genealogical generation, and w ith their other male and female classificatory m atrilineal kin, against all members o f the adjacent generations. In certain situations mothers are opposed to their ow n adult children, w h o, in seceding from the villages o f their husbands or brothers, disrupt and weaken those villages. W hen, for exam ple, Sondash m ade his ow n 4farm *, his m other remained w ith her husband, the o ld headman M w anaw uta ; and she scolded her son fo r * spoiling the village She also scolded her brother, w ho w ent w ith Sondash because he was given the office o f tnulopu or second-in-authority to Sondash, although he was too old to perform the duties o f a headman effectively himself. Thus persons are interlinked in one set o f relations w ith persons to w hom th ey are opposed in others. A lthough these cross cutting loyalties restrain and m itigate the pow erful tendencies towards fission in the residential unit, the uterine sibling bond in its closeness and exclusiveness still tends to assert itself at the expense o f all countervailing influences. T he nuclear group in alm ost every instance o f fission cited is clearly the group o f
226
Schism and Continuity in an A frican Society
uterine siblings and their children, although in m any cases not all the living siblings o f the leader o f the seceding group have been w ith him at the point o f fission. Sometimes one or more o f the leader’s sisters have been livin g w ith their husbands in other villages ; sometimes a brother has been elsewhere. T w o o f Matempa’s (III,D3) sisters remained at N sang’anyi and would not com e w ith him . Kafum bu, w ho founded a village w ith his w ife’s siblings, could not get his w ife ’s younger sister to leave Sakazao (fH p) her husband, o f M ukanza V illage, nor his w ife’s brother Kayineha (I,G6), w ho preferred to live uxorilocally at M ukanza Village. Another o f Kafum bu’s w ife’s sisters accom panied her husband Yim bw endi (I,Fi i) to the village Yim bw endi founded after his quarrel w ith Kafum bu (I,G i). B ut in time most o f the siblings o f such a leader w ill com e to his settlement. W hen his sisters are divorced or w idow ed they w ill jo in him w ith their children. The strong tie between uterine brothers also asserts itself. Gradually the uterine sibling group drifts together. O ften the children o f sisters w ho did not secede with their brother, jo in him when they grow up— boys w hen they m arry and girls after divorce. Thus the children o f Matempa’s sisters w ho did not accom pany him later built huts in his village. T h e structural principles, then, w hich govern residential affiliation are manifold, com plem entary, and also conflicting. M atriliny is in the final analysis dom in an t; but a peculiar cast is given to m atriliny by viriiocal m arriage, w hich reduces the span o f the effective matrilineal unit to the m atricentric fam ily. That the system is able to persist depends upon a number o f com pensatory principles : affiliation b y genealogical generation ; opposition o f adjacent, and alliance o f opposed, genealogical generations; and cross-cousin and grandparent-grandchild marriages. These are all means o f interlinking matricentric families in a wider system o f local ties and o f reducing the strength o f uterine siblmgship. It is im portant to note, however, that the effectiveness o f any one o f these principles varies from settlement to settlement and depends on such local factors as the length o f establishment o f the settlement, the fertility o f wom en and men, the strength o f marital ties, the reputation and astuteness o f the headman and o f candidates for headmanship, the age and experience o f these candidates, the local numerical strength o f
V arieties o f V illage Fission
227
sibling groups and m inim al lineages, and so on. Social change J produces alterations in die relative w eightings o f these principles S and gives rise to n ew types o f interpersonal and inter-group relationships. For instance, ties between members o f an elemen tary fam ily are strengthened and ties between m em bers o f a local maternal descaat group are weakened. Kinds o f struggle arise within the village against w hich the traditional m achinery for reintegrating a disturbed group m ay prove ineffective. I am aware that w hen I w rote that * virilocal m arriage reduces the span o f the effective matrilineal unit \ I was g u ilty o f over sim plification. In practice, the com plex interaction o f a number o f ecological and political factors, and n ot virilocality alone, is responsible fo r the reduced span o f the m atrilineal unit. The same set o f factors, w hich weakens the effectiveness o f m atriliny as a bond o f local affiliation, assists, i f it cannot be shown to determine, the em ergence o f virilocality as a com peting principle o f residential attachment. In societies governed by matrilineal descent, wide-span. local matrilineages tend to be found w here access to land or other valuable natural resources is lim ited. Such extended matri lineages tend to be associated w ith the possession o f fixed estates. In the natural region inhabited b y N dem bu, cultivable land is relatively abundant, w hile population density is lo w . Social groups are not constrained b y natural exigency to remain jerm anently in particular areas, in pockets and tracts o f fertile and. Indeed, since hunting is a h igh ly valued activity, and game tend to m ove aw ay from hunters, there are positive econom ic inducements to change the residential site periodically. These inducements are all the m ore effective for a population that does not invest in large, permanent houses, in long-grow ing and long-bearing trees, and in stumped land. N o r are Ndembu faced w ith the necessity o f settling near lim ited water-resources, for their territory adjoins the C ongo-Zam bezi watershed and m any streams and rivers take their rise in it. Nevertheless, given abundance o f land and lack o f investment in im m ovable property, it w ould still be theoretically possible for a large m atrilineal descent-group to m ove about together, and n o t to split up into smaller units. W h y does this seldom happen ? Part o f the answer lies undoubtedly in the individual istic productive system. The lim its o f econom ic co-operation,
i
"O •O
o o o o o o o 0 G G O o c n
G O
o C)
c o o o o G Q
G
G G €)
©
228
Schism and Continuity in an A frican Society
as w e have seen, are extrem ely narrow. M en hunt alone or in small bands. Com m unal working-parties to clear bush and hoe up mounds are infrequently mustered, and m any Ndem bu do not sponsor them. T he elem entary or polygynous fam ily can, and today often does, satisfy m ost o f its needs b y its own labour. N o t o n ly is access to resources virtually unrestricted, but there is no concept o f a jo in t estate w orked collectively by a unilineal kin group. N o one has a perennial interest in any one tract o f land, owned jo in tly b y his lineage. Furtherm ore, Ndem bu do not possess large herds o f cattle, the jo in t ownership o f w hich m ight under certain circumstances hold together a large nom adic group o f unilineal kin. The m ost valuable item o f m ovable property they possess is the m uzzle-loading gun, and o n ly a m inority o f m en ow n efficient guns. Since such guns are scarce they are h igh ly esteemed and are h o tly com peted fo r in situations in volvin g inheritance. They tend to divide rather than unite matrilineal kin. T he true problem , indeed, for Ndem bu society is not w h y local matrilineages are small, but rather w h y they exist at all. W h y do N dem bu live in villages, die cores o f w hich are male matri lineal kin, and not in fam ily homesteads ? In the past, at any rate, as w e have seen, one o f the reasons was provided b y the necessity to unite in strong defensive groups against slaveraiders. O ne response to slave-raiding was to build palisades, throw up earth-ramparts, and dig ditches around large villages. Such villages w ere, in effect, fortresses, in a state o f perpetual vigilance against siege. T h ey seem to have been attached to die same tracts o f land for long periods. T heir male members must often have co-operated to repel raiders, and to defend their wom en, children and gardens. O n the other hand, an alternative response to slave-raiding was to scatter in small groups into the hush, each group holding itself in readiness for instant flight. Y e t each such group, according to tradition, had a nucleus o f uterine or matrilineal kin. W h y w ere such groups w ider than the fam ily ? N dem bu have pointed out to me the practical advantages o f livin g in villages. T hey say that w hen one fam ily-head is short o f food, especially o f meat or fish, supplies o f w hich are. sporadic and uncertain, he can obtain it from a village kinsman or neigh bour. Later, when he * finds meat * he can m ake a return. In
V arieties o f V illage Fission
229
other w ords, livin g in villages under a subsistence econom y raises the general level o f consum ption and safeguards against individual shortages. N dem bu have also told m e that before the Europeans came N dem bu used to raid each other to obtain captives w hom they could sell to O vim bundu traders as slaves in return for guns, pow der and d oth . A large village was a m ore effective raiding and defensive unit than a sm all one in those troubled times. Then as n ow the local tie between uterine brothers was a very dose one. A brother w ould help one in raids, and i f a man was taken captive, it was his brother’s duty to recover him , b y force or paym ent, i f he possibly could. T he Icdandic m axim , which Dasent made the m otto or his translation o f N jal’s saga, w ould be equally appropriate to N dem bu circumstances : ‘ bare is back w ithout brother behind it.’ A gain, i f a man w ere killed, it was his brother’s duty to avenge him o r to exact blood-com pensation. As w e have noted on m any occasions, i f uterine brothers are to remain together, they m ust m arry virilocally. W e m ust regard matrilinea! descent itse lf as given in this sodety. M atrilineal succession and inheritance are regarded as axiom atic b y Ndem bu. M atriliny is sym bolized in m any kinds o f ritual, and is supported b y the dogm a o f descent ; and its im portance is emphasized in several folk-tales. I f male kin resided together fo r defence, offence, and m utual econom ic aid, the m ajority o f them w ould have had to be uterine and matri lineal kin. It is on ly the small span o f the effective local matri lineal unit that w e have to consider. It is probable that the external provision o f law and order b y the British authorities has contributed to the present small size o f local matrilineages. Large units fo r defence and offence in raids are no longer neces sary. E ven before the introduction o f cash, that solvent o f corporate kinship groupings, the individualistic tendencies in production m ust have resulted in the breakdow n o f large local groups. These, as w e have noted, had no permanent jo in t estates o r com m unal m ovable property to keep them together. I f a m an wished to secede from a village w ith his fam ily and kin w ho w ould fo llo w him , he could alw ays obtain residential and agricultural land elsewhere. These factors, then— abundance o f land, individualism in production, lack o f investm ent in valuable fixed resources, and
230
Schism and Continuity in an A frican Society
external provision o f law and order— do not provide favourable condition^ fo r the form ation o f wicle-span localized matrilineages. W hen, in addition, Ndem bu observe the custom o f virilocal marriage, a positive tendency m aking fo r the reduction o f the span o f the matrilineal unit m ay be said to be at w ork. It is o n ly b y describing and analysing a num ber o f cases o f village fission in terms o f the social drama, or, w here there are insufficient data, in terms o f the case history, that the precise effects o f these principles and o f social change in any given situation can be accurately assessed. For exam ple, w hen one examines the com position o f the seceding and rem aining sections involved in village fission, one seldom finds that any one principle o f social organization has clearly prevailed over the rest. In Nswanamatempa’s V illage one m ight have expected that Biscuit (UX,E7), uterine uncle o f Kam awu (M ,F4), and not Kam awu, his nephew, w ould have becom e the headman o f the n ew settle ment, since uterine uncles have authority over their nephews. B ut w hen the particular circumstances o f the case are considered, Biscuit is found to be a gentle, irresponsible drunkard, whereas Kam awu is a respected ihaku (law-man) w ith m any siblings and children. These factors outw eigh the values set on age and kin ship position as criteria influencing selection for political office. A m bition and ability becom e significant in any given instance, and have repercussions on the social structure o f the local unit. In the social drama w e can observe h o w particular individuals manipulate the organizing principles o f social affiliation to serve their ow n purposes. Sandombu, although childless and w ithout fertile siblings, exploited cleavages in other sections o f M ukanza Village and invoked the principles o f generation affiliation and the m other-child tie, to build up a follow ing. B y his hospitality to distant kin and strangers, he made apparent to all that he w ould give protection and assistance to persons w ho, fo r whatever good reason in N dem bu opinion, had been excluded from their ow n villages. Social prestige and authority in societies based on kinship principles are largely a function o f the relative fertility o f individuals and families. Sandombu was unlucky in this respect, and i f he had been a less forceful character, he m ight w ell have resigned him self to be a man w ithout a follow in g, a social nobody. B u t he was resolved to be a headman, and consequently had to utilize other means o f obtaining a follow in g than those
Varieties o f V illage Fission
231
which lay to the hand o f any fertile man w ith a fertile sibling group. In doing so he made m any enemies since he was com pelled to surround him self w ith strangers, social undesirables, or to attem pt to w in over b y hospitality those w hom m ore fortunate men w ould have regarded as their * natural ' supporters—-close blood-kin. N um erical analysis tends to ignore as irrelevant the unique features o f each instance o f fission and to stress regularities, the statistically norm ative pattern. Y e t it provides the background against w h ich each social drama assumes a significance. I f one finds that the m ost com m on unit t>f fission, fo r instance, is the uterine sibling group, and then, in a given case history or social drama one discovers that the unit o f fission is the minimal lineage o r the elem entary fam ily, one is led to inquire into the reasons fo r this apparent anom aly. I f the data relevant to the analysis are collected in terms o f the social drama, the lines o f alliance and cleavage in the specific village becom e visible ; and attention is focused on the particular norm s w hich are observed and broken, on the actual m otives w hich guide the behaviour o f the participants, and on the detailed econom ic, political and other interests w hich unite and divide them. Apparent excep tions to structural regularities discerned in the num erical data are som etim es found to be the product o f a com bination o f factors w h ich taken alone, or in their unhampered expression in different situations, also possess regularity. Thus a series o f marriages w ith in perm itted categories o f kinship betw een tw o uterine sibling fam ilies or between tw o m inim al lineages m ay decide w hether these groups secede o r remain together in a particular drama o f fission. Indeed the dynam ic interaction o f specific persons and groups in the process w hich I have called the social drama falls w ithin the province o f the sociologist no less than the analysis o f the statistical and ideal norm s o f social structure. In the social drama w e see social structure in action. O ur attention is draw n to the anomalous isolated instance, the apparent exception to statistical regularity, ju st as m uch as it is directed towards the manifestation o f that regularity. In consequence, w e are led to inaugurate a series o f enquiries into the nature o f the social mechanisms w hereby regularities are maintained, and into w hat happens w hen regularities are broken. W e are led to detect at ju st w hat points the fabric o f the social
T7 O
o o
o e o o o o o o
G O G C
o 0 G O G
G
C> C
C; O G
G O o
O
©
232
Schism and Continuity in an A frican Society
structure is weakest, and w hat means are taken to solve problems o f social integration posed by that weakness. As an exam ple o f this social handling o f structural vulnerability, let us take the recurrent situation o f the exceptional persons or groups w h o transfer loyalty to the faction that opposes their own prim ary kinship allegiance in a conflict w hich terminates in fission. In M ukanza Village, Kayineha (I,G6), brother o f Kafum bu’s w ife, was a case in point. W e w ere led to inquire w h y he did not secede w ith his other siblings, and biographical data w ere presented w hich showed that he was a personal friend o f M ukanza Kabinda, and that he had married N yam w aha’s daughter. W e also found that in the fight between the Kafum bu and M ukanza factions, Kayineha acted as peacemaker. Later he assumed an intercalary role between the tw o settlements o f M ukanza and Kafum bu. W e found that Sondash’s (III,Fi) m other, w ho did not secede w ith her son, acted in a similar capacity as a lin k between M wanawuta and Sondash villages. M atempa’s (ni,D 3) sisters w ho remained behind at N sang’anyi but whose children later w ent to their uncle, also occupied such an intercalary position, as did M wanawuta’s full brother w ho seceded to John’s setdement. In fact, persons w ho do not conform to the rule, m ay thereby acquire structural importance. W ithin the system o f internal village relations these persons appear to he the exceptions to the rule that uterine siblings secede together. B ut a study o f their behaviour at the point o f fission leads to a retrospective inquiry into their m otives for action, and an exam ination o f their subrole as intermediaries between groups to w hich they are con nected by ties o f close kinship, and groups w here they actually reside. W e learn that they perform an essential function in the w ider system o f inter-village relations, preventing the total e s t r a n g e m e n t o f groups divided initially in anger. Thus apparent exceptions to statistical regularities obtained from genealogical data on village fission prove to be themselves regularities within a w ider system o f social relations. V illage fission w hich divides the basic unit o f setdement, the village, tends on the whole to contribute to the integration o f the w ider system, the N dem bu socio-geographical region, by the provision o f links o f maternal kinship w hich jo in vicinage to vicinage, and senior headman’s area to senior headman’s area.
V arieties o f V illage Fission
233
The w ider system o f social relations gains at the expense o f the narrower. Summary M ale kin form the residential core o f villages : in a matrilineal society such m ale kin must im port their spouses or separate from one another : virilocal marriage in a spatially m obile society prevents the form ation o f deep lineages and gives a high degree o f cohesion and autonom y to the m atricentric fam ily w hich later becomes the principal unit o f fission. T h e looseness o f ties between m atricentric families leads to a high frequency o f fission, and their small size determines the small size and large num ber o f settlements. Within the village other principles o f social organ ization than the lineage become im portant, notably the opposi tions and alliances o f genealogical generations : between villages relations o f spatial propinquity and affinity becom e m ore heavily weighted than relations o f lineal kinship as factors m aking for cohesion in a vicinage. Ties o f m atrilineal kinship, on die other hand, are utilized to interlink separate vicinages. E cological and structural factors, interlocking in a com plex fashion, jo in d y determine the form o f setdement, the m ode o f its fission, and the w ay in w hich setdements are interlinked w ithin the w ider social systems o f vicinage and tribe.
C H A P T E R VIII TH E
STR U CTU R A L
IM P L IC A T IO N S
OF
V IR IL O C A L
M A R R IA G E W IT H IN T H E V IL L A G E HE first consequence...o f. jririlocal m ariiage_in Ndem bu residential structure is that the m ajority o f village .members tend.to. belong, to. a singlebilateralextended.faiiiily,.consis.tm g o f the Jheadman, h is. w ives and cHldren^ and ofJiis-siblings,, their es and childxen. There is no w ord in Lunda for * elem entary y * but the term ntanga connotes a group o f bilateral kin and their in-laws w h o are dependent on som e particular person. Ntanga, the singular, or antang a, the com m only used plural form , is always used w ith a possessive pronoun, as in ntang*ayami or antang*ajami, * m y dependent kin group * o r ‘ m y dependent kin (including in-laws) \ Its use can be extended to include all one’s kin w ith w hom one can trace genealogical connection, but in practice the term usually refers to one’s local kin-group and is restricted to kin and in-laws w ith a jun ior status to oneself. Thus in M ukanza V illage, all the villagers belong to M ukanza Kabinda’s ntanga. Kasonda’s ntanga includes his w ife, his minimal lineage sisters, and his and their children and grand children. It does n ot include his older sister’s husband w ho is an older man than Kasonda and tends to m ove his residence between his ow n farm in Kanongesha Area and M ukanza V illage. Sakazao’s ntanga comprises the w hole o f malabu lineage, and the spouses and children o f lineage members. Consequently his moral authority over ju n ior kin overlaps w ith that o f M ukanza Kabinda, since bis sister N yam ukola is M ukanza Kabinda’s w ife, and M ukanza’s children b y her are his sister’s children. The ntanga is n ot th erefo rean _independent,_clearly demarcated corporate unit, but varies-feom.senior..memberuto.senior-member, as each, is regarded as its point o f reference. Ntanga or antang*a must be distinguished from aumsoku (the dural form o f kawusoku, * a kinsman or in-law *), in that the atter term refers in practice to all one’s kin and affines whereas antang a is usually restricted to one’s jun ior kin w h o reside w ith one. O ne’s awusoku are, as it were, a pool from w hich antang a
T
S
{
234
V itilocal M arriage w ithin the V illage m ay be drawn. B u t the term aumsohu m ay be broken dow n into form er subdivisions according to the context o f situation. I f one wishes to contrast one’s kin from one’s afones, one m ay refer to the form er as aumsoku or tvusoku wakupwalika (* kinship b y birth ’), and to the latter as aka o r m atm ku, and mashaku (singu lars, mukti and ishahu). A k u o r aumku refer to in-law s o f an adjacent genealogical generation; mashaku describes in-laws o f one’s ow n genealogical generation. B oth categories m ay be further subsumed as wusoku wakusumhwang ana (kinship by marriage). B u t i f one wishes to distinguish the category o fo n e ’s kin w ho are respected (ku-lemesha) and feared (ku-tiya woma)9or w ho respect or fear one, from other categories o f kin, one also applies to them the terms aku and awuku . A ll one’s relatives, including parents and children, m ay be so described w h o belong to an adjacent genealogical generation. Brother and sister may also sometimes refer to one another as aku. T o m arry one’s muku is incest (chimaltvamalwa, chipikapika or hu-shim una). B u t the concept w uku admits o f a considerable range o f behaviour patterns between com plete avoidance, and respect tempered w ith affection and fam iliarity. In-laws o f opposite sex in adjacent genealogical generations should avoid one another ( bi-dichina ) ; a m em ber o f the ju n io r o f tw o adjacent generations should 4feel fear for ’ (ku-tiya tuoma kudi) his o r her m other’s brother o r father’s sister, his potential parents-in-law, w ho are often enough his or her actual disciplinarians; and a person should * respect * (ku-lem esha) his or her parents and siblings o f opposite sex. B u t one also * loves ’ (ku -ken ga ) one’s parents and siblings o f opposite sex. In general the verb ku lemesha, 4 to respect ’, is used to summarize the totality o f respectfear-avoidance attitudes regarded as appropriate betw een aku. B u t the best w a y to learn the idiom o f N dem bu kinship is to exam ine local system s o f kinship, beginning w ith the ntanga or bilateral extended fam ily o f the village headman. I must-repeat that, in its simplest and m ost fondam entd„fom i, the basic m embership o f an Ndembu. viUage.consists„ofa.]b^(hnan, his uterine siblings,, and his. and their children and grandchildren. In addition, it m aycontain spomes.ofsomejai^alfeofethes.epersons and perhaps some great-grandchildren of the headman. This group is the ntang*a o f the headman. T he spatial arrangement o f the huts ow ned b y its members gives the first clue to the
jjf
ya BB
O
o o o o c
236
G
c o o
o ©
c o c G O
o o o o
o
C; c
C) c
© &
.
S ch ism an d C o n tin u ity in an A frica n S o cie ty
character o f their mutual interrelations. As I described in Chanter Three, members o f. the_same_genealQgical_.gen.eratioii tencLto live in the same sendchcie,.o£.hu^-nhileutneinbers o f the adjacent generations live in the .opposite -semicircle^.. This means that members o f altemate_genealQgical_generations build together,, and their.hutsare.often, interspersed__The..arrangement o f huts in fact suggestsjhatThe_village-is. coru:eived-not. as a set o f hierarchically organized lineages. but _as ..a,. single. .extended fam ily ; fo r as a rule members o f different lineages-of variable span in weE-established villages build adjacendy, .so ,that extended fam ilial unity takes precedence over. Hneage-diftbcentiation in the spatial arrangement. Jelj*. sm allribjing.. vdkge.jo£ihe-type-W ^ are considering, m atriliiieal kin, atid tlie-children and- patriliaeal kin o f the headman, are interspersed. W ithin the ntang’ a, ego refers to his matrilineal kin either as akwamama or mwivumu detu (literally * in our w om b * or * in our lineage \ referring to matri lineal co-descendants o f a com m on ancestress), and to patrilineal kin as akwatata, w hich signifies persons descended through agnatic links from a com m on ancestor, alm ost invariably a common, grandfather. These are not patxilineages but i f m en always bring their w ives hom e, a man and his. son and.his son’s children m ay all be living in one village. T h eisntang’a, thus, has lineally tw o m ajor components. Its vertebral members belong to the matriiineage o f.th e headman. Female members o f the matriiineage maintain its continuity through-tim e, ..w hile its._male_meird)£rs-fo.rm.the_.residential core o f the village, at. any. .given.. moinent.._...B_u.t..through..virilocal marriage, the headman,- who-_is_freiju.enriy.. p-Qlygarnous, his brothers and..bis sons, have, children „who., together..constitute quite—a .high proportion, .o£....the-..po.p.ulariQn_Q£. a„_viOUiage (see Tables X and X I). In Chapter T w o (p. 44) I suggested that there is a functional relationship o f some im portance between the size and structure o f the local unit. N dem bu villages have a mean size o f about ten huts, containing about tw enty persons. Ent this small settlement, male village-kin outnum ber fem ale village-kin b y m ore than tw o to one (see Tables X and X I). Since the children o f brothers call one another siblings, the village membership has quite as m uch the appearance o f an extended fam ily as o f a local matriiineage. This virtual equality , between.farnily .atid lineage as. .principles
i -
t
<3
V irilocal M arriage within the V illage
237
o f local organization is at least p a rd ^ y xesponsible__£or_the m erging o f patrilineal and m atrilineal kin as jo in t members o f a single genealogical generation. Thus in m any situations^a man’s father and his m other’s brother have equivalent functions^and. are treated™ alike. B oth should, according .. to custom ,_contdhute the same sum o r the same am ount o f goods to a man’s, bride wealth. -W h e n a husband gives .bride^wealth- for...a-joroman, he should give an equal portion to_her,father..aiidj3JGtheiLs„.brother.
A womand^ph^es jhe..samejtexms™Xat«k«,..m«fewesi*«M to her husband’s father and his mo.theris. „brother.. ..and.» must equally avoid (ku-china) both. Conversely, a man should avoid not only~his w ife’s m other but also her father’s sister, to both o f w hom he-refers, and both o f w hom he addresses,-as-muku or mawenu. A man calls his o w n m other and also his m other's brother’s w ife mama (literally * m other ’) and greets them in the same w ay. H e greets his father and his m other’s brother in the same w ay also. T h e equivalence o f father and m other’s, brother, and of_m other and father’s sister, is expressed in a .number o f other situations as—w ell. -Ib is .-merging„..o£.^e^.Uy__iJistinct members o f the senior adjacent and authority-holding, generations is a further index o f the failure o f the m atrilineaL.principle to becom e.indisputably dom inant am ong N dem bu.... Intends to strengthen the bonds betweem m em bers.-ofihe-sam e-genealogical generation, and. at the same tim e, b y the creation o f a divided authority in the senior generation, to enable members o f the ju n ior generation, to exploit the division between father and m other’s brother, to their o w n advantage....Ih is_ d ivisio n o f authority^ând.. consequently.. limitation._of^unilareral ^onU:ol b y either father. orLmother’s brQther imdQubtedly contributes__tp the independent character orientation o f N dem bu, w hich,, in its. turn, helps to account fo r the high frequency o f fission. T h e Ndem bu_village, then^in its .s.Qcial.cQinpQsition-Eepj:esen||^ at once^a veiled -struggle, between- the_two_.powerfiiLpii||&pl^P o îfamiliaLan.d lineal organization, a n d -an -attem p t-to -i^ ^ ^ fi^ these b y a.jset..of com prom ise, form ations between..thei^il^rdey o f social control._In „the. .course o f the .stm ggle these_prinçiples m utuallyinhibit .one another,_s.o..tha.t.Hneages xemam-shaEow. and extended £amilies..hteak.up. o r.lo se .their.membership, through the pull o f m atrilineal affiliation....This, struggle. is itself a. manifesta tion o f the deeper opposition between male.. .andL..female. in
23 S
Schism and Continuity in an A frican Society
Ndem bu society. Each village represents an attempt to establish a patriarchal settlement in the teeth o f basic matrinneaL descent. B iit the male kin w ho reside together w ith theirw ives and families are themselves interrelated, b y matrstinea1.ties.which..persist,_while ties jvith their children. which_interlink_them„in .cof=residence are m ore, tenuous and friable, snapping, w ith the divorce, o f their wives5with_the mamageoftheix^daughters,and w ith th e frequent
de&^$^.o£.„sotis„^.,iheuL.i33^._jmdj3ineaL-kin._oir,.attaiiiing m ajority., .On..a, viU age.,geni^ogy jh e,Iin e,o£ descent,, .the spine o£ village conrinmty, is through jvon ien .whose_br_o_thers are marginal to it. ,. But in any given_village,at.any,specific.point o f tim ejthe line is broken or. in,rare,cases_even_imrisible.,.. For the residential core o f a viHage consists^o£men linked. to_,Qne,another through wom en, most o£w h om may, be,ab&ent,at..any_.one time, or w h o are already .dead. In the course ,o f events^.however, nearlyjevery. village contams some female,mLem.hers .o£ its nuclear m atrilineage,whoJiave,retum ed.to-it„after,divorce-O E-w idow*“ hood;. or-reside,diere_in.-cross=cousin-or-grandparent=grandchild
marriage, or whose husbands liv e uxorilocally These, w om en betray b y their presence A e -fm A m m u tty matrilmeaLcharacter or the* local unit. Thus,the kinship basis, o f. a villagers .as m uch, bilateral as uni lateral and represents a compr.omise .between,familial_ and. lineal principles o f organizations....-In. m y ..opinion,, .the..socio-spatial grouping..o f viliage kin by__genealo_gic_al.,ge_nerarion,_ as noted above,,'tends. to .neutraH2e sQme,o£ .the .tensiQns,arisin.g.£rom the co-existence, o f ..disparate principles ina..single, lo cal.im it...C om m on membership_o£a genealogical.generation aligns persons w ho belong to potentially opposed...descent .„categories,„. m other’s brother and father, w ife and sister,.husband,and.brother,.son and sister*s son, daughter and sister!s,daughters_memherSL.of,s.eparate ^grb^^^sf uterine siblings, or., members .o£.separate--miniinal or ^Maaqfe?3ipfeages__In m ost N dem bu vihages the headm an.is the 'H ^ t^ oj^ ll three m ajor modes o f organising, kinship-relations, y lineage, b y extended bilateral fam ily, and. b y genealogical generation. I£heJs,.an_astute.nianJhe_endeavours,ta-reduce the tendencies to disruption Q£.the_viUage,,which,originate.in contra dictions.between, these principles, by.encouraging-and taking part in theqjrow th o f a netw ork o f afSnal ties These ties are.form ed between persons belonging to opposed ...categories ...based on
f
V irilocal M arriage w ithin the V illage
239
differentprinciples^ of^ sacialorgam zatiori. Thus in M ukanza Village, the headman (I,F8) b y his m arriage w ith N yam ukola (I,H io) united the lineages o f n yach intang ’a and malabu . B y the same m arriage he strengthened the alliance between genealogical generations F and H (see A ppendix I). B y arranging cross-cousin marriages between his daughters b y N yam ukola and his sisters’ sons, Kasonda (I,G i5) and Sandom bu (I,Gxo), he sim ultaneously sought to increase the internal cohesion o f nyachjntang ’ a lineage and o f the village lineage o f widest span, that o f nyachipendi, sister o f the founder o f the village. A t the same tim e, both b y means o f his ow n m arriage w ith N yam ukola and b y means o f the marriages o f his daughters w ith his sisters’ sons, he tried to overcom e the dichotom y between his seminal children and his ju n ior lineal kin. Since N yam ukola belonged to the village lineage his children by her w ere fu ll village-m em bers. A n y children that Kasonda and Sandombu m ight have begotten on his daughters w ould also have been fu ll village-m em bers b y m atrilinea! descent, and the rivalry betw een Kasonda and Sandombu m ight have been re duced b y the fact that they had married sisters. Kasonda and Sandombu belonged to different m inim al lineages, each o f w hich represented a potential unit o f fission. B u t their children b y .4JJtiQrÌl tw/Ml\A h* ve belonged to the same N yam ukola s oaugnters m inim al lineage and w ould have been aligned w ith their grand father M ukanza in the same linked-generatton segment. M ukanza’s other m arriage, w ith Seliya, a rem ote matrilineal kinswom an, also ensured that his children b y her w ould belong to the village m atrilineage and at the same tim e strengthened his ties w ith Line (I,H i 7), a senior man o f m alabu lineage w ho had m arried Seliya’s sister. In this case too nyachintang ’a and m alabu lineages ,w ould have a jo in t interest in the children o f these m en. Kasonda’s divorce o f M ukanza’s daughter and Sandom bu s sterility thw arted some o f these intentions. B u t M ukanza, like other headmen, clearly understood the necessity ion forestalling b y m arital ties the potential disruption o f h js v village. Sodrces__o.f disruption .w ere .divisiom....arism^^i^hjav^ m atriliny, conflict between_sem inalchildren ^ancLsisters’ j^ pf and struggles betw een adjacent_genealogical^generarionll^ T he contradictions between these three principles o f village
240
Schism and Continuity in an A frican Society
organization— lineage, fam ily, and-generatioifc the term inology.. of__the Ndem bu.. and influence. jdxeJhehavioural noxms governing the interrelations .ofJdn. It is w orth repeating that the simplest w ay to approach the analysis o f Ndembu kinship is through the exam ination o f the spatial structure o f the typical village. The village hut-circle consists o f tw o semi circles, occupied respectively b y members o f adjacent genealogical generations, and composed respectively o f members o f alternate genealogical generations. Bach linked-generation segment calls all the members o f the opposite segment collectively aku. Aku o f the senior generation le v y respect from aku o f the junior generation and exert authority over them. Betw een aku who are in-laws, and, w ithin this category, between aku o f opposite sex, the relationship o f respect is at its most form al and rigorous. Aku o f opposite sex avoid one another, and aku o f the same sex exhibit extrem e constraint in one another’s presence. But affinal aku do not, in practice, enter into a relationship which confers a high degree o f authority on the senior over the junior generation. H ostility in the relationsjhip.pxedQminates.joyer the factor, o f social control, resulting in .av.oidance....and....c.Qnst:raint rather than in authority. N o t infrequently such aku physically attack one another. For exam ple, one o f the prisoners at the Bom a in 1950 was a man w ho had murdered his son-in-law after a fierce quarrel. Benson o f N g ’om bi V illage (mentioned in connection w ith Social Dram a II, p. 118) once beat his m otherin-law severely after she had upbraided him for taking a second w ife and neglecting her daughter. These examples indicate, h o w near to-xhe-surface-hostility m ay comein_this. relation ship,jan d ^ hy _the..noxms^oxeming..it stress the javoidan.ee. .of. m utual.interaction. ...a,s,far.. as„p.QSsible..... B ut a relationship o f this.kind .can hardly., he. a. su itab leh n k.in .a chain o f authority. ... A -relationship.. o£. superordination-subordination must_also involve xo^operation.,..There is. a chain .of-comm and. B ut .there...can .be little . co-operation..betweeru Jcm^who avoid one .another or. experience-.extrem e..constraintJuGL-oneuanother’s presence. Am ong. Ndembu,.^authority—relations_ar.e...carried in the Jdnship..structure. Thus that category o f aku, over w hom one effectively exerts authority and w hom one in practice obeys, is made up o f consanguineal relatives o f adjacent village genea logical generations. W rfhin._this„category,. again.r_..one..finds a
V irilocal M arriage within the V illage
241
range^of behaviour patterns. O n the w hole, m atrilineal kin o f the senior adjacent .generation exert greater^controLover- their junior relatives than do paternal.kin. B ut the; sharp differendatioii be.tween the jXLOthec’ s hrotlier as a severely authoritarian_iig.ure and the. .£ather as..a..herievolent protector^ found in m any matrihneal . sQcieties,_. .sjui:h_._as_ the Trobriand Islanders, is not m ade.. by.„N dbm bu. D ue to the partial assimilation o f functions associated w ith th e tw o positions, the m other’s brother has m uch less control and...the. father rather more than am ong the Trobrianders. I have, for exam ple, seen fathers administer beatings to their youn g children, and I have know n sons w ho left the villages o f harsh fathers to stay w ith their mothers’ brothers. B ut the m ajority o f fathers assist their ow n children financially ; help them to m arry ; represent them in co u rt; teach them the skills o f hunting, housebuilding, and the blacksmith’s c r a ft; and instruct them in custom and law . The father is head o f the domestic gardening team in bush-clearing operations, and m ay call on his children to assist him in hoeing his ow n garden. H e takes his sons w ith him to carry his p ro visions and his kills when he goes on hunting trips. There are a num ber o f folk-tales w hich reveal the existence o f tensions in the father-son relationship. O ne tale describes how a father killed his son in the bush w hen the pair failed to find game, and h ow the father was detected in trying to pass o ff his son’s body as the dried carcase o f an antelope. A nother relates h ow a father cursed his son fo r w alking beside him like an equal instead o f behind him like an inferior w hen they w ent honey collecting in the bush. As the result o f that curse the son could not find a single hive. E ventually the lad’s m other persuaded him to treat his father w ith respect and thus induce the father to revoke his curse. A son, other than an infant, m ay not sleep in his father’s h u t; nor, by extension, m ay C h ie f Kanongesha spend the night in the village o f any o f ms senior headmen to w hom he stands in the perpetual relationship o f ‘ father On the other hand, it is said that the father-son relationship is free from mutual sorcery accusations, although the case o f Chibwakata and Kasamba (p. 160) where a classificatory son was alleged by his * father ’ to be trying to bew itch him , seems to be an exception to this rule. I have never heard that a father had been accused o f bewitching his own son or vice versa, although in divination
242-
Schism and Continuity in an A frican Society
paternal kin o f the victim s are occasionally named as sorcerers or witches. I have often witnessed acts o f indulgence and kindness per form ed b y men on behalf o f their sister’s sons. W hen, for instance, the uterine nephew o f one man I knew carelessly wrecked his uncle’s bicycle, the latter freely forgave him , telling me that i f he w ere too severe on the lad, when the tim e came for him to m ake a village o f his ow n his nephew w ould not come w ith him . 4People are m ore im portant than things,’ he added. I have know n men to pay their nephews’ fines in adultery cases. B ut there is no doubt that uncles often punish nephews quite harshly. For exam ple, Sakazao, usually the gentlest o f men, grew exasperated one day in M ukanza V illage b y the persistent naughtiness o f Chikim bu,1 his sister’s son, and twisted his arm out o f join t. Sisters’ children are usually at the beck and call o f their uncles in the collection o f firew ood, the carrying o f food from kitchen to chota, and so on. B u t the father is in the final issue his ow n child’s protector. In Ikelenge Area I once saw a man angrily threaten his sister’s son w ith a beating after the latter had knocked down the form er’s son. T he case cited on page 189, in w hich a father refused to allow his w ife’s brother to sell his children into slavery, w ell expresses the difference between avuncular and paternal attitudes. The marked hostility between uncle and nephew emerges in m any sorcery accusations I have recorded between these categories— most o f them in connection w ith succession disputes. T he relationship o f father (tata) to daughter {mwana wamumbanda) is often one o f great affection ; indeed, I have only once seen a father beat his young daughter. In fact, such an action is roundly condemned by N dem bu w ho say that it is the task o f the m other to punish the daughter and o f the father or m other’s brother to beat the son. M others’ brothers m ay soundly rebuke nieces w ho have been involved in adultery cases, as m ay mothers, but fathers seldom do so. M others frequently beat young children, but once a b oy has been circum cised he comes under the general control o f the men 1 Chikim bu had transferred his allegiance from Sakazao’s group to Sandombu’s group. The relationship between uncle and nephew was strongly influenced by die political struggle between village factions. But the political factor did no m ore than exacerbate a relationship that was already tense.
V irilocal M arriage w ithin the V illage
243
o f the senior adjacent generation and the m other loses her punitive role. B u t m others exert quite a considerable authority over daughters, levyin g their help in gardening, the preparation o f food and the carrying o f w ater supplies. It is the m other w ho receives the greater share o f gifts, as distinct from cash, bestowed for her daughter b y a future son-in-law , and in the past it was for her that her son-in-law used to build a hut and m ake a gardens Jh the context o f m arital relations the m other is sometime, 1 g ? .' a dken o f as the real * ow ner o f the child ’ (mwenimwma ). This Ose and authoritarian bond betw een m other and daughter is *?-' 5 - dos probably strengthened b y the fact that lineal descent passes through w om en. T his bond is n o t broken b y death, since the ritual histories I have collected show that the m other’s spirit is -* <, C-- thought frequently to afflict her daughter w ith reproductive vV,~ troubles o r illness i f the daughter neglects to m ention her name when praying o r to m ake her an offering o f food or beer. Again :fe. ’ a m other’s spirit is thought to afflict her daughter i f she quarrels 14 w ith relatives. In fact, the m other-daughter tie is the second most frequent category o f relationship between afflicting spirit l l ' ;and patient w hich Í recorded. T he father’s sister (1tata wamumbanda or taiankaji) is in certain contexts equated w ith the m other, hut she is treated w ith greater ÜÉ: "'' constraint, since, w ith preferred bilateral cross-cousin marriage, she is a potential m other-in-law . For that m atter the m other’s brother is a potential father-in-law , and this tends to diminish his authority over his sisters’ children i f they m arry his children. A fter m arriage o f this kind, his nephews have little to do w ith him directly, and his nieces avoid him . So for w e have been considering relationships between primary kin w h o call one another aku. ’TKe^dassifcatGr/, extension o f kinship terms. tends_to_dififuseLOver_the. w h olegrou p-ofsen ior aku, the control exerted b y prim aiy km o v e r foeiiL jmiiQrs in the adjacent ^eneration, thus m itigatinn . its. narrowi sev.erity in particular „relationships. This co n tro l .by ,a .who generation o f village kin over.another . w hole.-generation is especially niarked in the case o f m en w h o form .in som e respect a separate, m oral com m unity o f males.1* O ne o f the most
T
*$T t^tS »*
?IHv
1 W om en do n ot form a coherent m oral com m unity since with virilocal m arriage m any arc unrelated to one another and do not belong to the m atriiineal group o f the village.
r r
o
o o o o o o c c
o o o o
244
Schism and Continuity in an A frican Society
im portant functions o f the boys* circum cision ritual o f Mukanda is, in fact, to rem ove boys from the authority o f their ow n individual mothers, in the sphere o f the Idtchen, and to bring them under the collective control o f the m en o f their village, in the sphere o f the chota, the m en’s court and forum* A g irl remains largely under the authority o f her ow n m other until her husband has made the kazundu paym ent, at a variable period after marriage, w hich enables him to rem ove his w ife * from her m other’s knee \ This..is__becanse...thebasic.format.jpf.a._village....consists essentially, o f a set o f structured relationships. betw een,m de„kin w hojconlxol m d obey, one another in-econom ic.andjuralm atters. Another aspect o f this .collective control is-connected_with the conception o f a village., as._a -unit which-transcends- its-narrower com ponents, the uterine siblm g groups .and elem eritary Families discussed above.
o
e o ©
c © © C)
W ithin.eachJhiked-gm erado.n_segm ent.in„aviUage..the„m ost general behavioural characterisrics.ma^ he sh d .tp. be^fajiiiliarity and-equality. B ut the precise...w:dghting,Qf..such characteristics varies in each segment. Im thegrandparent-grandchild segm ent,1 the generation .components are.geneaipgicaliyjunequal. and often differ_considerably in_the mean age_of their_xes.pe.ctive^jnember” ships; This fact tempers, the fam iliarity o f their, intercourse w ith respect, .since. it_is _a.general_principle.jo£.Hdembu-sociaLlife that the _eldest-bom. .is entitled , to the greates.t_respect._B ut these components have. in.-coxnmon the fact-that .both are_.o.pp_osed to the^mtervemng_.genealogieal. generad
PLATE V T h e A l l ia n c e
of
A l t e r n a t e G en e r a t io n s
A ll the children, w ith one exception, playing around Sakazao, the nominated successor o f Headman Mukanza, are daughters’ children o f his sisters ISTyamukola and N yatioli. The exception is the little girl standing behind him and clutching a maize-cob. She is Kasonda’s daughter and is Sakazao’s cross-cousin and joking partner. A ll the children present in fact have a jok in g relationship (wusensi) with Sakazao, their ‘ grandfather ’.
V irilocal M arriage within the V illage
24$
~'■ ineradicable narrowness ofJSldemb.n_corp.otate. group, relations, is a latent source o f tension between them. That tension exists is again revealed b y certain sorcery beliefs. I have been told that at night a sorcerer sends o u t his ilomba : snake-familiar to listen invisibly to w hat his grandchildren l (ejtkulu) are saying about him in private and to report hack to him any expressions o f anim osity. In ritual perform ed to placate female spirits w ho have returned to plague their livin g female relatives w ith infertility or illness, grandm others form the largest category o f punitive spirits. Q n th e other hand, in custom „the grandparents, are the genial i advisefSL.and. instructors m any matters. Especially, is this.the case w here sexinstxuction.is.concem ed. Aku and siblings o f opposite sex m ay n ot speak openly to one another on sexual topics but must em ploy circum locutory speech (kudidyika). B u t this prohibition does n ot extend to grandparents (cmkaka). In feet m any instructresses o f girls undergoing the puberty ritual are * grandmothers \ These w om en (ankongu) instruct their novices inter alia in the techniques o f sexual inter course and in other matters connected w ith sex and reproduction. Grandparents and grandchildren m ay sleep in the same hut and witness each other’s sexual activities. Y o u n g children w ho are considered too old to remain in their parents’ huts at night for this reason are sent to sleep w ith their grandparents, sometimes in the same bed. Thejrelationship. between, grandparents, and. grandchildren*, like that between cross-cousins, is summed up in the. iSTdemb.u term w usensV' Wusensi m ay be translated as ‘ jo k in g .relationship *. Joking partners m ay revile one another, excitin g amusement but not anger. T h ey m ay claim any articles o f one another’s property they m ay fancy or call on one another’s assistance ,in.,work^with out paym ent., or. return.—Joking -partners -of-opposite-sex- may indulge_in sexual.play, even-in. public. B ut. the jo k in g relationshipjw hich exists between grandparent.and-grandchilcLis-of a mild and~tesfrained type, and tends to. b e asym m etrical. T he grand parent tends to initiate jo k in g activities such as reviling, o r seizing food carried b y the grandchild. A gain, although a young child is often allow ed to take considerable liberties w ith his or her grandparents, w hen the child reaches adolescence he o r she must behave w ith greater respect towards them. „The relationship
;o ■o o o o o o o o
v./ V.../
o o o G
c o c
C)
© G
•©
246
Schism and C ontinuity in an African Society
he!3»een.grandfether ,.aiid„. grm 4
V irilocal M arriage w ithin the V illage
247
tion. Contradiction exists in die relationship in that a grandson w ho is regarded in one sense as a younger brother is at the same tim e a potential brother-in-law . Sim ilarly a granddaughter w ho, as the sister o f a kind o f younger brother, m ust be regarded as a sister, is also a potential w ife. B ehavionrassociated w ith these respective ro les.is contradictory. ^ T h e cornequent confusion is one o f the. elements in the jo H n g relationship .w h ich is nearly alw ays„a coincidentia oppositorum. This^confusion and contradictoriness-seems...£a_m£-.ta_brmg out the essential nature o f the principle o f organization by genealogical generation as a com prom ise form ation.betw een the lineal, and fam ilial modes o f organization. I f .paternal and maternal .kin are partially, merged, as. m em bers o f a sin g leg en ea logical generation it .is possible to preserve-their distinctive kinship„eharacteristics in that gen eration ; but in the. follow ing generation, given the continuing -strength ~o£ -the—generation principle, such distinctions becom e-blutred. In a society w ith strong . lineage organization, one aspect .of., w h ich_ is., lineage exogam y,-m em bers^of-tliegenerationsjuniorto. ego. are.divided into tw o categories : (1) Lineage fe llo w s.with.,.whom..copulation., is- incest. (2) Others, certain o f whom , are -potential-spouses. B ut am ong _Ndemb.u,_members.o£generatfohs-junior- to ego are divideddnto tw o other categories : (1) Thos£ in adjacent generations, w ith whom -there is hostility, fear,-form ality, authority, etc. (2) Those^in...alternate generations,, w ith .w hom ..there,is in form ality, trust,. equality, .. and.the... like, „and. w ho are potential spouses except fo r ow n children’s .own children. M arriage w ith one’s ow n sister’s daughter’s daughter is frow ned upon but sometimes occurs, as in the case o f Nswanamatempa, w ho married his ow n sister’s daughter’s daughter N jonka (III,G l ; see Appendix HI). B u r the generation principle, trium phs.over the lineagepsrinciple to the extent that a person may. marry-a..q.uite close JmeaLkmsman o r kinswoman, belonging to.-an. alternate-generation. O n the other hand, one m ay not m arry a com paratively ,distant paternal or cognatic kinsman or kinswom an o f an adjacent generation.
248
I | j! [ j | I j [, !
Schism and Continuity in an A frican Society
The jo k in g relationship between alternate, generation kin is made: up. o f ..two m ajor components— One. such com ponent is that in-effect a person m ay m arry someone regarded.as.his. or_her classificatory sibling ; fo r true siblings this is tabooy.yet hereLit is perfectly legitim ate, to do so. Embarrassment is- resolved by laughter. T he neighbouring Lwena tribe regard this Ndembu custom w ith as m uch abhorrence as Ndem bu display towards the reputed Lwena custom o f wife-sharing. T he ..other com ponfcnt. consists in the. fact that two. categories, o f .genealogically asymmetrical kin are regarded a- equal and allied, in contradiction to the general prin cip le. that seniority -by birth or genealogical order_confers authority. Thus one’s equal ,is. a t the same time one!s.inferior, a paradox again resolved b y .mutuaL-joking. In both linked-generation segments in a village one o f the tw o com ponent categories consists o f prim ary and classificatory sib lings. In m any contexts siblings are equated, in others they play opposed roles. T h ey are equated as members o f a uterine sibling group, the basic unit o f fission and nuclear unit o f a new settlement. T he term mwivumu detu, * in our w om b is most often applied to such a group, the children o f a single m other, although it m ay be extended to all w ho claim descent through real or putative matrilineal links from a com m on ancestress or even to those w ho w ithout remembering the name o f that ancestress claim maternal relationship w ith one another. B ut basically mwivumu detu refers to one’s ow n sibling group. There is no general term for * brothers \ only terms for ‘ older brother ’ (yaya) and 4 younger brother * (mwanyika). A man does not discriminate between an older or younger sister, but calls both by the term muhela. O lder and younger sisters distinguish one another by the terms yaya and mwanyika respectively. A wom an calls both older and younger brothers by the term manakwetu. Kwetu is the locative, not the possessive, form , and means 4at our place ’. This is interesting, since it denotes that a wom an regards her hom e as being where her brother is residing. The plural form * kwetu \ 4at our place,* suggests that the sibling group is regarded as a spatial unit. Sometimes a wom an w ill refer to her brother, whether older or younger, b y the term yaya w hich is usually applied to an older sibling o f the same sex as ego . This implies that she recognizes his authority, since w ithin the genealogical generation seniority goes w ith prim acy by
ft-.
V iriiocal M arriage w ithin the V illage
249
birth. In the past I am told that a wom an used to call her brother iyala, literally * man ’, a term now sometimes used as a synonym for ‘ husband * (mfurnu). W ith in a genealogical generation seniority in age and not in genealogical position determines whether parallel cousins call one another '‘ old er’, or * younger sibling \ O n the other hand, sublineages o f a village matrilineage are rated as senior (ivumu Jamukulumpi) or ju n io r {ivumu dakanst) according to the sibling order o f their apical ancestresses. Thus one m ight find a man in a senior lineage addressing his parallel cousin in a ju n io r lineage as ‘ older brother ’ since the latter was the earlier born o f the tw o. Hence in each linked-generation segm ent there is a contradiction among lineal kin o f the same sex between the principles o f genealogical and chronological seniority. B oth principles are invoked b y candidates for village headmanship. Nevertheless,..despite the.. close -and,„persistent_ ties~..that unite members, o f .a single.... uterine, sibling, group,..pow erful..tensions between its m em bers exist....The. closer ,the .tie, .the_greater the am bivalence o f feeling. Betw een older and yo u n g eru terin e brothers..mutual, hostility ...is revealed., b y an... analysis...o£.soxcery accusations. B y adelphic succession a younger..brother m ay succeed to headmanship on the death o f his.Qlder .brother so that he is a .natural target fo r suspicions w hich , themselves , m ay tend to influence the result o f a ..divination. A... further source, o f hostility resides in. the fact that an .old er .brother is .regarded b y N dem bna'shis-father’s-lieutenant.witliin.the__siblingL.group. He m ay be delegated b y his father w ith the task o f punishing a younger brother or o f admonishing him severely for bad be haviour. In the boys’ circum cision ritual, this role o f the older brother is dram atized ; either a father or an older brother is considered to be the m ost suitable person to act as * shepherd * (chilombola) to a novice. Part o f a shepherd’s duties consist in administering punishm ent to his charge fo r breaches o f lodge customs, on the instructions o f the senior official o f the seclusion lodge (mfumu watubwiku), w h o is him self usually the father o f the greatest num ber o f novices. A n older brother can command the services o f his younger brother in, for exam ple, the carrying o f loads, the collection o f firew ood, hut-building, and many other w ays. B u t ju st as the shepherd at circum cision protects and attends to the wants o f his charge, so also does the older
ft
a © © 0
©
©
© ©
© ©
€1
© ©
©
© © © © m
© #
o ©
© ©
n
o o O
o o
o o o o o o Ç)
o O
o o 0
o o 0
o 0
c
0
.0 o
250
S ch ism and C o n tin u ity in an A frica n S o ciety
brother give protection and assistance to his younger brother in the affairs o f life. O nce when Daudson M w evulu’s (I,H7) younger brother S ig n 1 was insulted at ah1Nkula ritual as being too young to take part in the popular m odem chikinta dance which today accompanies most rituals, and in w hich the younger element dom inantly participate, M w evulu im m ediately started a fight w ith Sign’s detractor and then knocked dow n the latter’s father w ho had intervened on his behalf. In m ore form al matters older brothers w ill assist their younger brothers w ith bride-wealth, w ith court fines, and w ith deathpayments, and w ill share their food and possessions w ith them w ithout question. As brothers gro w older their relationship becomes m ore egalitarian and reciprocal and often som ething like a division o f skills emerges between them. Thus Mukanza Kabinda became a headman w hile his younger brother Kanyom bu acquired a reputation as a ritual specialist and herbalist. In another village M ulila, the younger brother, acted as headman, w hile Itota, the older brother, devoted m ost o f his tim e to hrmting. In Ikelenge Area Nswanandong’a was at once headman o f a village and a hunter w hile his younger brother M akw ayang’a participated fu lly in. the cash econom y at various times as a casual labourer, kapasu or N ative A uthority policem an, and cash-crop farmer. U terine brothers usually spend the w hole o f their lives to gether. M others and sons m ay be separated b y the divorce and rem arriage o f the form er. Sisters are divided from brothers by virilocal marriage. Children m ay be separated from fathers by divorce, sex, matrilineal pressures and vim ocal m arriage. Uterine brothers are united b y descent and generation. Thus the link between brothers is deep and abiding, despite the possibilities o f conflict inherent in it. W hen a man founds a new settlement his m ost faithful supporter is his brother, for their sisters have another lo yalty— to their husbands. In die relationship between parallel cousins the ties are less close and the cleavages m ore clearly evident. Parallel cousins belong to different matricentric fam ilies and m ay have been reared in different paternal villages. W hen different groups o f uterine siblings build together and becom e co-m embers o f a 1 Manyosa’s sons in Mukanza Villagè.
V irilocal M arriage w ithin the V illage
251
village genealogical generation, the divisions initially existing , between them are n ot im m ediately or easily overcom e. Even although it is custom ary w hen m eat is divided to giv e an equal share to all kin w ho bear die same kinship term, in practice people tend to favour their closer at the expense o f their m ore distant kin and this becomes a source o f jealousy and back-biting, especially between sibling groups. Sim ilarly, although it is expected that close kin should contribute and receive the bulk o f marriage paym ents, fines, etc. and that distant kin should contri bute and receive lesser amounts, m uch w rangling actually takes place between tw o o r m ore uterine sibling groups w ithin a minimal or m inor m atrilineage about the precise sums given and received. B ut it is w hen succession to office drops from the last living member o f the senior village generation to the ju n io r adjacent generation that relations between its com ponent uterine sibling groups tend to becom e m ost strained, it w ill be remembered that in M ukanza V illage the relationship m ost fraught w ith tension was that between Sandom bu and Kasonda, parallel cousins belonging to different sibling groups. It was pointed out in the analysis o f Social D ram a I that according to N dem bu opinion succession to office in a chieftainship o r in a w ellestablished village should not be confined w ithin a single narrow span lineage, so that a man was succeeded b y his uterine nephew. In point o f fact, how ever, a headman’s ow n sister’s son often strongly urges his claim and this brings him into conflict w ith his male parallel cousins and also w ith their siblings w h o support them. This kind o f conflict probably influenced the secession o f Nswanakudya (II,D io) from Shika V illage w hen his parallel cousin seemed lik ely to succeed, and that o f M atem pa (III,D3) from N sang’anyi V illage, w hen the present headman did in fact succeed to headmanship (see Appendices II and III; also p. 208 and p . 2 1 2 ).
The tie between brother and sister is also a very close one, and like the preceding tie, contains several possibilities o f tension. Brothers and sisters are frequently reared together in their father’s village, but the tie o f spatial propinquity is usually broken b y die virilocal m arriage o f sisters after puberty. Brother and sister are linked b y descent and generation, but are divided b y sex. In certain respects the relationship o f brother and sister resembles that between aku (in-laws) belonging to adjacent generations.
252
Schism and C ontinuity in an A frican Society
Strong taboos against incest divide them and introduce constraint into their behaviour towards one another. N either m ay speak overtly o f sexual matters to the other. O n ly as very small children m ay they sleep in the same hut. A man was once strongly suspected o f sorcery w hen on a visit to his m arried sister he sat on her bed and it broke beneath his w eight. T o obtain a peculiarly potent hunting m edicine a man must com m it incest w ith his sister, and the suspected possession o f such m edicine by a hunter causes him to be regarded w ith abhorrence. One N dem bu oath believed to be pow erfully binding in a village court has the form , * I f w hat i say is untrue I have had sexual relations w ith m y sister/ This is considered to be a statement o f the impossible. Y e t in spite o f this incest barrier sisters are equated w ith w ives in several contexts. A t a chief's installation ritual, fo r exam ple, the chief-to-be is secluded during the night before his accession either w ith his senior w ife or, i f she is in a state o f pollution due to her pregnancy or menstruation, w ith his sister. Again, in the Nkula ritual, perform ed for w om en w ho are thought to be afflicted w ith menstrual disorders b y the spirit o f a deceased relative, the patient is escorted from the small spirithut built behind her dw elling-hut to the ritual fire b y a male ‘ helper * (chaka chaNkula), w ho m ay be either her husband, her brother or her son. A t the boys* circum cision ritual either the m other or a father's sister o f a novice m ay cook for him at the sacred fire (ijiku daMukanda). In their ow n generation the principal conflicts o f interest between brother and sister spring from marriage and the fam ily. A brother desires his sister's children to sw ell his follow ing, but, ow ing to the strong ties that interlink members o f a matricentric fam ily it is difficult to fu lfil this wish unless his virilocally resident sister is divorced from her husband. I have recorded several cases in w hich a husband demanded compensation from his w ife's brother in. a N ative C ou rt for retaining her in his village when she had paid him a visit. O n the other hand I have heard complaints from brothers that their sisters have remained w ith their children in their husbands' villages so that the brothers have been unable, though senior men, to found their ow n settle ments. W hen brothers and sisters inhabit the same village, disputes sometimes arise between them over their respective groups o f
V irilocal M arriage w ithin the V illage
253
children. Sisters accuse brothers o f neglecting their nephews and nieces in favour o f their ow n children. Quarrels arise between brothers and sisters over the distribution o f bride-w ealth received for daughters o f the latter. I have often heard sisters bitterly scold their brothers fo r failing to assist them w ith m oney, and for givin g cloth to their w ives but not a rag to their sisters. The sister has considerable pow er in this relationship, and her threat to leave her brother’s village often has the effect o f bringing him to heel. For exam ple, C haw utong’i o f Mukarxza V illage, in censed that Kasonda, her first parallel cousin, had n o t bought clothing for her younger children, le ft the village fo r a w hile to stay at M bim bi V illage, a short distance aw ay. She w ould not return until he had made up a dress fo r her daughter on his sewing-machine. Despite the possibilities o f discord inherent in the relationship between siblings, w ithin each linked-generation segm ent prim ary and classificatory siblings, m atrilineally related to the headman, confront the other members o f their generation as a unitary group. Thus in the senior segm ent the headman and his matriHneal generation mates confront the adult members o f the * grandchild ’ category, w h o are usually num erically inferior and m ay be divided am ong themselves b y varying maternal and paternal affiliations. Some m ay be m atrilineal members o f the village, others seminal children and children’s children o f male village members. In the ju n io r segm ent prim ary and classifica tory m atrilineal siblings confront their cross-cousins, the seminal children o f male village members o f the senior segment. Since the m atrilineage is to some extent a persisting corporate group w hile the tie to the father is personal and ephemeral, in this segment also the m atrilineal sibling group tends to have greater internal cohesion. This group is linked to the co-members o f its segment b y cross-cpusinship. Cross-cousmship, like the grandparent-grandchild relationship, is a jo k in g partnership. Perhaps on account o f the egalitarian nature o f the tie betw een cross-cousins in N dem bu society, the jo k in g is o f a m ore boisterous typ e than that betw een kin belong ing to alternate generations. It is sym m etrical, sexual in content, derogatory in manner, and often contains reference to the sorcery or w itchcraft o f the jo k in g partner. Cross-cousins m ake free w ith one another’s possessions, those o f opposite sex indulge in
254
Schism and Continuity in an African Society
sexual play, and cross-cousin marriage is the preferred form o f marriage. N dem bu m en appear to m arry their patrilateral and matrilateral cross-cousins w ith equal frequency. B ut a distinction must be made between the kinds o f behaviour that typically occur between close and distant cross-cousins. Cross-cousins w h o inhabit the same village behave to one another m ore as though they w ere brothers and sisters than jo k in g partners. Their relationship exem plifies the statement previously made that in the village there is a convergence between lineal and fam ilial principles o f organization, and that the organ ization b y genealogical generation represents a compromise between them . As members o f a com m on generation crosscousins are united in opposition to the adjacent generations. As children they are jo in tly under the authority o f the senior adjacent generation, as adults they jo in tly exert authority oyer the jun ior adjacent generation. Together they co-operate m communal working-parties at the beginning o f the gardening season. Male cross-cousins m ay bunt together, female cross-cousins m ay pound in the same m ortar. Their fam iliarity in daily tasks and the sim ilarity o f their structural position in the organization by generation tend to reduce the element o f unlikeness derived from their distinct lineal affiliations. I f they lived in different villages the feeling o f dissimilarity w ould be m uch stronger. U nfam iliarity and hostility are components, o f.th e r a ille r y between jo k in g partners. Cross-cousins m a y . be regarded as unlike and equal. Relations, between siblings m ay be described as parallel relations. Siblin gsare regarded zs.equivalent.to one another. T h ey are alike and equal. V illage cross-cousins also tend^to interact in parallel rather than in „com plem entary and opposed relations. T hey tend, .in fact, to. behave...to„Qne. another like siblings. This m ay result in a reduction_of_joking and perhaps in a dampening dow n .o f the._sexual_attraction between opposites. Cross-cousin m arriagebetweem .closeJdum isrelatively infrequent. O n the other „hand marriage between persons who call one another cross-cousms..-.(dLyo»ytl,_but--who^cannot trace precise relationship between one. another, is. quite-com m on. This kind of. m arriage supports m y jprevious stetement tha^villages w hich claim com m on matrilinea£descent.soon.fbrg.et.their precise m am lineal link but can still class. one._another!s-members b y genealogical, generation. Thus. the. m e m o f .one- village know
Virilocal Marriage within the Village
2 $$
roughly w hich w om en in the other village fall into the marriage able category and. w hich w om en fall into the forbidden categories. Am ong the Lam ba o f the N dola D istrict o f N orthern Rhodesia whom M itchell and Barnes found to possess a h igh frequency o f virilocal m arriage (38-1 per cent o f extant m arriages), and among w hom , therefore, cross-cousinship m ust constitute an im portant ingredient o f village kinship, prim ary and traceable cross-cousin marriages form ed an extrem ely lo w proportion o f the total marriages recorded (7 prim ary, 2 classificatory crosscousin marriages out o f 360 extant and dissolved marriages, or 2'5 per cent) -1 It m ay w ell be that am ong m atrilineal peoples with shallow local lineages and virilocal m arriage an identification tends to arise betw een the children o f siblings o f opposite sex on the basis o f co-m em bership o f a genealogical generation w hich inhibits frequent intra-village cross-cousin m arriage. A m ong Ndem bu, at any rate, there is a m arkedly greater tendency to m arry distant than close cross-cousins. A nd it is a matter o f observation that jo k in g behaviour reaches its fullest expression between distant cross-cousins, and betw een cross-cousins, whatever their genealogical distance, w ho norm ally reside in ^ w idely separated villages. JSach. village. contjdns.„.g_s?pa^£e arSEf^** autonomous system o f kinship relations,.jcon&txuedjcm ..^^^j® t^^! tional principles w h ich m um aU y.m odif^ one_ anQ tbei:^ Sl^ % ^ l&t^ principle—then, achieves.....its..m axim um - expression.— Concrete behaviour is their mean w o rk in g .. In specific_siimtLQns,, how ever, one or other m ay become, dom inant Marriages_do. .in fact take place between village cross-cousins.... -When, children build leaf-huts in the h u s b a n d -p la y athusbands_and—w iv e s-9xrosscousinsdbut n ot brothers ^ a sisters jxume__foe_marital_roles, In other kinship categQries_we..have-seen^examples-o£.die,.kinds o f situation in w hich, fo r instance, fatherss tancLopposecLta .mothers’ brothers; and children _to. nephews. JBnt-imorder that the village m ay be a-going, concern, and a viable..:unit,_.it_is. .necessary that confficting nm tiilineal. and patrilateral loyalties_shauldLb.e recon ciled. O rganization b y genealogical... generation., is...Jthe__major means by.w hich. this reconciliation is effected. 1 Mitchell and Barnes, The Lamba Village, p. 41.
256
S ch ism and C o n tin u ity in an A fr ica n S o ciety
Summary Virilocal-..mamage..imites a . nunib_er o£ narrow-spaii .maternal descent groups within the bilateral extended fam ily, to form the locaLunit o f residence. Each such descent.gro.up. is. a. matricentric fam ily w hich in the course o f the social, process strives-to dis engage itself. from all .other ties of--local-affiliation- and-achieve residential autonomy. It is—possible, to..regard most other categories o f local kinship ties as. co.unterbalancing..the .centrifugal tendency o f the matricentric -family. I n .its early—stages the matricentric fam ily is typically linked through.its_apicaLmember, the mother, to a village lineage to which, h er husband-belongs. Its members then have a personal tie o f loyalty to their father, to tbe-bilateral extended family. (Kto«^,ii.) .ofwliich-he-is-a. member, and to. his matrilineal .village itself, .regarde
© V irilocal M arriage within the V illage
257
0
matrilineages and o f the village as a w hole. V iriloral marriage w h ich w as a tth e root o f their in tcn scco h esio n n o w operates to separate-the. siblings b y sex. Nevertheless, asrwe have seen, the tendency to jo in t secession evinced b y the g ro u p -o f- mature uterine jsiblings frequently asserts itself-to retard the grow th of* large villages.
© ©
©
© ©
© © © &
© €>
©
#
© © ©
C H A P T E R IX P O L IT IC A L A SP E C T S O F KINSHIP A N D A F F IN IT Y "E have already discussed the w ay in w hich ties o f maternal descent are utilized to establish links between villages in separate vicinages. Such ties have a political function. Because ties o f kinship, w hich are often coterminous w ith ties o f friend ship, and o f econom ic, ritual and jural co-operation, enjoin coflaboration and are almost always means to establish residential affiliation, they have a high political value in a decentralized, m obile and unstable society. Even m ore im portant in this respect are affinal ties w hich interlink exogam ous and potentially conflicting groups. In the preceding analysis o f internal village relations, w e have seen h ow intra-village marriage in a longestablished village allies uterine sibling groups, belonging to the same generation, by means o f cross-cousin marriage, and inter links uterine sibling groups belonging to alternate generations b y grandparent-grandchild, marriage. In this case kinship ties, both lineal and cognatic, are supplemented b y affinal ties w ithin the settlement. In this chapter links o f in-lawship between villages w ill be examined as a mechanism for establishing political relations between them. T w o m ajor aspects w ill be considered : (i) the effects on inter village linkages brought about by the particular form o f marriage am ong Ndem bu, and (2) the range o f such ties, whether they interlink villages w ithin a vicinage or whether they associate separate vicinages w ithin a connuhium. Table X X , based on a sample o f 165 extant marriages in m y V illage Census about w hich I was able to obtain inform ation from both m arital partners, shows to w hat extent virilocal marriage is the prevalent form . N early seven out o f ten mar riages w ere virilocal as com pared w ith seven out o f a hundred w hich w ere uxorilocal. Som e uxorilocal marriages represented the first stage o f a first m arriage in w hich the husband was still conform ing to the tradition o f initial uxorilocality. Others, like the marriage o f N deleki in M ukanza V illage (see pp. 101-2), w ere special cases in. w hich the husband had abandoned his matri258
P olitical A spects o f K insh ip and A ffinity
259
lineal village on account o f fears o f w itchcraft or because o f serious disputes w ith his kin. T h e structural effects o f virilocal m arriage w ithin the village have already been discussed in Chapter Bight» w here it was found that a relationship existed between this forn i o f marriage» the close cohesion o f the uterine sibling group, the shallowness o f the m atrilineage, the organization b y genealogical generation within the village, the tendency towards the m erging o f fam ilial TABLE X X R
jeseoence o f
H u sba n d
and
W
if e
Sample of 165 -extant marriages : Residence
N o. in sample
Percentage
Virilocal . . . . Intra-village . Uxorilocal . Other . . . .
ns 27 12 11
697 i 6‘4
Total
16$
IOO'O
.
.
.
.
7 ‘3
6-6
and lineal principles o f intra-village organization, and the high rate o f individual m obility. B etw een villages its principal consequence is to mesh together b y ties o f affinity, w inch ulti m ately giv e rise to links o f consanguinity, a num ber o f spatially distinct settlements, vicinages and even senior headmen’s areas. M atriJiny is the foundation o f each village’s continuity in tim e ; matrilineal descent is therefore the dom inant m ode o f attachment b y kinship. V irilocal m arriage scatters the nuclear w om en o f the m atrilineal kin-group o f each village through m any other villages, so that the dom inant bond o f kinship is utilized not m erely to consolidate the local unit, but to strengthen the ties between separate villages. Thus each village contains not on ly its ow n nuclear group o f m atrilineal kin but also a num ber o f m atricentric fam ilies belonging to the nuclear matrilineages .o f other villages. Conversely, through the virilocal m arriage o f its ow n w om en, the nuclear m atrilineage o f each village at any one tim e has a num ber o f its cells, consisting o f m atricentric
260
Schism and C ontinuity in an A frican Society
families lineally related to it, dispersed through other villages. It is as though each village entrusts other villages w ith the task o f fostering its ow n future membership and simultaneously brings up the potential personnel o f other villages. B ut there is no precise correspondence between the villages to w hich each village exports w om en and from w hom it receives w om en. In other words, there is no m arked tendency towards the form ation o f connubial clusters o f villages w hich constitute closed inter m arrying groups. Interlocking ties o f affinity connect separate chiefdoms. Kanongesha’s Ndem bu interm arry freely with Musokantanda’s Kosa, and w ith Kazem be’s C ongo Lunda. And w ithin Kanongesha’s chiefdom , senior headman’s area is inter w oven w ith senior headman’s area b y the busy shuttle o f virilocai marriage. I f w e consider each remembered matrilineage separately w e find that it has a local nucleus, consisting typically o f a m inor lineage in w hich m ale members outnum ber females by about tw o to one, and a w id ely scattered peripheral member ship in w hich the proportions tend to be reversed (if absentees, both m ale and fem ale, at urban centres are excluded from the reckoning). In certain matrilineal societies that practise viriloca! marriage, such as that o f the Trobriand Islanders described by M alinowski, it is custom ary fo r the sister’s son to return to his m other’s brother’s village on attaining puberty or social m aturity as defined by other criteria. N o such custom exists among Ndem bu. A son w ill remain w ith his father after attaining man hood i f he obtains any m aterial advantage from so doing, or i f he has a strong tie o f affection, w ith him . For exam ple, Gideon o f M ukanza V illage, son o f the headman, prefers to remain w ith his father, even although his m other, from Shika V illage, has been dead for several years. Gideon is in his m iddle thirties and says that he w ill stay in M ukanza until his father dies. Daughters w ill often return to their fathers’ villages after divorce, especially i f their mothers also reside there, instead o f taking up residence w ith their mothers’ brothers. This duality o f residential choice results in a state o f endem ic and barely concealed conflict between a man and his w ife’s village matrilineal group, w h o w ant his w ife and children, and between him and his sister’s husband’s village lineage, w ho seek to retain his potential local follow in g and heirs. This conflict o n ly becomes acute as the children and sister’s
P olitical A spects o f K in sh ip and A ffin ity
261
children gro w up, but it exists at every stage o f a marriage. There is really no w ay finally to resolve the situation except b y divorce o r death. In die final issue village continuity depends on m arital discontinuity. In practice, o f course, since each village is in effect a set o f variously linked uterine sibling groups w ith attached individuals, who m ay be kin o r strangers, and since the units o f affinal conflict are uterine sibling groups, losses in som e groups are compensated by gains in others. For a tim e indeed m arital continuity is an indispensable condition o f village continuity, since children replace the lost sister’s children, both in the uterine sibling group and in the village. B u t in the end, i f the sisters” children do not return, the village w ill perish as a m atrilineal descent group. The village o f M bim bi, fo r instance, d ose neighbour o f M ukanza, was, in 1954, in a fair w ay towards becom ing extinct, since M bim bi’s sisters* children w ere either livin g w ith their fathers or had gone to other villages, his o n ly livin g daughter was hving virilocally, and his son had left him to reside w ith his m other s brother fo r fear o f the w itchcraft o f his father’s senior w ife, w ho was thought to have killed several children in the village. Each village is linked to m any other villages b y single marital ties and in this w a y it extends the geographical range o f its political interconnections. O n the other hand, one finds the not infrequent establishment o f a relationship o f reciprocal marital exchanges betw een tw o villages. T his arrangem ent aims at reducing conflict since each o f the tw o in term arrying groupings rears the future membership o f the other. It is in the interest o f both villages to m aintain in being the marriages that interlink them , since the breakdow n o f a m arriage in one village m ay lead to the exertion o f pressure b y the m atrilineage o f the other on their virilocally resident kinsw om an to leave her husband and bring her children to her o w n village. Nevertheless, as w e shall see, m arital ties, far from establishing enduring friendship betw een villages, m ay g ive rise to frequent disputes between them . fii order to Bring out the political function o f inter-village m arital linkages, the field o f affinal relations o f a single village w ith other villages w ill first be considered, follow ed b y an analysis o f the ties o f affinity and o f kinship affiliation, inaugurated
n ;0
O O
262
o
b y inter-village marriage, existing between the headmen o f com ponent settlements in a single vicinage.
o o o
•O "O
o o
■ Q o
o
o
■ c [o:
Schism and Continuity in an A frican Society
Since I have analysed the internal structure o f M ukanza Village in some detail, I shall discuss the external relations o f the same village constituted b y affinity in similar detail. In m ost respects this village is typical o f the consolidated and traditional Nderabu village. I regard all livin g members o f nyach intang ’ a and m alabu lineages, together w ith N yaw unyum bi and her children, o f n yach ula lineage, w ho have made their hom e at Mukanza, as * village members \ B oth extant and fruitful completed marriages, whether com pleted b y death o r divorce, between village members and members o f other villages w ill be con sidered.1 Persons w ho w ere absent in urban areas in N ovem ber 1951 w ill be excluded from the investigation. R eciprocal m arital exchanges have taken place between M ukanza and Shika Villages for several generations. Shika V illage xTxatxilineage belongs to the w ide m atrilineal descent group from w hich the incum bent o f die Kanongesha stool is elected. There is a distinct tendency for villages w hich par ticipate in the Kanongesha chieftainship to interm arry w ith K aw iku villages. Thus Shika V illage and its offshoot Nswanakudya V illage tend to interm arry w ith N sang’anyi, w ith Kasai, and w ith M ukanza Villages, all K aw iku settlements. Sim ilarly there is a tendency for Chibw akata and its offshoot N g ’ombi V illage, both o f Lunda origin, to interm arry w ith K aw iku villages. B oth Shika and Chibw akata refer to themselves as * the husband ’ (1mfunrn) o f N sang’anyi. Nsang’anyi, on the other hand, calls him self * the husband * o f Chibw akata. In actuality, although both parties, the one descended from the original invaders, the other from the autochthonous people, claim superiority over one another b y the use o f the term * husband \ both are able to m ake this claim b y virtue o f the fact that the chain o f reciprocal interm arriage began w ith an exchange o f sisters between the heads o f the respective groups. ¿1 one sense, therefore, the reciprocal m term arrying between M ukanza and Shika V illages is an aspect o f the political rdation1 See M ap 6, M ukanza Village M arriages, for spatial range o f extant mar riages and fruitful completed marriages o f m atrilineal membership.
P olitical A spects o f K in sh ip and A ffinity
263
ship between Luncfa and K aw iku. B ut, m ore im portantly it, is a relationship between tw o N dem bu villages. A t one tim e Shika V illage belonged to the same vicinage as M ukanza V illage, and it was at this tim e that M ukanza Kabinda (I,F8), then, a young man, married N yatungeji (II,D3), sisters daughter o f Headman Shika Ikubi (II, C z ). In the same year N yattm geji’s brother Kahum pu (II,D i ; see pp. 190-1) m arried M ukanza Kabinda’s sister N yam w aha (I,F7 ; see p. 118 et seq.), thus interlinking n ot only the tw o villages but also the uterine sibling groups o f M ukanza Kabinda and Kahum pu b y reciprocal interm arriage. Although I have not been able to trace the exact genealogical connection, M ukanza Kabinda called both N yatungeji and Kahum pu his * grandchildren * (ejikulu), w hich suggests that previous interm arriages had taken place between the tw o villages. M ukanza Kabinda’s m arriage was term inated b y the death o f N yatungeji in 1943, and the circumstances associated w ith her death became a source o f perennial discord in the relations between the tw o villages. W hen a death occurs am ong N dem bu the lineal kin o f the surviving spouse have to pay the lineage o f the deceased w hat is fo r them a substantial sum, called mpepi, o r * death-paym ent ’, in com pensation fo r the loss they incur o f his or her labour potential, or in the case o f a w om an, o f her fertility in addition. Mpepi also includes paym ent for various ritual services perform ed b y the village lineage o f the deceased for the surviving spouse, the m ost im portant o f w h ich is to rid the latter o f the spirit o f the dead spouse {ku-fumisha mukishi windi) w hich w ill otherw ise d in g to the survivor and afflict him or her w ith bad dreams and illness. Such a spirit m ay also jealously attack the survivor’s next w ife or husband w ith disease. N o w in order to avoid paying m pepi1 it frequently happens that a man o r wom an m il divorce a spouse w h o is fatally ill. T he lineage o f the deceased cannot then claim that the death occurred w hile the m arriage was extant, and that therefore according to N dem bu notions responsibility for the w elfare o f the deceased was vested in the surviving spouse and his o r her village lineage. T he practice o f divorcing an ailing spouse has undoubtedly distorted ' r-fr* ;il V
1 It should be m entioned that N dem bu villagers believe that the payment o f mpepi is forbidden under a N ative A uthority regulation, but they tacitly agree to maintain the custom clandestinely. B u t i f a sick spouse is divorced there is no remedy at law .
z> S 2 t _____
f
CHINGONGU—
Uns*. ncuBi
vT MATQROKOSHI
IUP* WILLIAM
W o P o litic a l A sp e cts o f K in s h ip an d A ffin ity
265
the divorce ratios in Table V II, since a significant proportion o f the divorces recorded must be in reality m asked w idow hoods. W h en N yatungeji seemed to be on the point o f death, M ukanza Kabinda divorced her, and carried her, w ith the aid o f kinsmen, back to Shika V illage on a stretcher. H e alleged that someone am ong her m atrilineal kin in Shika V illage had bew itched her * in order to eat the mpepi hinting that Headman Shika him self was to blam e. H e m en affected“ to take um brage because the m arriage-paym ent (nsewu) he had made for her in cloth was not returned. In the long process through w hich an N dem bu m arriage becom es consolidated a num ber o f paym ents in cash and kind are m ade. T h ey include the small chijika muchisu 4w hich closes the door * (today about 2s. 6d.) w h ich signifies b etroth al; (he kazundu, 4taking a w ife from her m other * (today about 5s.), w hich gives the husband his righ t to take his w ife to his o w n v illa g e ; and the m ost im portant, the nsewu, 4the arrow ’, the marriage paym ent proper, form erly paid in cloth, livestock o r a gun, but n o w in cash. Nsewu is generally regarded b y N dem bu as a precaution against a d u ltery; in formants have com pared it w ith a fence round a cassava garden to keep out w ild pigs. I f a man has paid nsewu fo r a wom an he is entitled to com pensation from his w ife’s lo ver both in a N ative A uthority C o u rt and in a village court, i f adultery is proven. M oreover, a man is entitled to the return o f his nsewu on divorce, w hichever party is adjudged culpable. I f he wishes to retain w ith him im m ediately one o r m ore o f his children he w ill not seek the return o f his nsewu from his w ife’s kin. Again, i f he wishes to m aintain friendly relations w ith his w ife’s village kin w h o w ill help to lo o k after his children, he w ill ‘ leave the nsewu w ith them \ In the case o f N yatungeji no death-paym ent was m ade and no m arriage paym ent was returned. M ukanza’s tw o surviving children, including Gideon (see Social Dram a IV , p. 148), b y N yatungeji, rem ained in their father’s village. A lthough no o vert breach o f social relations took place between the tw o villages, an undercurrent o f hostility remained which manifested itself in subsequent situations. Kasonda (I,G l5 ), son o f Kahum pu (II,D i) and N yam w aha (I,F7), m arried his father’s classificatory sister’s daughter M angaleshi (II,Ea), o f Shika V illage lineage. She is know n as
o o
© © 0
© © ©
©
©
©
©
©
© ©
©
© © ©
©
,!i ■ D
D O
n :o
G ;D t)
.G ;C G
;0 G f.^
266
Schism and Continuity in an A frican Society
his cross-cousin on both Other’s and m other’s sides,1 has borne him five children, and their marriage is a stabilizing factor in the relations between the tw o villages. Kasonda’s older daughters have both acted in the capacity o f ritual handmaids {tunsonselelu) to girls undergoing the puberty ritual at Shika V illage, and w ill themselves be secluded there. T h ey pay frequent visits to Shika unaccompanied b y either o f their parents. M angaleshi herself goes to Shika to take part in any m ajor ritual that is perform ed m ere. H er classificatory brother Gideon and Kasonda have built adjacent huts in M ukanza V illage, and Kasonda’s con nections w ith Shika have stood him in good stead both in his internal intrigues and in his intercalary role betw een the tw o villages. It is im portant for the M ukanza people to maintain friendly relations w ith Shika V illage, for it is possible that a m em ber o f that village w ill some day succeed to the Kanongesha chieftainship.2 T he husband o f Kasonda’s classificatory sister C haw utong’i (I,G i2 ), N deleki (II,D8), is also the son o f a Shika man, and he and his w ife have also built adjacently to Kasonda on the other side to Gideon. These persons, linked to one another through kinship ties o f varying character w ith Shika V illage, form a sm all pressure group under the leadership o f Kasonda w ithin M ukanza V illage. W ith in M ukanza too they support the interests o f nyachintang ’a lineage to w h ich they are affinally or filially attached: A t the same tim e they provide a num ber o f linkages w ith Shika V illage w hich b y their m ulti plicity and strength prevented the conflict betw een M ukanza Kabinda and the elders o f Shika V illage from developing into a permanent breach o f friendly relations between the tw o villages. T w o members o f malabu lineage w ere also affinally linked w ith Shika V illage in 1951, and conflict arose from both relationships. N yam uw ang’a ( I ,G i7 ; see Social Dram a IV , p. 148) was m arried to W illiam M atorokoshi, (II,D7), son o f the late Shika Ikubi. In that year he divorced her on the grounds that she was a w itch and wanted to kill him b y her w itchcraft. Since N yam uw ang’a was feared for the same reason in M ukanza 1 Kasonda stresses this double tie to support a politically valuable marriage. But Mangaleshi is not his true matrilateral cross-cousin. She is classificatory sister to Gideon (II,El), who is Kasonda’s matrilateral cross-cousin. 2 See pp. 207-8, 322.
P olitical A spects o f K insh ip and A ffin ity
267
V illage itsel£ and since die pair had no children, little was said about the m atter. B ut Sakazao (I,H9), senior elder o f malabu lineage, resented having to return W illiam ’s nseunt, since he was a poor man. H e was said to have upbraided N yam u w aag a for her reputed occult activities and to have demanded that she contribute to the repaym ent o f nsewu. I did n o t hear whether she did so, but it is probable that she did, since she w as living in Sakazao’s quarter o f the village in the dry season o f 1951 when Social Dram a IV took place. N yam u w an gV s sister Kalusa (I,G i9) was m arried to Kasonda’s father’s uterine brother N tololu (I,D2), and this pair w ere the parents o f Ikubi w h o , as described in the account o f Social Dram a IV, was supposed to have been killed b y the w itchcraft o f N yam uw ang a. N to lolu at that tim e was livin g iixorilocally in M ukanza V illage, having quarrelled w ith his classificatory brother, the headman o f Shika, Shamupaku (II,D5.) A fter Ikubi’s death N to lolu returned to Shika V illage w ith Kalusa, bitterly angry w ith N yam uw ang’a and in fact at odds w ith the people o f M ukanza V illage as a w hole. The death o f Kalusa, sister o f N yam uw ang’a, a t Shika V illage hi 1953 touched o ff a social drama w hich made o vert m any o f the hidden tensions both w ithin M ukanza V illage and between M ukanza and Shika Villages. SO CIAL D R AM A W Two Headmen Dispute over a Death Payment (my own observations) O n July 6th, 1953, a messenger arrived one somnolent afternoon at Mukanza Village with news from Shika that Kalusa (LG19) was dead. Instantly w ailing broke out from all the Mukanza women who rushed into the central clearing between the huts and sat or stood each separate from one another. In a short while the plaints became organized into songs o f grief, led by Manyosa (I,Gi3), who sat with her hands crossed on her shoulders, shaking and nodding her head while tears ran down her free. Am ong her rhythmic cries could be heard brief phrases, 4 She died all alone \ * m y father and mother and m y unde Kanyombu are dead*, ‘ a death by andumba [a term for witchcraft frmiliars or tuychela]’ . A t first all the women had been freing in the direction off Shika Village. When the last phrase was uttered all turned in the direction o f Sandombu Farm with anger
P olitical A spects o f K insh ip and A ffin ity
269
on their faces. Men had joined the bewailing group, although in keeping with custom they did not weep. Sandombu (I,Gio) came slowly towards them from his ow n place. He came forward into the middle o f the dispersed group and a harsh argument developed between him and Nyakalusa, the slave wife o f Sakazao (LH9), privileged as we have noted already by virtue o f her role as an *out sider * to speak openly what was in the minds o f all. Chayangoma (i,H i5), classificatory sister’s son o f Kalusa and Nyamuwang’a (f,G i7), and classificatory brother o f Sakazao, all in malabu lineage, also joined issue with Sandombu. Chayangoma since he was, as we have described, a blind man, as well as a close lineal kinsman o f Kalusa, was also free to speak his views. They attacked Sandombu for harbouring a witch in his farm, meaning Nyamuwang’a. Nyamu wang’a had recently visited her sister at Shika Village and they recalled how Nyamuwang’a’s sister Shimili (I,Gi8) had died at Chimbili Village after Nyamuwang’a had visited her, and how Xkubi (I,Hi4), Kalusa’s daughter, whom in the words o f Mangaleshi (II,E2), Kasonda’s wife, * Kalusa had followed ’, had died after receiving a malediction from the old witch. Sandombu shouted back at them that there had been no proof that Nyamuwang’a was a witch. She had loved her sister, he said. He returned to his house in animated conversation with Chayangoma (I,H i5) who was led along beside him by Kasonda’s small son who was incidentally his father’s confidant. Later the men assembled in the chefta. It was decided that they should send round to the other villages to buy beer for the funeral gathering (Chipenji) o f Kalusa (LG19) which would be held in Mukanza Village. It was also decided to send a small delegation to Shika to ascertain more fully the facts o f the situation and to make arrangements about the funeral gathering. Feelings had calmed somewhat, following a rumour that Headman Shika (HJDs) was contemplating a witchcraft action in a chief’s private court against unspecified ‘ Mukanza. people It was suggested that in this w ay he wished to evade the payment o f mpepi to Mukanza Village, since i f he could get his charge upheld the latter would be culpable and his own group could not in any way be held responsible. Mpepi is paid by the kin o f the relict, not in com pensation for their witchcraft, but inter alia for failing to protect the deceased spouse from real or supernatural harm. But i f her own kin had bewitched her, Kalusa’s arnnes need not pay them mpepi. This rumour, which may have been inspired by someone in Mukanza, had the effect o f uniting the village against Shika. They discussed the question o f mpepi, which is divided into two consecutive payments, a small sum called kutotafu (* to take away the dead ’) for ritual services rendered to the w idow or widower, and the mpepi proper, and decided to ask only 2s. for the former since they could not obtain
C7
D :o o c:.
o c.■?
o o (■ ; r-
o o
e o o •o ■ c (• •( y C
tJ ;© 1,
270
Schism and C ontinuity in an A frican Society
enough beer or provide enough food at that time for a long funeral gathering (in the past this sometimes lasted three or more months accompanied by elaborate ritual). The mpepi proper was to be left over for negotiation between the tw o villages after the all-night Mudyileji dance which concludes the funerary ritual. The emissaries to Shika were as follow s: Mukanza’s sons, Gideon (H,Ei) whose mother came from Shika, and Zachariah (XJ3), uterine nephew o f Sakazao (I,Hp) ; Short (I,Hi6), younger uterine brother ofChayangoma (I,Hi5) ; Zuliyana (IJz), Mukanza’s daughter and Zachariah’s sister ; and Nyakalusa. The women were included only because the deceased was a woman. Village unity had been restored to such an extent that Nyamuwang’a (I,Gi7) was nominated as executrix (1nyamuju, * mother o f the dead ’). She would receive and distribute Kalusa s property, sell her Shika gardens, and take charge o f the mpepi paid by Headman Shika. This nomination was a demonstration o f the solidarity o f the Mukanza group in the face o f the outside world. Normally this office would devolve upon the uterine brother or mother’s brother o f a dead woman but Kalusa had no brothers and Nyamuwang’a was her closest uterine relative. Sakazao, as head o f malabu lineage, might have been appointed but since he had recendy been fined for not returning the marriage payment to the divorced husband o f Kalusa*s daughter Evinaki (I,Hi3) and was looking round desperately for cash from whatever source, it was thought inadvisable by the majority o f village elders to entrust him with the office o f executor. B y the mechanism, therefore, o f the chota debate a crisis which threatened to activate the hostilities latent in the village was brought under control and the ranks closed against Shika Village. The next stage o f this social drama followed the month o f the Chipenji gathering, during which Ntololu (II,Dz) kept the customs o f widowhood at Mukanza Village where a grass hut (nkunka) was built for him. O n the night o f Mudyileji (the final dance), Headman Shika and other elders o f his village, together with many women, came to Mukanza. Nearly every village in the vicinage o f Mukanza Village sent a contingent. The general atmosphere was euphoric and no hint was afforded o f the acrimonious debate that was to take place on the following day. Both Mukanza and Shika danced, and much apparent good-humour characterized their relationship round the beer pots. Next day the real business began, the controversy over mpepi. Three mahaku were called in to judge between the contending parties, Headman N g’ombi (see p. 200), Headman Mulila who had come some years ago from Angola into the vicinage, and Headman Nswanamatempa. None o f these men had close consanguineal ties
P olitical A spects o f K inship and A ffinity
271
witli cither Shika or Mukanza, although N g’ombi had married Mukanza’s daughter Koniya (I,Ji). But N g’ombi*s reputation for scrupulous fairness made him immune from the charge o f favouritism. Shika called in Headman Nswanakudya (IIJD10), whom, it w ill be remembered, he called * older brother * (p. 208 et seq.), to assist him in putting forward his point o f view. He was also supported by the widower Ntololu (D^Dz) and by a classifrcaeory sister s son from Shika Village. O n the Mukanza side, Mukanza Kabinda himself spoke, assisted by Sandombu, Chayangoma, and Sakazao. Nyamuwang’a refused to discuss the question o f mpepi, saying that she had been ‘ despised * (ku-sawula) by the Shika people on many occasions and would not hear* herself insulted again. Her place was taken, curiously enough, by Nyakalusa, who had originally accused her o f witchcraft. Once more Mukanza people had closed their ranks. The discussion took place outside the chota since it was not about internal village affairs. Sandombu opened the debate by demanding a sum o f ^ 5 as mpepi. Ntololu, he alleged, had made it clear during the Chipenji funeral gathering that he intended to retain with him in Shika Village his tw o daughters by Kalusa, so that their services would be lost to Mukanza. Furthermore, * bad words * had been spoken by Shika people about witchcraft in Mukanza and it was only just that they should make amends by paying a big mpepi. He went on to say that X5 was not such a large sum after all b y modem standards and mentioned cases in which £ 7 and even ¿£io had been paid. Shika, a big, Blustering man, who, h is said, had once tried to bribe Kanongesha to appoint him to the Chibwikaship (see Chapter Eleven), hotly disputed the sum named, pointing out that the two villages had for many years been in the habit o f exchanging wives, and had come to he interrelated by many ties. It was an unfriendly act to make heavy demands on one’s kin. Besides, he recalled that when Nyattmgeji (H,D3) was on the point o f death Mukanza had divorced her, only 2$. 6d. to Shika. Again, at the beginning o f Chipenji sent 1 0 5 . to Mukanza by Ntololu to contribute to the cost o f divination into the death. ‘ W hat did the chicken say ? ’ he went on, referring to the poison test (ng'ombu yamwaji) administered to fowls in one mode o f divination. It would appear that Mukanza Kabinda had not gone to a diviner but had * eaten * the money. D id Mukanza think that he was a muloji (sorcerer) to ask him for such a large mpepi ? Mukanza him self then spoke, saying that he had no quarrel w ith Shika but that the latter must know that it was the custom nowadays to pay a large mpepi since * every kind o f food and beer now cost money, and cloths, pots and pans were very dear ’ . He had not gone to a diviner, for, i f such a matter came to the car o f the Boma, all o f them might * receive a very heavy case *. But
m
272
Schism and Continuity in an A frican Society
he had taken the money already paid into account when he had decided in consultation with his fellow-villagers on the amount of mpepi he would ask for. Other elders now spoke on both sides, presenting variants on the main arguments. After a time the judges were consulted and they suggested that Shika should pay £2. 10s. Headman N g’ombi said that bitter things had been said on both sides and it had to be taken into account that Mukanza had divorced Nyatungeji in order to avoid the paymcnj o f mpepi. But since many people in both villages were kin, it would be a bad thing if one group started to nurse a grievance against the other. He thought that Mukanza should be satisfied with half the amount demanded, in view o f the Nyatungeji affair, o f kinship relations between the villages, and o f what Shika had already paid as a contribution towards a divination which had never taken place. Sandombu, speaking on behalf o f Mukanza Village, refused to accept this, saying that while they would no longer press for f s they would be content with nothing less than .£3 10s. After more argument Shika agreed to pay this sum. But he told the Mukanza people that they were mean and selfish to drive such a hard bargain. He said that he could only pay j£ i that day, as he was a poor man (kazweng’i) but that he would pay the rest later in instal ments. He then left the village * in a huff* (ku-fundamoka) with his followers. Kalusa’s children remained at Shika Village, and for more than six months there was a marked reduction in visiting between the villages, and much mutual backbiting. In the last chapter I shall describe how one o f the functions o f the Chikamba ritual I witnessed was to effect a reconciliation between the villages. Analysis Social Drama VII illustrates a number o f points already made. Am ong these m ay be noted : (1) the w ay in w hich inter-village disputes help to prom ote intra-village unity— strikingly brought out by the change o f front shown by Chayangom a and N yakalusa, w ho began by accusing Sandombu and N yam uw ang’a o f witchcraft and abetting w itchcraft and ended by jo in tly arguing the case for M ukanza V illage, and even, in Nyakalusa’s case, by acting as representative for the one initially stigm atized ; (2) the interest shown by representatives o f the w ider social system o f the vicinage in w hat was prim arily a m atter concerning tw o inter m arrying villages ; and (3) the role o f m atriliny in providing links between villages in different vicinages, exem plified by
P olitical A spects o f K insh ip and A ffinity
o
273
N sw anakudya’s advocacy o f his classificatory brother’s case. But its m ain interest for us in this chapter is the ligh t it throws on the kinds o f cleavages and ties between interm arrying villages. E very breach in a m arriage, whether made by divorce or death, is accompanied by conflict between the villages, but the existence o f other ties o f affinity and kinship prevents the conflict from rupturing all ties between them. Cross-cutting affiliations ally those w ho are in conflict. Kasonda and M anyosa, for exam ple, w ould not argue on behalf o f M ukanza V illage. T h ey said that their father Kahum pu came from Shika V illage and that he was the full-brother o f N tololu. Kasonda said that in addition his ow n children belonged to Shika V illage through their m other. H ow then could he speak against men w hom he called 4father ’, and w ho w ere grandfathers o f his ow n children, i.e. members o f the same linked-generation segm ent in Shika V illage ? Again, M ukanza him self spoke w ith great m oderation, fo r his three children b y N yatungeji belonged b y m atrilineal descent to Shika’s follow in g. B y another reckoning, through M ukanza’s marriage w ith M angaleshi, he was a * co-parent-in-law ’ (nktilu nami) w ith Headman Shika, since he was muhu to M angaleshi, and Shika was, in the context o f affinity, muku to Kasonda. Persons w h o call one another nkulu nami, m ay sleep in one another’s huts, receive hospitality from one another, and in general behave w ith considerable fam iliarity, although there is no jo k in g between them. T o m y m ind this is another ex pression o f the solidarity o f genealogical generations. Sandombu, on the other hand, had no close kinship or affinal tie w ith Shika Village to prevent his speaking the M ukanza case unreservedly. Shika recognized in him the authentic voice o f M ukanza Village, when he asked fo r ^ 3 10s., and agreed to pay up to preserve peace between the villages because he knew Sandom bu’s demand to be a final offer. B ut although the villages did not break o ff relations it was felt by the Shika people that M ukanza V illage had had the better o f the bargain. Resentm ent lingered ; Shika people complained in m any villages o f the meanness o f the M ukanza group, coupling their reprobation o f the village w ith disparagement o f the K aw iku as a w hole. T he latent conflict between Lunda and K aw iku, the invaders and the indigenous people, never entirely damped dow n, began to sm oulder. M oreover, the people o f
©
o
0 ©
©
o © @
© ©
©
© © G ©
© ©
o o
c ©
n n
o o
e f) o o c
O o Q O c o o
C' .0 €
1;
9 O
274
Schism and Continuity in an A frican Society
M ukanza felt rather uneasy about their ow n conduct in the affair. M embers o f other villages hinted that they had not behaved w ell in demanding such a large sum as ntpepi. W hen it was decided to hold the Chihamba ritual for N yam ukola 1 a special invitation was sent to Shika V illage to attend. N tololu’s children came and a number o f children linked to both villages w ere chosen as candidates for initiation. A t the ritual level both villages w ere tem porarily reconciled. B ut outside this relationship o f m arital exchange w ith Shika V illage the people o f M ukanza had a w ide and extensive range o f marital connections w ith other villages. Even divorce and death do not break such ties, i f the union has been productive o f children, for children act as a focus o f com m on interest between the maternal and paternal groups. Children w ho live w ith their m other often exchange visits w ith their father. In custom, a man must show no animosity towards his ex-w ife’s present husband w hom he calls nkulu nami, the same term as he applies to his child’s spouse’s parents or uncles and aunts. This rather tense relationship o f outward friendliness prescribed b y custom enables a man to visit his ow n children at his divorced w ife’s hom e, w hether she be livin g in her ow n or in her new husband’s village. The customs are directed against the outbreak o f physical violence due to jealousy. B ut i f a couple have had no children, the relations between their respective villages w ill nearly always cease on their divorce. Affinal ties are strengthened enormously w ith the advent o f children, in whose w elfare the groups to which each spouse belongs by various principles o f attachment have an interest. I exam ine the extant inter-village marriages o f M ukanza people and the fruitful marriages o f present members w hich ended in divorce and death. Starting w ith members o f n yach intang ’ a lineage, in order o f seniority, M anyosa (I,G i3) was married to Chikasa, o f M pawu V illage in the deposed Governm ent SubC h ie f N tam bu’s area. She had previously been married to Chisem pi, w hom she divorced and b y w hom she had five children ; he was a member o f a village in the vicinage in which Nyaluhana (see p. 48) is principal headman, in C h ie f Kanongesha’s
€
O C) k)
1 See p. 310.
^ÎÆKJgSm
276
Schism and Continuity in an A frican Society
area. Her oldest son lives w ith his father, although he sometimes visits Kafum bu V illage, and she goes to see him w hen she visits Kafum bu V illage w hich is near Sawiyem bi V illage, w here her form er husband and her son reside. A ll her other children, except R.osina (I,H8) w h o lives w ith her husband in Kafum bu, reside in M ukanza V illage. Chaw utong’i (I,G i2 ), whose husband N deleki (II,D8) lives uxorilocally, provides a link with N g ’om bi V illage through N deleki’s matrilineal connection there and w ith Shika through N deleki’s hither. Kasonda (I,G i5), by his divorced w ife M ika, w h o had a baby b y a lover, main tained connections w ith N tam bu’s area. These connections, though snapped b y divorce, between Kasonda and M ika, were renewed b y Kasonda’s m arriage w ith ex-C h ief N tam bu’s sister’s daughter’s daughter Luciana. nyachintang ’ a lineage had maintained relations b y affinity w ith the ntambu chiefly lineage for several generations. T he Ntam bu chieftainship was reckoned equivalent in status to that o f Kanongesha in Lunda tradition : C h ie f N tam bu sent tribute directly to M w antiyanvw a as Kanon gesha did in the past, and the tw o chiefly lineages interchanged spouses ju st as Shika and M ukanza V illage lineages did. The father o f M ukanza Kabinda and N yam w aha, Santapu, was the younger uterine brother o f an incum bent o f the N tam bu chief tainship. T he marriages o f N yam w aha’s son and daughter maintained the connection. Kasonda had in fact m arried quite astutely. His children b y Mangaleshi m ight succeed or bear successors to the chieftainship o f Kanongesha. His children by Luciana m ight succeed to the stool o f N tam bu. Kasonda cherished hopes that Governm ent w ould eventually restore that chieftainship to its form er position in the N ative Authority. His son by Luciana m ight then perhaps becom e a Governm ent C h ie f in that case. I f Jerry, his son b y M angaleshi, became Kanongesha, Kasonda m ight be the father o f tw o chiefs. Sandombu kept up Intermittent contact w ith the daughter or his first w ife N yalundaw u. This daughter, regarded b y Sandombu alone as his, had m arried into a village in Ikelenge Area and had a small daughter. W hen Sandombu visited Ikelenge Area on business he always stayed w ith his * daughter *. Sandombu’s sister M angalita (I,G i i) had form erly been married to a man from nearby Nswanakudya V illage, but in 1951 she was living at K afw eku V illage w ith her new husband. K afw eku, form erly a
P olitical A spects o f K in sh ip and A ffin ity
277
senior headman’s village, had previously been linked w ith M ukanza b y the uxorilocal m arriage o f Kajata (I,Fro), o f malabu lineage. It is in M w ininyilam ba Area, far from M ukanza, on the border o f A ngola. Bibiana (I,H4), m other’s sister’s daughter’s daughter o f Sandombu, also lived w ith her husband, a crippled ex-askari, in Chim bila V illage near Ikelenge’s capital village. She had three sons and tw o daughters. T he sons regularly visited M ukanza in the school holidays, and w hen I left, Bibiana, divorced b y her husband in 1954, was talking o f settling in Sandom bu Farm. Bibiana’s first parallel cousin, N yam pesi, was m arried to a store capitao w orkin g for a European trader, near the Bom a. She made frequent visits to Sandom bu Farm. H er husband had a useful role from the point o f view o f M ukanza V illage in placing the younger m en in paid jobsM anyosa’s oldest daughter Kandeieya was m arried to M akanjila, son o f Headm an M bim bi, w hose decrepit village im m ediately adjoined M ukanza. M anyosa’s son Daudson (fH y ) had tw o w ives. O ne was the daughter o f an O vim bundu im m igrant w ho had m ade a farm near M ukanza. T he other war. Sakutoha’s daughter (see p. 191 et seq.) from her father’s adjacent settlement, w hich paid tax through M ukanza. H er other daughter Rosina (I,H8) was m arried to Kantam oya (I,H 3), the younger brother o f the headm an o f K afom bu V illage. In malabu lineage Sakazao (I,Hp) was connected to Kafum bu by m arriage through N yaluw em a (I»C4). H is slave w ife N yakalusa no longer seemed to m aintain contact w ith her Angolan hom e village. N yatio li (I,H n ), his sister, had lived in M w ininyilam ba area fo r m any years w ith a husband from Nyanseng’i V illage. H er daughter Yana (IJ4) had married a man from the neighbouring village to N yanseng’i, Kam bim hi V illage. In 1951, w hen Y ana was still livin g virilocally, I visited K am bim hi V illage, far from the m otor road along a path w inding over troublesom e spurs o f rock, w ith a number o f M ukanza people ; and w e w ere all given hospitality by Yana and her husband, in the form o f fo o d and beer. Yana’s brother Pearson (I,j5) was m arried to a g irl from N g ’om bi V illage in the vicinage. T h eir younger sister Serita had previously been m arried to a man from Ibulu V illage in C h ib w ik a A rea, b y w hom she had had a daughter w ho was livin g w ith her father. R ecently
278
Schism and Continuity in an A frican Society
she married a man from N swanakudya V illage in the vicinage o f M ukanza. N yam uw ang’a (I,G i7) o f M ukanza had form erly been married to a man from Kafum bu V illage, by w hom she had borne a daughter. This man subsequently died, and his death was popularly supposed to have been caused b y N yam uw an gV s w itchcraft. B ut her daughter Ikayi (I,H i2) still visited her father’s village from tim e to tim e. N yam uw ang’a’s sub sequent marriage to a man from Shika V illage has already been mentioned. N yam alita (I,G zo), the surviving uterine sister o f N yam uwang’a, lived viriiocally for m any years w ith her husband, a farm headman in Ikelenge Area, and later obtained w o rk as a W ard M aid at the Kalene M ission Hospital. H er younger children, like those o f Bibiana (see p. 164), frequently visited Mukanza V illage in school holidays, and it was expected by M ukanza people that they w ould look on it as their hom e when they grew up. Chmema (I,G 2l) and his uterine siblings remained in Ikelenge Area after Kahali Chandenda (I,F5 ; see p. 180 et seq.) had returned to M ukanza V illage. Together w ith some disgruntled claimants for the M ukang’ala chieftainship he established a new composite village, called after him. H e and his siblings married in Ikelenge Area and I shall not enumerate the marital and consanguineal ties that n ow interlink the matrilineage o f M ukanza through Chinem a and his siblings w ith villages in Ikelenge Area. Some o f Chinem a’s uterine nephews have n ow returned to M ukanza, including Chayangom a (I,H i5) and Short (I,H i6 ; see Social Dram a V II, p. 269 et seq.), neither o f w hom w ere married at the tim e o f w hich I am w riting. Line’s senior w ife came from Sawiyem bi V illage, near Kafum bu V illage, and he frequently accompanied her on visits to her village. His jun ior w ife, M aria, as noted on page 186, belonged to the lineage o f n y a c h u l a , led o ff from M ukanza V illage b y Yim bw endi, as described in Social Dram a V I (p. 181). O f the nyachujla people w ho remained in M ukanza V illage, N yaw unyum bi (I,G2s) through previous marriages w hich w ere fertile provided links w ith villages in C hibw ika and Kanongesha areas. H er son Kenson (I,H 2i), w ho resided at M ukanza Village, was married to a wom an from nearby Chibw akata V illage. She herself, as already mentioned, was m arried to Headman M bim bi, and her youngest daughter (I,H2o) was married to a man from Kasai
P olitical A spects o f K in sh ip and A ffinity
279
Village, a K aw iku village in another vicinage (see p . 191 et seq.). Aram (IJ 7), Line’s sister’s son, was m arried to a w om an from Chikang’a V illage, in the same neighbourhood as Kasai Village. O f N yam ukola’s children, her daughter K oniya (I,Jr} was married to Headman N g ’om bi, w hile her son Zachariah (IJ3) was m arried to the sister’s daughter o f Farm Headman Sakutoha (see pp. 192-3), close to M ukanza V illage. Affinal ties, then, linked M ukanza V illage to other villages in the same vicinage, to villages in adjoining vicinages, and to villages in distant vicinages and even in different Governm ent Chiefdom s. B y m arriage the village was interconnected w ith villages in C h ibw ika Area fo rty o r fifty miles to the south, and with M w ininyilam ba Area about the same distance to the north west. Genealogies suggest that similar w ide ranges o f marital ties interlinked villages in pre-European times. Num erous close ties o f kinship and affinity w ith Shika V illage can be placed at one end o f a scale, on the other end o f w hich w e find single threads o f connection radiating in all directions to settlements near and far. Table X X I, based on village census m aterial, gives some idea o f the spatial range and dispersion o f m arital ties. I was un fortunately unable to w ork out in m any cases w hat marriages had taken place between members o f villages belonging to die same vicinage. M ukang’ala is classed as a * chiefilom *, although the incum bent was deposed b y the Governm ent in 1947. The high proportion o f extra-chiefHom marriages in the Ikelenge sample is probably due in large measure to the high im m igration into the area from A ngola, the B elgian C on go and from other parts o f M w inilunga D istrict (see p. 9 and p. 59). The table, despite its deficiencies, does show how w id ely the netw ork o f affinity, w hich interlinks villages, vicinages and chiefdom s, is spread. Kinship and Affinal Connections between Headmen in a Vicinage Before discussing the netw ork o f affinal and kinship ties interlinking the headmen in the vicinage o f w hich M ukanza Village form s part, a few w ords should be said about polygyn y and about the polygynous marriages o f headmen in particular. Table X X II gives the age and conjugal status o f all members o f
Marriage
Ike Ikdenge
Muk Mukang’ala Nya Nyakaseya Mwi Mwininyilamba
op
TABLE XXI Spatial R angb
P olitical A spects o f K in sh ip and A ffinity
281
24 villages and 6 farms, w hile T able X X III compares the age and conjugal status o f 28 headmen w ith those o f 107 commoners o f thirty years old and over.1 M ore than h a lf the headmen had tw o or m ore w ives w h ile on ly about a sixth o f the com m oners had tw o or m ore w ives. Headmen have greater obligations than commoners to offer hospitality, and the m ore w ives they have the better are they able to do this. B u t m arital ties betw een headmen and members o f other villages are politically stronger links than those between com m oners, since the headman represents a group and sometimes inherits a tim e-honoured title. M en often, like Kasonda, anticipate headmanship b y judicious polygynous marriage w ith im portant maternal descent groups ; others find that they m ust m arry m ore than one w ife to fulfil their duties after they have succeeded. Since also the children o f head men often remain w ith them , the polygynous headman w ith fertile w ives is enabled to beget a follow in g w ithin the social system o f the village o f w hose lo ya lty he can be almost sure. C onflict and tension between co-w ives is, how ever, a frequent cause or excuse fo r divorce so that p olygyn y, w hile it offers advantages to the am bitious, m ore than doubles the possibilities o f conflict, never absent from the N dem bu m arital relationship. Before 1952 o n ly five villages com prised the vicinage (chi tting*Hi or itung a) to w hich M ukanza V illage belonged. These were N s wanakudya, Chibw akata (in a different site, n ow in a Governm ent Forest R eserve), N g ’om bi, M ukanza, and M bim bi. N o w N sw anakudya and Chibw akata w ere interlinked by a number o f m arital and consanguineal ties. T he headman o f N swanakudya was the son o f a Chibw akata man, K aluw a, w ho was the m other's brother o f the headman o f Chibw akata. Thus the tw o headm en w ere cross-cousins. Headman Nswanakudya married N yachipoya, classificatory sister o f Headman C h ib wakata. W h en N g ’om bi split o ff from Chibw akata, N yachipoya’s uterine siblings w ent w ith N g ’om bi. Headman N swanakudya’s oldest son b y N yachipoya later settled in N g ’om bi V illage w here he was mulopu o r deputy o f the headman. Head man N g ’om bi’s m other was older uterine sister o f N yachipoya. Several other marriages interconnected the three villages but 1 The lay-out o f these tables is based on Table X IX in Mitchell and Barnes, T he Lam ba Village , p. 4 6 .
■;. ,'*r* *\>
N,
'.v^X
t
TABLE XXII A ge
Sample of 24 villages and 6 farms:
and
C onjugal Status
Number in sample Age
Married
Unw,
Males : o-Pu PU-19 20-29 30-39 40-49 50-59 60-
Percentage in each age group
95
18 15 3
— —
Total .
U!
Females : o-Pu PU-19 20-29 30-39 40-49 50-59 60-
20 — —
Total .
IIP
99
—
—
D
— —■ —2 3 —
— — — — — —
Total Unw.
1
2
3
4
__
— —
— — — — I 2 —
— — — — __ —
— 31 29
5 5
21 14
6 14
5
24
5
5
5
IIP
35
— — 8
— — — — i
— — — —
3 4 5
Married
R
6
3 9
26
13
2 il 49 52 27 ï9
10
190
D
R
Total
Total
—
— 3<> 34
95
18 51
ioo-o 100*0 29-4
— __ —
—
—
35
— — —
— —
77
5*1 9*7
—
158
299
43*8
17
i *7
39*8
— ~~ — — —
2 31
ÏOI Si
98-0
49 52
57 55 32
— — 14*0
— — — —
2-0 60*8 86*o
3*1
84-4
—
—— — — — — ~
—
—
_
—
—
—
—
39*2
27 19 10
27 25
— — — •— —
190
348
34*2
5*5
14*3
n
-7
1-0
o*3
52*8 100*0
— „ — —
— — — —
— — —
__
— — — — —
—
—
—
—
2-0 100-0 60*8 100*0 86-0 100-0 94*5 100-0 84*4 100*0 70-4 100*0 40-0 100*0
9*8
12*8 19*4 4 (5*7
677
94*5
12-5 iS -5 24*0
n *l 36*0
70*4
7*5
3 *7 ,
54 *6
40-0
Past i . A ge
Age
and
I li . !■ ,.fi ■ PIS p£f!
C onjugal Status
op
3
4
I
—
I
— —
—
—
—
X 8
II
4
12
H
Sample of 107commoners: 3
3
2* A ge
2
29
3
20
5
4 5
5
14 U
6 i
5
76
16
i i
2
54*6
100*0
ip P îlS ;
ÜS"*
Headmen
9
1 2
i
16
i
28
28
—
~ —
—
C onjugal Status 33
—
R
Total
i
9 ï6
and
D
Total
X 2
i
i
i
Married Total Unw.
2
“
Percentage in each age group
R I
6o~
3-2
■ !
Married
D
— —
iíM^h p § i i l l
Number in sample
Unw.
_
<5*6 —
6o*8 4 (5-7 68*5
Total
4
— — 100*0 — — 100*0 — 70*6 100*0 — 87*2 IOO’O __ 90-3 100*0 — ioo-o 100*0 2-9 85*7 100*0
74*4
I
Sample of 28 headmen
Total .
— —
3
30
PPPp V ;
30-39 40-49 50-59 60-
~
I
31
Vj X ? P
Total .
3
28 30 30
tip
30-39 . 40-49 50-59
2
—
— — — — 14'3
39
I
38
26 21 14
29 21 IP
94
IO7 Legend
7*9
2*8
— of
2
100*0 50*o 88*8 68*8 25*0
3
4
100*0 100*Q
50-0
—
42*8
50*0
Total
6*2
100*0 100*0 100*0 100*0 100*0 100*0
3*6
3*6
100*0 100*0
3*5 4*8
—
100*0 100*0 100*0 73*7 100*0
1*9
—
87-8 100*0
ii-i
C ommoners (male) 5*3
76-3
10*3
4-7
26*3
17*2 28*6
4-7
71-0
14*9
Unw. Unwedded D Divorced, and not remarried R Relict; widower or widow, not remarried ! j 2, 3, 4 1, 2, 3, 4 Spouses Pu Puberty
86*8 100*0
10*5
69-0 66-7 68*4
5*3
89*7
284
Schism and Continuity in an African Society
those mentioned w ere politically the m ost outstanding. N g ’ombi Village, it w ill be remembered, split o ff from Chibw akata Village in the 1920’s. M ukanza V illage was linked w ith M bim bi V illage in a number o f ways. T he headman o f M bim bi was the son o f a m em ber o f malabu lineage in M ukanza V illage, and his son w as married to Kandeleya, daughter o f M any osa o f nyach intang ’ a lineage, the favourite * granddaughter ’ o f M ukanza Kabinda. M bim bi’s second w ife was N yaw unyum bi, m other o f M ukanza Kabinda’s second w ife and Line’s second w ife. The prim ary relationship between M ukanza Kabinda and M bim bi was that o f classificatory * father ’ to * son ’, since M bim bi’s father Ibeleka was classificatory ‘ older brother ’ to M ukanza. Through subsequent marital ties M ukanza could be considered the classificatory son-in-law o f M bim bi, since M bim bi had married M ukanza’s m other-in-law N yaw unyum bi. B y matrilineal affiliation M bim bi belonged to the descent group from w hich incumbents o f the M ukang’ala senior headmanship w ere chosen. Betw een these tw o groups o f closely related and intermar ried villages, N sw anakudya-Chibw akata-N g’om bi on the one hand, and M ukanza-M bim bi on the other, stretched threads o f genealogical connection. Headman N g ’om bi was the husband o f M ukanza’s and N yam ukola’s daughter K oniya, form er wife o f Kasonda, M ukanza’s uterine nephew. Few m arital ties directly interlinked Nswanakudya and M ukanza, but the close relationship o f marital exchange w hich existed between Shika, from w hich N swanakudya had seceded, and M ukanza, b y classi ficatory extensions o f kinship terms and behaviour, made it possible for M ukanza Kabinda and Headman Nswanakudya to call one another ishaku, * brother-in-law *. Kasonda called Nswanakudya ‘ father ’ since N swanakudya was the classificatory * older brother ’ o f his father. Kahum pu. T hrough the marriage o f his sister N yam w aha to the previous headman o f Chibw akata, M ulow a, M ukanza could also call the present headman o f Chibw akata, Chonku, classificatory younger brother o f the late headman M ulow a, his ‘ brother-in-law ’ (ishaku). Kasonda’s father’s father came from Chibw akata. A few marriages between younger people interlinked the tw o villages. B u t on the whole few ties o f kinship and affinity connected M ukanza V illage with these three villages. M bim bi once m arried a w om an from
P olitical Aspects o j K insh ip and A ffin ity
285
Shika V illage w h o died m any years ago, but o n the strength o f this form er m arital tie, he could call Headman N swanakudya * brother-in-law * (ishaku). I could not trace any genealogical connection between M bim bi and Chibw akata. In 1952 Governm ent m oved a num ber o f villages out o f the vicinage in w h ich N sang’anyi was the principal village w hen their form er territory was declared a Forest R eserve. These villages w ere told to build in the vicinage o f M ukanza V illage. M ap 3 (p. 21) shows the distribution o f these villages and their gardens in 1953. It is instructive to note the w ay in w hich kin ship, affinal and historical connections influenced the choice o f their n ew sites. Nsang*anyi w ould not build near its offshoot Nswanam atem pa, in view o f the tension that still existed after the secession o f the latter (see p. 212) as recently as 1940. N or w ould N sang’anyi build dose to M ukanza, w h o claim ed to be mwenimbu, principal headman, o f that part o f the vicinage aborning the Governm ent m otor road. N sang’anyi was the senior and m ost ancient K aw iku village and could n o t recognize the local pre-em inence o f any o f its offshoots. Nswanamatempa broke up as described on page 216 into three settlements, all o f w hich built near M ukanza, a fellow -K aw iku but distantlyrelated village. A nother group o f villages and farms, settled on the ve ry verge o f the Forest Reserve, w ere interrelated w ith each other and w ith villages already in the vicinage. Itota, the head man o f one, was m arried to the sister o f the headman o f another, named Lambakasa. Lambakasa and his sister w ere in their turn siblings o f Headm an M bim bi. T heir father cam e from M ukanza Village, from malajbu lineage. Lambakasa was a fu ll brother o f M bim bi ; Ifcota’s w ife was their half-sister. H er full-brother Nsanji built a small farm opposite the road from Lambakasa, his half-brother b y the same father, and on the same side o f die road as Itota. In this group also was a form founded b y a wom an headman N yam um w em a w hose daughter was Lambakasa’s junior w ife. N ext to N sang’anyi V illage on the Itota side was Kanyahu Village. Headm an Kanyahu had m arried N sang’anyi’s uterine niece. K anyabu’s ow n uterine niece had once been m arried to Sakazao o f M ukanza Village, where; indeed, she had died. Her daughter b y a previous husband was the w ife o f Gideon, the son o f M ukanza Kabinda.
286
Schism and C ontinuity in an A frican Society
These villages then, w ere intricately interlaced by affinal and kinship ties. N o ties o f either type existed between Mukanza and Itota. In this case, how ever, the headmen made the contract o f blood-brotherhood (wulunda). M ukanza used this relation ship to secure for nothing the services o f Itota as senior ritual specialist in the Chihamba ritual, to be described in Chapter Ten. A fter 1952 also, a few small farms grew up in the vicinage. Anderson M ulum bi (see Social Dram a V , p. 160 et seq.), son o f a form er female slave o f Chibw akata, made a farm between M ukanza and the Nswanamatempa group, and set up in business as a sewing-machine tailor. His w ife was sister’s daughter’s daughter o f Nsang’anyi, and m any disputes arose between the K aw iku and Anderson over his cruel treatm ent o f her. Anderson’s sister, married to an O vim bundu ritual specialist w ho had intro duced the new ritual o f Tukuka into the vicinage, occupied the next farm. N ext to her again was another O vim bundu, a petty trader whose daughter was married to Daudson, M anyosa’s son, o f M ukanza V illage. Beside him was the small farm o f Muchona whose w ife belonged to the Nswanakudya village lineage. He was a close friend o f the O vim bundu w itch-doctor and they jo in tly officiated at several kinds o f ritual. N ear Chibwakata, the younger uterine brother o f the headman, M akayi, made a fam ily farm w ith his tw o w ives and children. In the absence o f strong overall political authority, such ties o f kinship and affinity becom e o f the utmost im portance in giving some degree o f cohesion to a fluid and volatile society. B ut since strong possibilities o f conflict exist in these very relationships themselves, kinship and affinity, unsupported by other kinds o f linkage, do not constitute reliable political bonds between villages. Even where, as in the case o f M ukanza and Shika, conflicts w hich arose between tw o villages over a broken mar riage could to some extent be contained b y the countervailing influence o f other ties o f kinship and affinity w hich interconnected their members, tensions w ere set up w hich seriously reduced the frequency o f interactions between them . Indeed, such tensions m ight, w ithin each village, give rise to incessant bickering between its ow n nuclear lineage and affines and kin belonging to the lineage o f the other. In other words, the instability in the marriage relationship, arising from the conflict between matrilineal descent and virilocal marriage, was at the same time a
P olitical A spects o f K in sh ip and A ffin ity
2S7
source o f instability in the political relationships between groups allied by marriage» Kinship and affinity in such a society divide as often as they bind. T h ey provide brittle foundations for a social p o lity. T im e and again one notices the tenuity o f the kinship and m arital ties that hold together the com ponent elements o f local groupings o f w hatever kind. Villages constantly shed uterine sibling groups w hich becom e potential nuclei o f new settlements. Sometimes they divide at a higher order o f lineage segmentation. Individuals m ove from village to village in pursuit o f their im m ediate interests. M ost w om en m arry ou t at least once. Betw een villages, strife is endem ic at the level o f consanguineal and affinal relations. T he best neighbours are not one’s closest kin- Since m any kin participate in marriage and death paym ents, and since these are nearly alw ays subjects o f dispute in w hatever context o f social relations they m ay be considered, conflicts between villages over them are continually taking place. Such conflicts seldom reach a solution satisfactory to both parties. The contradiction between principles o f social grouping within villages inhibits the grow th o f deep lineages, tends to rip the m atricentric fam ily prem aturely from the m atrilineage, sets adjacent generations at odds w ith one another, divides sisters from brothers, and enfeebles the bonds o f m arriage. Betw een villages, the same contradiction engenders disputes, makes o f the vicinage an unstable grouping, and disperses kin throughout the w hole N dem bu region. U nlike the societies described in African Political Systems as * segm entary V in w hich lineage organization form s the permanent fram ew ork o f social relations, the cohesion o f N dem bu society depends on the strength o f its mechanisms fo r resolving conflict w inch arises from the weakness and instability o f lineage organization and from the ambiguous character o f kinship ana affinal ties. 1 Fortes, M. and Evans-Pritchard, E. E., * Introduction * to African Political System (1940), p. 10.
CH APTER X T H E P O L IT IC A L L Y IN T E G R A T IV E F U N C T IO N O F R IT U A L T^fe^ibeen^sjaggjes.te4..b^.Pro.£ess0r.Gluckmat)uL.thatm,s0cieties w ifoout-govem m entaL-m stitations ,-.~sociaL. cohesion-—appears to dep
I
1 * Political Institutions *, in T he Institutions o f Prim itive Society, p. 67. also Fortes, T h e Dynam ics o f Clanship Am ong the Tallensi, pp. 241 et seq. 288
See
© @
The Politically Integrative Function o f R itu al
289
expense ,o f its significant local unit,.the village_Frequentfission, associated w ith w ide-ranging spatial_m obility o£ -theu-secedmg group, tends to create through tim e, numerous., .sets* o f .local kinship nuclei, each set consisting o f villages w hich have arisen from a "com m on source or from one another, by:.fission. T he villages in each set, though ,.widely., scattered , in., space,.,.may be interlinked b y b elief in com m on m atrilineal..descent,,, or b y know ledge o f actual kinship connections. T h e separate sets are interlinked b y affinal, vicinage and other ties. Thus .fission and m obility, w hile they break up villages, interlock the nation, w hich has n o —effective overall -political-authority..- But-, when w e examined m ore closely the political ro le -o f ues.of-km ship and m arriage enjoining co-operation w e found that theseties-too. w ere o f an unstable and friable nature. . It w ould seem then that links o f kinship and- o f political and econom ic co-operation per se are insufficiently strong to hold together village and nation. A nd yet N dem bu are conscious and proud o f their .unity as a discrete group, and as a sub-group o f the great Lunda nation, the unity o f w h ich is expressed in -the kingship o f M w antiyanvw a. Such a consciousness o f national unity .obviously depends on the jo in t recognition b y all N dem bu o f a. com m on system o f values and a com m on set o f norm s regulating behaviour. Their unity_is a m oral, rather than a political, unity.... By_.what .social mechanisms are these com m on norms and values inculcated and periodically re-anim ated ? In this book I shall say nothing about the educative process w hereby youn g children are indoctrinated in the basic values o f their group. N o r, except in passing, have I been able to deal w ith the jurid ical mechanisms. M ore im portant perhaps than either as a means o f keeping continually before the people their com m on norm s and values, in terms o f sym bol, m im e and precept, is the com plex system o f ritual.1 The.profusicm of_types_and .fre.quency-.of performance, o f ritual in N dem bu.society are, in. a w ay,, confessions, o f failure in the pow er o f seculat mechanisms to redress and absorb, conflicts that arise in.and between local and kinship gro.ups....X have.suggested 1 1 1 do not intend here to make a cultural analysis o f Ndembu ritual but simply to isolate from the ritual complex those sociological features which are relevant in this book.
0 i]
©
© © 0
©
© ©
©
..
© © © ©
©
290
Schism and Continuity in an A frican Society
that the contradiction between virilocal marriage..and...matrilineal descent is a crucial .determinant, o f structural instabiiity^at-all levels o f organization. I w o u ld postulate that where the m ode o.f postmarital residence, is inconsistent., w ith ..the.mQde_o£jo5ckonmg descent in unilineally organized .sQcieties, local groupings in such societies w ill be inherendy unstable, unless other ..factors intervene, such a-s lim itations on access to resources. I. have tried to show how the tenacity and tendency to autonom y o f the uterine sibling group are related to virilocal marriage, and h ow the tenacity o f tins group tends, in situations o f conflict, to jeopardize the cohesion and continuity o f villages. I have also tried to trace a relationship between virilocal marriage and the high divorce rate, w hich, in its turn, impairs political relations between inter married villages. So radically disruptive o f local cohesion is this contradiction that not infrequently the com bined operation o f all countervailing relations is powerless to prevent absolute schism in villages. M ost Ndem bu rituals tacitly recognize the. instability o f villages and .o f the relations betweerL villages,, but .positL-the .ultimate unity of-all-Ndem bu in.a.snigle.m oraLcom m unity. The.instability o f kinship- and political relations is. recognized. in the fact, that the dominant social elem ent in the-.composition, of-ritual assemblies is not .a kinship group but an association o f adepts..who belong to m any, kinship groups... Also,..the.. .dominant-symb-Qls in the cluster -.of-sym bolic objects, and activities, .associated-with each ritual-do not reflect or express major aspects o f the.social structure, but rather the values w hich all Ndem bu possess in.com m on, such as, for example, the fertility o f women,, crops .and. animals, huntsmanship, health, and the pow er o f the ancestors-to .bestow or w ithhold such benefits. The ultimate unity o f aU-Ndembu is expressed in the com position o f ritual assemblies. Thus, all Ndem bu, all Lunda even, have the right to attend any perform ance "of ritual o f whatsoever typ e.... Indeed,-i£ theyJaave. been initiated into its special techniques and esoteric, know ledge, its * mysteries * {mpangu), they have the right -.to. Jparticip.ate as officiants (ayimbuki},1* whether-they- are- related-to-the- -principal subjects o f the ritual or not. Itris also expressed.in the recurrence 1 Singular cbim bukt. This term has the dual significance o f * doctor ’ or ‘ leech * in so far as the ritual is regarded as therapeutic, and ‘ adept * (perhaps even 4priest * for a senior adept) in so far as it is a national cult.
The P olitically Integrative Function o f R itu a l
291
throughout the ritual system, o f sym bols represeating.theJbistorical origin jo f the N dem bu in M w antiyanvw a’s em pireandLofsym bois representing gérontocratie authority in general and o f m atriliny in general rather than specific political positio n so fiau th o rity or specific matrilineages. T ke.u n ity _o_f_aUNdembu is. n ot maintained, by. pohtical control fro m a strong centre o f authority. N o r is their polity^ o n eco n sisting o f hom ologous unilineal descent , groups, formed., b y a process o f fission into structurally eq u ivalen tan d juxtaposed segments. M oreover, there is no considerable overlapping o f territorial and kinship groupings such as one finds in segm entary societies. In such societies force tends to be distributed according to the like but com petitive interests o f hom ologous segments. In both the pyram idal and segm entary types o f society, ritual relatiom .tend .both to .express and toJbuttressthe_major political relations,.and ritual sym bolism tends, to express.parricular political offices .or to stand, fo r the unity o f specific corporate g roups o f kin. In bo.th states and .segm entary societies,..pohtical,. kinship, ..territoriaLand ritual relations_tend to some extentJo._hejcotermino.us.1 In these respects, both types, differ markedly..&QmJhs„&idfeinhin. A m ong Ndem bu, relations established b y ritual .jc.ut_.sharply across kinship and territorial ties, and. even across tribal .affiliation,.. smce_members^_of_other Lunda.graups,_such.as ...the. Kosa o f Musok^htanda and.the N dem bu oXlshinde (and .evsnXvfena w ho practise, analogous rituals), can participate in a n y cult into which they have been initiated. „B y establishing ties o f co-participation in cults w h ich operate independently o f kinship and locallinkages, the rituaLsystem compensatesjto.some extent for the lim ited range o f effective pohtical. controLand fo r the instability, o f kinship and affinal, ties to w hich, political. value.is_attached. It w ould «trem almost t a be the consistent id io m o f N dem bu social structure that certaim basic sets o f ties w h ich tend„to. be interwoven_so_M to reinforce one another in m any .small~scale societieSy.shauld here be separated from one another ; further,_thexohesiom o£the total society tin som e measure depends, not on the. juxtaposition and correlative operation o f the ties, but on «heir disjunction, and even o n . their opposition. Thus m aternal descent. .groups, have 1 Fortes and. Evans-Pritchard, * Introduction * to African Political Systems, pp. 16 et seq.
. j
j j j j
] j
I j
{ \ I j
I !
292 ji J.
j( |
| i i! i
|■ j' 1 |j | | | j
Schism and Continuity in an African Society
no local centres^juid Yicmages^are..made. up.of.yfflages^m o f w hich are unrelated b y .lineal descent.^. .Descent..and..pQSt-marital residence are not congruent... .. SpatiaL.an
The P olitically Integrative Function o f R itu al
293
A—dismctiorLisjiiade^y^Njdembu.het^eeatlie.spiat^mMfeisfei) which, afflicts,..and its. m ad^afeafflictian._T h e„sp iiit ..that afflicts is a Jm'own ..and. named Jieceascd_i^attx:e.j>£jdie_aiflicted person or p atien t (muyej.i)__XKe.p^ife»i_is..ai:jth.e.J5amc..tiHie-a^£^^|d3i:e for.initiation into, the cu rative.cu lt;... the., doctor- is. im adep.Lxti that cult—^^.-^hfcj«oJe..<^^irho«„re£ers,.to_ceitain.characteidsdcs_of this spirit, .whidht are.conrelate.d.wifo.Q^^ the sufferer’s m isfortune-.or illness. Thus a spirit w h ich afflicts a wom an in the m ode o f Nkula (or as N dem bu pu t it, w hich * comes out in N kula \ as in the phrase mukishi warni wedikilili muNkula, * m y spirit em erged in Nkula ’), 4 sits in her back ’ (a euphemism fo r her reproductive organs), and afflicts her w ith menstrual disorders. These disorders include m enorrhagia, dysmenorrhoea, and amenorrhoea. T he w om an m ay dream that her dead relative has appeared to her equipped like a hunter, wearing the red w ing-feather o f a lourie in her hair, as only those w ho have shed blood, such as hunters, manslayers and circum cisers, are entitled to do. Sim ilarly the m ain sym bols o f the ritual fall into the class o f red objects : red d a y , the bark scrapings o f a tree that secretes red gum , the blood and entrails o f a red cock, lourie feathers, and the like. T he nam e o f the m ode o f affliction in this ritual is called Nkula. „The nam e o fjth e ritual perform ed to treat the afflicted person is also railed Nkula. N o w the treatment.is carriedjouthy^a.nunffl.exLaf„^doctors„’_QiL^adepts * (ayimbuki or ayimbanda), both m ale and fem aleT w h o have them selves been closely associated w ith .previous per£brmances_o£Lthe Nkula ritual. T h e w om en doctors, m ust at one .time^havejbeen patients..theimselves and m ay not „become, practising- doctors unless they are .generally considered to h ave been.cmed_byjhej:reatirient. The mate-doctors. m ust have previ9jLidy_.ac-te4 ^ of ckaka-JzhaNkula,. that is. to. say , m ale .ritual. asshtant..o£imsi..Nkula patient, a role„w hich may. bep.eife«:medh^hec„hmb.aad,Jbjrother or son, „aiid.iixvolves foe.perform ance„ofa„few ^taaLtasks,~ such as guiding the patient out backwards fro m jh e spirit-rhut .to the ritual .fire. T h e doctors or adepts are arranged in a loose hier archy o f ritual prestige and each perform s a different ritual task. Betw een doctors_.and patients^there. m ayhe.„no„. ties-ofokinship or affinity at all— iCties. interlink them , ,w hich in secular situations w ould -be associated w ith m utual avoidance, -these_.ties are reckoned inoperative in the. context o f the ritual,_so_.that a man, X
294
Schism and C ontinuity in an A frican Society
in his role o f doctor, m ay splash w ith medicine the almost naked body o f a wom an w hom he calls 4 m other-in-law * or ‘ daughter-in-law and w hom he norm ally avoids. I£_a. patient has been treated successfully,. sheJs.jentidcd.^to_take^paxt-iix future performances o f the N ku la ritual.., as,.j^}w uot-.6octCH ^(chim bttki wanyanya, * Hide doctor ’). H er ritual status becomes higher the m ore she participates in rituals o f this type, and the m ore she learns about the medicines and techniques. T o whatever part o f the N dem bu region she m ay m ove afterwards she w ill still be entitled to participate as a chim huki (adept) in N kula . The spirit w hich had once afflicted her is n ow thought to give her healing pow er {ngovu yakuuka ) as a doctor, as w ell as to make her forme. It is ‘ her spirit * (m ukishi windi). ThisLjspirit is thought ,to have heen itself.an Nfe«/a patieiit aii(fjdoctox, during its lifetim e. T he ritual itself has exoteric and esoteric phases. D uring the exoteric phases, w hich typically take place in the village clearing, a big secular dance often accompanies the curative process proper, and as m any people as are able to leave their other tasks in the vicinage, and, indeed, in villages in other vicinages w ho have received tidings o f the ritual, com e to dance, sing, drink, gossip and lo o k on. T h e headman o f the village and the close kin o f the patient bu y and brew as m any calabashes o f beer as they can to satisfy the clamorous demands o f the throng o f dancers. I f the secular dance is a success, the prestige o f the headman w ill tem porarily be high, but i f there is insufficient or unfairly distributed beer and food for im portant attenders, it w ill sharply decline for a tim e, whether the patient recovers or not. The esoteric phases o f the ritual are attended b y doctors and adepts only. T h ey take place either in the secrecy o f the patient’s hut, or in the bush, w here medicines {yitumbu) are collected w ith song and sym bolic action, and w here towards the end o f several phases o f ritual, a chicken o r goat is sacrificed. In the past doctors drove the uninitiated aw ay, b y threatening them w ith bows and arrows, from areas in the bush w hich they had sacralized for ritual purposes. Bach culturally-defined mode^Taiflii^ojn^hy_a^spirit is associated-with a cult, whose adepts coflaborate-to-placate-the offended spirit. Since the m ost com m on canse.,o£,afflictk>n„is_the neglect in g or ‘ forgetting * (ku^vulamena) ,of.-the-spirit^the_m ost im -
The Politically Integrative Function o f R itu al
295
p cu tan tasp ect o f the_ process afip lacation i& .to hring..it. back to m em ory, * to m ake it kn ow n before m any peopleban d * to m en ricm its name in th ek hearing \ in 'so c io lo g ic a l terms, thismieans that in. a .m obile .and fissile society. th.ere_is.. a., strong tendency towards structural am nesia,-winch_.is^countered b y ritualslw M cli^ ontm ually. reviv.e.-tha.memos^-jo£-deacL-persons through whom , the Uving™ ^e_significandy„._jmerconnected. W hen the senior doctor addresses the .afBictmg^spint. at various points in a ritual he mentions a num ber-of-significant kinship connectionsof.thedeceased.. -Sometimes heLgives_a b rief outline o f the kind. o f beha\aaur w hich. clivinatioix.has .suggested as the cause ,o f the spirit’s anger__(This„o.utline, incidentally, gives the mthre^pologist-Jvaluable.-mfbrmation_about. the.Jkind&_of,conflict diatL often arise in d ie. kinship system A .. In_the..context, o f the ritual, situation, dus ritm g-ofLgenealagical. ties m ddiscussion o f contem porary conflicts, has .the dual_effect-on-l>idembjU..of reviv ing kinship a n d a ffin a l lhdcages.-which-.are_ be.gmuiug .to w ear thin, and .o f invoking the participation.-of tbe wider.-S.oeiety, represented b y the adepts draw n fro m . all .over the.. .N dem bu region, in the conflicts o f the. narrower-unit^of-the village. B y means, o f ritual,, the. im m ediate past and the__poHticorrkinship relations Originating in it are kept .alive,-in. .spite, .o f „the social forces „w o rk in g .in_the _app.Qsite__airection. The other cults have die same general structure and functions as Nkuia.1 In each, there is a band o f doctors and adepts, w h o share out am ong themselves the various ritual tasks. In each, there is a patient o r subject o f ritual, w h o is at the same tim e a candidate fo r initiation into the cult. In each, a named, deceased relative o f the patient is divined as the cause o f the affliction. This spirit afflicts in a specific m ode. T h e mode is com m on to all members o f the society : the spirit belongs to particular persons and narrow kinship groups. T he name o f the m ode is the name o f the ritual. T he same spirit m ay afflict the same livin g relative in several modes, or m ay attack several relatives in the same m ode. For exam ple, N yaluw em a o f M ukanza V illage was afflicted b y her m other's m other’s spirit w ith reproductive 1 For a preliminary classification o f Ndembu rituals see my * Lunda Rites and Ceremonies O ccasional Papers o f the R hodes-Livingstone M useum , New Series, No. 10 (1953)« A detailed, analysis o f Ndembu ritual is in preparation.
296
Schism and C ontinuity in an African Society
troubles in the mode o f Wubwangu, and w ith illness in the mode o f Chihamba. She is now an adept in both cults, and as w e shall see an im portant doctor o f Chihamba. O n the other hand, N yam ukang’a o f m alabu lineage in M ukanza afflicted her daughter’s daughter N yam asung’a, her daughter’s daughter’s daughter N yam ukola, and her daughter’s daughter’s daughter’s daughters Zuliyana and Yana, w ith illness in the m ode o f Chihamba. T h e spirit is m ost frequently a m other or a mother’s m other in w om en’s cults. Thus the cults act as a m nem onic o f direct matrilineal kinship links. V irilocal marriage, w hich may scatter wom en far from their m atrilineal kin, is to some extent counteracted by cults that stress m atrilineal ties between dead and living wom en. B ut sometimes a wom an m ay be afflicted b y a spirit from a different village sub-lineage from her ow n. N yatioli, N yam ukola’s sister, for exam ple, o f malabu lineage, had tw ice been afflicted b y N yachintang’a, apical ancestress o f the other village lineage, in the modes o f Isoma and Wubwangu. Thus the cause o f affliction is thought to be connected w ith a breach o f kinship norms in some specific kinship group. But the treatment o f the afflicted person is carried out b y an association o f ritual specialists w ho need not necessarily be related to him or her. Sinee-most.. adult. members.o£,any_.village-.are,.adep.ts,in_.atJeast one cult, and. since .members., o f ail...c.ults_miay_.he_.found_evjs;qrwhere in the N dem bu and Kosa regions, it YoHows .that the-total ritual system provides a set o f. interconnections—which~in effect perform a . pobticaLfonction. This w ill be clear i f I present a highly schematized picture o f the w ay ritual interlinks persons and groups in a vicinage, w ith reference to cults concerned with female reproduction only. There are four main cults w hich are today perform ed to placate spirits interfering w ith female repro-i duction : their names are Mkula, Chihamba, Wubwang*u and Isoma.1 Let us for convenience call these cults A , B , C , and D . Let us suppose that there are four villages in the vicinage, each containing eight adult w om en belonging to the village descent group. N o w let us suppose that in each village tw o wom en belong to each cult association. This means that the wom en 1 There is another cult Kalem ba which used to be frequently performed but is now extremely rare.
T he P olitically Integrative Function o f R itu a l
297
belonging to a single kinship group in each village are divided by membership in four a its , and that four cu lt associations interlink the four villages w hich, let us suppose, belong to four unrelated descent groups. N o w let us suppose that in each village there are eight w ives o f male village members, each w ife from a diffèrent village in a different vicinage. I f tw o w ives in each village belong to each cult association, each pair o f w ives is linked to tw o fem ale village members b y cult affiliation, and to fem ale members and w ives o f m ale members in all the other villages in the vicinage. T h ey themselves, since they originate in other villages and vicinages, act in effect as representatives o f the N dem bu as a whole in the context o f the ritual. N o w let us suppose that in each village one m an is a doctor or adept in each cult. In his ow n village he is linked through cult member ship w ith tw o kinswom en and tw o fem ale affines, none o f w hom , perhaps, is his ow n w ife. H e is cross-linked w ith one man and four w om en in each o f the other villages in his vicinage, and divided from all the other m en, and from three-quarters o f the w om en, in his ow n village, b y membership o f a cultassociation. Siblings and parents and children m ay be divided from one another b y cult affiliation, w hile the status o f members o f adjacent genealogical generations m ay be reversed during the period o f perform ance o f a ritual. W h en ritual is being per form ed in connection w ith cult A , members o f cults B , C , and D are excluded1 from the esoteric phases o f the ritual, although senior doctors o f these cults m ay be given sacralized beer or food in token o f their high ritual status in other situations. Sim ilarly w ith the other cults. Thus in a society characterized by the poverty o f its status system in secular social relations, the status differentiation w ithin each cult and the exclusion o f unin itiated persons, w hatever their rank or standing in other contexts, from its central mysteries, provides some compensation for the frustrations o f ambitious urges, or for the occupation o f an inferior status in secular life. I have o n ly m entioned one section o f the ritual system, in dealing w ith cults relating to fem ale reproduction. In addition, 1 Gluckman has shown in * The Role o f the Sexes in Wiko Circumcision Ceremonies \ that exclusion from the arcana o f a ritual is sociologically as significant as participation in them.
zgS
| ! j,-
Schism and C ontinuity in an A frican Society
there are tw o distinct but interlocking categories o f cults relatm g to huntsmanship ( Wubinda and Wuyang’a), one o f which (Wubinda) contains five separate cults. Social bonds o f mem bership in the hunters’ cult likewise interlink individuals in different localities, and even cross tribal boundaries. Then there are curative cults o f various kinds, a cult w hich concerns itself w ith the initiation o f diviners, and a number o f anti-witchcraft cults. T oday, new cults, often organized b y O vim bundu and Lw ena im m igrants, are directed against disease and misfortune caused b y the m alevolent influence o f living and dead Europeans, and, in some areas, these are beginning to oust the traditional cults. T o the m anifold ties interconnecting people in these cults m ust be added the links created by the boys’ circum cision ritual, w hich jo in m en in opposition to w om en,1 and by. the girls* puberty ritual, w hich jo in w om en in opposition to men. Furthermore, there is the M ungongi association w hich carries out im portant duties during the funeral rituals o f its members and w hich carries out an elaborate initiation ritual o f its ow n in connection w ith the funeral rituals. These_cidfclm lcages-can-.be-pohti.c»llyxSk:tive..only-if.ritual is frequently performed..^ I .do_not_know_j^hether_there has been any increase in the frequency o f performance o f rituals in recent years ; but I have personally witnessed thirty-one performances o f no less than fifteen different categories o f cults o f affliction and anti-w itchcraft cults, as w ell as m any life-crisis rituals, and have been notified o f the performance, in localities w here I was staying at the tim e, o f m ore than tw ice that num ber— mostly in m y second period in the field. It is likely that rituals were perform ed at least as frequently in the past, for a num ber o f rituals in the bow-hunters’ cult (Wubinda) have becom e extinct, as w ell as ritual perform ed b y the w om en’s Chiwila funerary association, and a cult called Musolu, perform ed b y a ch ief or senior headman, to bring rain i f there had been a drought at the beginning o f the rainy season. M any cults in vo lve the per form ance o f tw o successive rituals, separated b y a period during w hich the patient undergoes partial seclusion from secular life. T he first o f these rituals is in the w om en’s cults called ku-lembeka 1 C f. G luckm an, M ., * T h e R o le o f the Sexes in W ik o Circumcision Cerem onies *.
1
■
, i
The P olitically Integrative Function o f R itu a l
299
or ilemhi, and the second, and m ore elaborate and im portant, ku-tumbuka. A candidate w h o has undergone ku-lembeka m ay play a m inor role in the cult, but cannot becom e an officiant until she has passed ku-tumbuka. O ften a person, is considered cured, that is, freed from the ban o f her fertility im posed b y the spirit, after ku-lembeka, and m ay never undergo ku-tumbuka. T h e 'fa ct that m ost cults in vo lve tw o separate perform ances o f ritualJtnultipHes the num ber o f occasions o n w h ich each kind o f ritual m ay be perform ed__Again,. rituah„aresQme_times con sidered tohavef^edim theirm irpose__fQ ra_num ber^Q fjte3soiis. T he diiriner J w h o „p xj^ q o b ed .jfe..^ been deceived b y -a . w itch._O r he m ay have divined fo e.rig h t. spirit but difiijwmng_mo.de o f affliction,_O r severaLspirits_ m ay have been afflicting .the .patient^simuitaneously, in,..diSm nt_m odes o f affliction,.sodiatrituahm m tbe_successively_perform eclby.seyeral cult associations before a cure is efiected»_Thus^.personsuentitled to participate .as adepts in cults, are comtmtly._being_mobilized, and.new candidates fo r initiation, as._patients, .o r in .fo e guise o f patimts,- are constantly., undergoing..treatm ent. Som e w om en, such as Y ana (I,j4), daughter o f N yatioli (I,H zi) In M ukanza V illage, have been initiated into all the w om ens cults. Yana, w h o was o n ly thirty-four in 1954, was a patient in the ku-lembeka phase o f Nkula w hen she was about seventeen years old . Nkula in this case had been perform ed after the m iscarriage o f her first child. H er husband h ad sought out a diviner w h o had named her mothers’ m other N yam asung’a as the afflicting spirit. Yana considered the treatm ent successful, for her next child was bom successfully. A fte r about ten years she had menstrual troubles and her husband w ent w ith Sakazao her uterine u n d e to consult a diviner. H e diagnosed the renewed anger o f N yam asungV s spirit as the cause o f her trouble and recommended that the ku-tumbuka phase o f Nkula should be perform ed. A fter this, her m other N yatioli, w ho had been the senior fem ale doctor in the ritual, taught her the fu ll * m ystery * (mpang*u) o f Nkula, and its fu ll curative aspect (wumbuki). The follow ing year she became ill during pregnancy, and divination indicated that her m other’s m other’s m other had afflicted her in the m ode o f Isoma. T h e Isoma manifestation o f a spirit can cause abortions and miscarriage. T o prevent these she under w ent the fu ll ritual, in this case called ku-tumbuka, o f Isoma,
3©o
Schism and C ontinuity in an A frican Society
perform ed b y Kanyom bu, M ukanza Kabinda’s brother, and at that tim e her mother*s husband* There is no ku-lembeka at Isoma, B ut w hen "she bore her child w ithout mishap, the kwidisha, or ‘ causing to com e out * (i.e. from the grass seclusion hut), was perform ed, to celebrate the success o f the treatment. Yana also underwent as patient a perform ance o f IWubwangu, k.U'-tumbuka on ly. Wubwangu is most frequently perform ed for a w om an w ho has had twins, fo r her and their w elfare, or for a wom an w h o is expected to bear twins because she herself or her m other is a tw in. Yana passed through ku~tumbuka only, be cause ku-lembeka is perform ed during pregnancy and not after parturition. O nce the tw ins have been bom , ku-tumbuka is enacted. Yana's twins both died and the treatm ent was reckoned unsuccessful. Nevertheless, Yana claim ed that she w as entitled to act as a ju n io r adept (
T he P olitically Integrative Function o f R itu a l
301
ritual com m on in Ndem bu wom en but it is no exaggeration to say that m ost N dem bu w om en are initiated into at least tw o cult» concerned w ith fem ale reproduction between puberty and the menopause. D oes the frequency w ith w hich such ritual is perform ed indicate a h igh rate o f reproductive disorders ? I obtained the impression that it was indeed high, but not exceptionally high for A frica. I w ould like to postulate that the h igh frequency w ith w hich curative and ‘ gynaecological * ritual is perform ed by N dem bu is ju st as m uch socially as biologically determined. M isfortune, .illness and reproductive. troubles. am on g ,Ndembu, i f severe enough, are associated ..with the activities—o f spirits, witches and sorcerers. W h en these are vregarded.as.jdhe. causes o f affliction, one does not have to lo o k very fa r before..one. .finds confflcfc in social .relations. E very one o f the social dramas I. have usecLto illustrate and carry forw ard-the analysis, contains .at least one re£erence i:cLa supematural agency, incon nection w ith a social dispute._N dem bu—society,-.as-~w e-have-show n,—is-to m w ith
pe£eimi.aLdUspur_es>-there—i s ^ between its dom inant p rin cip les-o f sociaL_organization, anci in secularJife there is little to bind-together..m ore-than a small number o f people in habitual co^operation. U n ity in . such a societyisinaintained by m aking each-outstanding case, o f p er sonal m isfortune the occasion on w hich the .m oral norms and _values shared b y all N dem bu are prom inently displayed in a number o f waysrrrin prayer, precept, sym bol, m im etic action, .and in the rituahassociation o f those w h o have. suffered_regardless„o£. their kinship or other interlinkages, T he com m on m isfortunes o f m ankind..are. m ade occasians_for .restating_the_.common. values o f N dem bu. A m on g the .Ndem bu,.as w e have seen, each little m obile local unit_is_tom w ith continuous dissension. Nevertheless put, o f cojnfflct its e lf u n ity is engendered._Each, severe case of illness bringSrintQ^ the_ppen s.om ^raiikling dispute w ith in a group^-between-specifrc.persons w ith opposed such disputes .are irresoluble sincejhey. spring^from cont in the soeial structureitself.— Since they are ultim ately irresoluble the m ost that can be-done is to. brm g about.a.tem porary palliation, to produce, as it w ere, the illusion -o f harmony^ within- the dis turbed-group. This is. done b y stressing .in. each.little. localized
$02
Schism and Continuity in an African Society
performance o f ..ritual the co m m o n in tere^ .o fth e whole, sodety^ and,, what.. Fortes., and Ey^-Eritcharid,.calJl^
The P olitically Integrative Function o f R itu al
303
same. w ay.. J3u t afHiction is potentially...a baoii>^.not_merely a pre&ent'pain, fo r it is th e ra y a fro a d ta ritual em m encein a cult. Affliction b y a . spirit follow s a breach o f regular custom ary relations betw een livin g and^dead,_i.e_relations^ o f worship invob^gjthejaei^tioningjo£jrfie. .names, o f,the dead in. prayer, o r a breach between, livin g, and.Jiving,_vdh.emklmquarid,._which is fflen^puhished b y the dead. "Thus the unity or. a ll Ndem bu is only- perceived in situations.. arising_aut-ofothe-&tttf£&_of specific relations, usually couched in terms o f. .. kinship.— A— society continually threatened w ith disintegration is continually per form ing reintegrative ritual.. R itu a l among_Ndemhu_does n ot express the kinship and political structure as in a firm ly organized society ; rather it. compensates for„ffleird efIcien cies.in a labile society. In order to bring out m ore fu lly the im plications o f this analysis I propose to exam ine in som e detail the social background o f an im portant ritual w h ich took place near the end o f m y second tour o f fieldw ork at the village o f M ukanza. * Chihamba9 Ritual at Mukanza Village In this analysis I try to isolate social relationships and processes from their cultural integum ent as for as possible. I hope to make a detailed exam ination o f the cultural structure o f N dem bu ritual in a separate study. B u t som e account o f the main cultural features o f die Chihamba ritual is necessary, i f w e are to grasp d early its sociological im plications. T h e Chihamba ritual is said b y N dem bu to be a very im portant one {chidyika chalema nankashi, * a very h eavy ritual *). Som e claim that it is m ore im portant than the hoys’ circum cision ritual, others deny this. B u t all agree that its ku~tumbuka phase is usually attended b^r m any m ore people than com e to any other ritual except boys* circum cision. M any calabashes o f beer are brewed b y the village sponsoring it, and a large am ount in cash and kind is paid to the ritual officiants afterwards. Chihamba is a specific m anifestation o f an ancestor spirit. It is an exceptionally dangerous one, fo r w hile most manifestations o f spirits cause m isfortune, in fertility or disease only, a spirit that has * com e out in Chihamba 9 can kill the person it afflicts. T h e Chihamba manifestation o f a spirit afflicts its victim in several ways. M ost com m only it causes pains in the w hole body,
304
Schism and Continuity in an A frican Society
especially in the neck, and a feeling o f extrem e cold. Sometimes it is said to induce decay in crops planted by the object o f its wrath. Again, a man w ho is ‘ caught * by Chihamba may experience bad luck in hunting. It m ay also afflict wom en with reproductive troubles. T he Chihamba manifestation is, in fact, a sort o f compendium o f all the misfortunes that can happen to a person. B ut Chihamba is not sim ply the manifestation o f an angry ancestor. Another category o f supernatural being is also involved. A t several points in the ritual a strange supernatural being is represented in various ways. It speaks indecencies in a throaty vbice and brusquely questions the candidates. This being, called Kavula (a name know n in theory only to initiates), is not the spirit o f a dead person, but has ‘ his ’ ow n independent existence. Kavula is an archaic term for the lightning, and is perhaps etym ologically connected w ith nvula, * rain \ A t one phase in the ritual he is said to enter the hut o f the principal patient, as the Ughming does, through the grass * top-knot * {ntungu). O n the evening o f the first day o f the ku-tumbuka phase, a doctor impersonating Kavula, screened from the view o f adepts and candidates, addresses them in the principal patient’s hut in a throaty voice. A fter enquiring o f the candidates w h y they have com e to visit him , and after reviling them, he confers on them ritual names w hich o n ly Chihamba adepts are entitled to use. N ext day a contraption is made sym bolizing Kavula in a secret place called isoli w ithin a sacred enclosure in the bush about a quarter-of-a-m ile from the village. The contraption consists o f an inverted m eal-m ortar containing sym bolic objects and a fram ework o f sticks to w hich rattles are attached. O ver this is spread a blanket or skin whitened b y cassava meal. A concealed string attached to the fram ew ork can be manipulated by a male adept to ‘ make Kavula dance \ Just before nightfall the candi dates, each in order o f ritual status, but wom en before men, are brought to ‘ greet Kavula their * grandfather ’, ‘ the one who know s everything and must be praised \ O ne by one they crawl on their stomachs to the leaf-hedge surrounding the white image o f Kavula, banging their heads on the ground as they advance, first on one side and then on the other. W om en squeeze their breasts * to give m ilk to the ch ief ’ as they m ove forward. Kavula is compared w ith M w antiyanvw a at this moment. W hen he reaches Kavula, each candidate is told *to kill Kavula \ by striking
The P olitically Integrative Function o f R itu al
305
him on the * head * (really the inverted m eal m ortar) w ith a Chikamha rattle-—each candidate has such a rattle, o f special form elaborately carved for him or her before the perform ance. The candidates are then led aw ay b y adepts and brought bacfc to find the blanket gone and the m ortar running w ith the blood o f a sacrificed chicken. T h ey are told b y the adepts that they have killed * the c h ie f’ (mwantci) and that they have seen Kavula*s blood. T he w hole o f that day they have been driven, w earing only a waist-cloth, between the sacred enclosure and the principal candidate’s hut b y the adepts. T h ey have been called ‘ the slaves * (andungu) o f Kavula and have been forced to wear sym bolic slave-yokes (mpanda). Nkaka means * ow ner ’ (o f slaves and dom estic animals) as w ell as * grandfather *, It seems reasonable to infer that one o f the m any meanings o f Kavula is gerontocratic authority in general. Kavula jokes (ku-senseka) w ith adepts and candidates, as a grandfather does w ith his grandchildren. H e reviles them sexually in a thick, m irthrevoking, yet nevertheless, rather terrifying, voice. Some o f is esoteric names, Samasenga for exam ple, said b y informants to be derived from ku-senguka, 4 to m ultiply ’ (o f fruit)* suggest that he presides over fertility and that his authority has its benevo lent side. His associations w ith the rains, w ith cassava, and w ith other cultivated crops, support this view . H e can rem ove sickness and m isfortune and bestow health and fertility after his m ystical death. B u t it is diear that there is am bivalence in the relationship between Kavula and candidates. H e it is w ho strikes like lightning w ith disease, w h o treats the candidates like slaves, w h o forces them to grovel before him , w h o taunts them w ith teasing questions and reviles them , and w h o finally is bloodily killed b y them . This am bivalence is n ot surprising in view o f the strains and conflicts in the kinship system, and between slaves and their masters. Kavula is said to be a male. T he spirit o f the afflicting ancestor, on the other hand, is always fem ale, and its name is found by divination. O ne inform ant told m e that * Kavula is the husband and X (the aifiicting spirit’s nam e) is the w ife. T h ey jo in together to catch the patient.* In the Chikamha ritual there is norm ally a principal patient or candidate (the term muyeji has, o f course, both meanings), always a wom an, w ho has been ill or
g
306
Schism and Continuity in an A frican Society
unlucky in the ways mentioned above. A divination has been m ade for her and the diviner has said that ‘ such-and-such a spirit has com e out to her in Chihamba \ O ther sick w om en in the village m ay then also be thought to have been afflicted b y the same spirit and they are also nom inated as candidates. O n the m ost im portant day o f the ritual, that on w hich the candidates are chased back and forth like slaves from the sacred enclosure to the hut o f the principal candidate, the pursuing doctors may capture any persons o f any age o r sex w ho happen to cross their path and add them to the group o f candidates. People often get caught deliberately in order to be initiated, in spite o f the hum iliations to w hich they are subjected. Parents urge their young children to get caught, for this renders them im m une from capture at subsequent performances o f Chihamba and gives them the right to enter the sacred enclosure, i f not to approach closely to the central shrine. The chasing goes on all day until sunset. Each tim e the candidates are driven back into the sacred enclosure from the village, they are made to sit dow n w ith their backs to the central shrine in a long row , w hile adepts crouch beside them on either side asking each in turn a num ber o f riddles all concerned w ith the identity o f Kavula. M ost candidates m ake ridiculous mistakes in their answers at w hich the adepts laugh derisively. Each tim e they return they advance a little nearer to the central shrine where the sym bolic representation o f Kavula is being m ade. T hey are told that they w ill see * som ething terri fyin g*. U ntil the final obsequious greeting o f Kavula no candidate ever mentions his name. There is a fiction that only the initiated kn ow it, although I have obtained evidence that uninitiated people have heard o f it. Eventually th ey are told by adepts w h o are their friends to m ention the name o f Kavula when asked b y the senior doctors, w ho are always male. T he wom en claim that Chihamba was once their ritual but that the men * stole it from them *. M ale doctors and adepts concern themselves principally w ith the m ore esoteric aspect o f the ritual such as the construction o f Kavula's representation, w hile the w om en adepts act as a sort o f police force guarding the sacred enclosure from desecration b y the uninitiated, chasing the candidates and seizing those w h o cross the path between village and sacred enclosure. Thus m en are the m ore important adepts, w hile w om en are the principal candidates.
PLATE VI T
he
I n t e r r o g a t io n
of C h ih a m b a
C a n d id a t e s
The candidates sit modestly in a line with their backs to the small sacred enclosure (isoli) where Kavttla (see Chapter Ten) is represented. Nyamukola is the candidate nearest to the camera. Manyosa stands on the right, arms akimbo, holding her adept’s rattle while other adepts ask the patients, riddles. The child-candidates can be seen top-left, harangued by Headman Itota, who enacted Kavttla the previous night. Nyamuwang’a, the suspected witch o f Social Dramas IV and VII, is the bare-headed female adept near the centre o f the photograph.
G
T he P olitically Integrative Function o f R itu al
307
Kavula is impersonated in the patient’s h u t behind b y a screen b y the m ost senior m ale adept or an experienced adept w ith a suitable vo ice fo r reviling in the style required. Another senior adept makes the blanket-covered fram ew ork that represents Kavula in the bush. O ther m ale adepts take turns to manipulate the string that makes the structure shake and rattle. N o w om an m ay pass behind the screen o f leaves w here Kavula thus * dances *, but the leading fem ale adept m ay sit d ose to the front o f it. Both, m en and w om en adepts m ay chase the candidates, although m ore w om en than men do this. B oth m en and w om en m ay question the candidates, although men ask the m ost esoteric questions. A fter Chihamba is over, a special form o f friendship similar to blood-brotherhood and blood-sisterhood (wulunda) is made between adepts and those w h o have been candidates in the ritual, called tmbwambu or wulunda waChihamba. This umbwambu pro vides yet another means o f interlinking members o f different villages, fo r members o f the same villages are seldom united in this w ay. T h e adept w ith w hom a candidate is join ed in ritual friendship is the one w ho has helped him or her to answer correctly the final riddles put b y the senior m ale adepts, before the killin g o f Kavula. This necessarily abbreviated account o f the m ain features o f the hu-iumbuka phase o f Chihamba m ust suffice as an introduction to the sociological analysis o f a specific perform ance o f this ritual. It w ill be remem bered (see Chapter fiv e , p. 164) that the village o f M ukanza was rapidly approaching the crisis o f fission in 1954, near the end o f m y second tour o f fieldw ork. It seemed lik e ly that the village w ould undergo a three-w ay split, w ith Sandom bu revivin g the nam e o f Kahali V illage for bis farm and obtaining bis follow in g from Bibiana and her children ; Kasonda d ividing o ff w ith M anyosa and C haw utong’i and founding a fa rm ; w hile Sakazao, w ith o r w ith out Chinem a, and supported b y m ost o f m alabu lineage, w ould succeed to the name and position o f M ukanza. It w ill also be recalled that the relationship under m ost strain in this situation was die marriage between M ukanza Kabinda and N yam ukola w hich interlinked tw o village lineages and also tw o genealogical generations. M ukanza and N yam ukola w ere a m ost devoted couple, w ere the parents o f m any children, and w ould never divorce one
© o
© @ ©
©
©
© ©
©
©
€>
© © ©
©
©
© ©
© © ©
© o
© G 0
o
p :o o o .o :p o
p
io .c o
.o t) o O' &
o ,o
o
o (} c o ..o
308
Schism and C ontinuity in an A frican Society
another. H ostility between village lineages and generations did n ot express itself in marital quarrels between them but in criticism directed by other villagers against N yam ukola. M ukanza him self was rarely attacked ; in some w ays he seemed to be above the battle, as a personification o f the values o f village unity and persistence. B u t N yam ukola became involved in several quarrels, some o f them quite sharp. M ost o f her disputes were w ith members o f the intervening genealogical generation. Social Dram a V (pp. 157-61) includes a quarrel between Sandombu and N yam ukola. She also had a dispute w ith Chaw utong’i w hich led to the latter’s tem porary defection from M ukanza Village. N yam ukola had com plained publicly that w hen she had been to Kalene Hospital for treatm ent o f her leprosy, Chaw utong’i’s sons, Kasenzi (!,H 5), w ho w orked fo r (he European headmaster o f Sakeji School, w ith K aseloki (I,H6), w ho was a schoolboy at Kalene M iddle School, had o n ly given her a small piece o f beef between them. Chaw utong’i, w ho had recently visited Kalene Hospital, herself accused N yam ukola o f lyin g, and claim ed that her sons had given N yam ukola a lo t o f meat. As (he quarrel progressed m utual recriminations became m ore and m ore savage. Ghaw utong’i accused N yam ukola o f w itchcraft, claim ing that the latter had developed a grudge (chitela), the activating principle in witchcraft, against Kasenzi, because he had once, on receiving the present o f a goat from his father- and m other-in-law after the first pregnancy o f his w ife, charged N yam ukola and M ukanza some m oney fo r a portion o f its meat instead o f givin g it as an obligation o f kinship. A fter saying this she demanded com pensation from M ukanza for the slander his w ife had uttered against her sons. M ukanza became angry him self at the accusa tion o f w itchcraft and refused to pay. Kasonda then came along and said that it was true that Kasenzi and Kaseloki had given little help to N yam ukola. Chaw utong’i, b y this tim e in a tow ering rage, retorted that she could n ow see that she had no brothers to help her in M ukanza V illage, so w h y should she stay there ? She prom ptly ordered her docile husband N deleki to carry the fam ily’s scant possessions to M bim bi V illage where there was a vacant hut. Later, Kasonda became alarmed that he w ould lose Chaw utong’i ’s support fo r his projected farm , and took her part against N yam ukola w ho, he told m e, was an inveterate scandalmonger (mukwakukapekelela). It w ill be recalled that
T he P olitically Integrative Function o f R itu a l
309
N yam ukola had once accused Kasonda o f bew itching M ukanza (p, 140) and that he had then been defended by Chaw utong’i. B oth quarrels bring out the strength o f the com bined lineal and generation attachments, and the opposition between adjacent generations. N yam ukola was at the centre o f sundry other quarrels, and it is d ear that she occupied a position in the social structure subjected to considerable strain. In the latter part o f 1953, N yam ukola’s leprosy becam e worse and she began to suffer considerable pain from a large spreading sore on her left shoulder and upper arm. W e sent her to Kalene M ission Hospital and w hen she returned she seemed to experience a slight im provem ent in health. B ut she was still unable to do a full day’s w o rk in her gardens and in the preparation o f cassava meal. A private divination was made into the cause o f her affliction, and the verdict was that her m atrilineal great-grand m other N yam ukang’a (I,E5) had * caught her in Chihamba \ N yam ukang’a had been the daughter o f M alabu and the apical ancestress o f a m inim al lineage w hich induded Sakazao ana his sisters and their children, and w hich was structurally opposed to another m inim al lineage o f m alabu lineage that included N yam uw ang’a and Line. B u t her role as afflicting spirit probably did not so m uch derive from the contem porary social situation as from the fact that when N yam ukola was a young girl she was initiated at the ku-lembeka phase o f Chihamba perform ed for her m other N yam asung’a. T he spirit afflicting N yam asung’a in Chihamba was then divined as N yam ukang’a. It was said in M ukanza V illage that N yam ukang’a * wished to m ake N yam u kola a full doctor ’ (chimbuki wachikupu) in Chihamba, and w ould make her ill (mukishi wakamukatisha) until the ku-tumbuka phase was perform ed for her. B u t it is possible that N yam ukang’a was divined because her genealogical remoteness from the Hving enabled the inhabitants o f M ukanza V illage to assert their unity as a long-established village against the centrifugal forces o f factional struggle that threatened to destroy it. T h e fact that N yam ukola, the person m ost involved in current disputes, was singled out as the senior candidate in an im portant ritual, as the focus o f the unity o f a big ritual assembly, also suggests that the people w ere intuitively aware o f the threat to their cohesion and w ere dealing w ith it at its source.
3 io
Schism and C ontinuity in an A frican Society
I suggest that the people o f M ukanza V illage had three main aims in sponsoring Chihamba. T h e first aim was deliberate : it was to cure N yam ukola and her fellow-patients and to initiate other members o f the vicinage into the im portant cult m ystery. T he second aim , hardly conscious but proceeding from a col lective sense o f disturbance, was to reintegrate the village o f M ukanza itself, tom , as w e have seen, b y m any dissensions. The third, also hardly conscious, w as to effect a reconciliation w ith villages w ith w hich M ukanza V illage was at the tim e on poor terms, notably Shika and K afum bu Villages. T he first aim was directed towards the physical and m oral w elfare o f indi viduals, the second towards the betterm ent o f internal village relations, and the last towards the creation o f a clim ate o f friend liness and co-operation in the w ider systems o f inter-village dealings. A n analysis o f the social com position o f the ritual assembly on the day o f chasing and questioning w ould clearly reveal, even i f supporting evidence w ere lacking, that the last tw o aims w ere present. I list the candidates and adepts approxim ately in order o f im portance, and give the* village affiliations o f each, dividing them b y sex. C a n d id a t e s Name
W om en: Nyamukola Ziiliyana Yana Masondi Kalusaim Girls: Seliya Sakwimba Sam Mwendiana Didiya Delu Rene Rene’s twin sister Monica ? Men : Makanjila Chilayi Mungongu Wankie
Village
Mukanza Mukanza (Sandombu Farm) Mukanza Muchona Farm village in Ikdenge Area Sbika (mother from Mukanza) Mukanza (mother from Shika) Mukanza (mother from Shika) Kafumbu Shika Ntambu Ntambu Mukanza (mother from Shika) Chibwakata Mbimbi Nswanamatempa Nswanamatempa N g’ombi
R itual Status
afflicted by spirit ft ft IS captured by adepts ft
** »>
»? »?
9t ft
Nam e
Village
Boys : Kafwila Muryila Chikimbu Samuvvika Bisheki Jerry ? ? ? ?
Mukanza Mukanza (Sandombu Farm) Mukanza Kafumbu Mukanza (mother from Shika) Nswanakudya Kasai Chikang’a Chibwakata
311
R itua l Status
captured by adepts
99 99
*« ♦1
9> 99
Nam e
Village
Itota Mukanza (Sandombu Farm)
F.H. F.H.
99
99 99 99
99 99 99
99
99
99 99
99 99
junior adept
oldest adept (sat near isoli) principal female organizer wife o f Itota who impersonated Kavuta
Shika Mukanza (Sandombu Farm) Mbimbi Chibwakata (mother from Kamawu) stands for Village Headman. FJHL stands for Farm Head.
*9 99
99 99
'JJL
1 H.M.
senior adept ♦* 99
w
rsyamuieji Nyamuwang’a Nyawunyumbi Nyamalicbi
L J O O
Mukanza (mother from Kafombu) Itota (sister o f Mbimbi and Lambakasa)
9V
w v
Nyakantemba
Kafombu
organizer o f ritual arrangement senior adept
GLO
Nyaluwema
Sakutoha Muchona Kasai Lambakasa Kamawu Kamawu Kanyabu Kasai Chikang’a (Koshita Farm) Nswanamatempa N g ’ombi Nswanamatempa Nswanamatempa N g’ombi Kamawu N g’ombi
Kavuta
v
Sakutoha Muchona Sachinjungu F.H. Lambakasa H.M. Kamawu Biscuit H.M. Kanyabu Kilisha FJL Koshita HJM. Nswanamatempa HJM. N g ’ombi Chisanji Mboyunga Ndeleki Kapitula Spider Chanza Women t Katcndi
R itu a l Status
I C C L v L v L
A x>epts M en: H.M.1 Itota F.H.* Sandombu
CL C i LL C C
T h e P olitically Integrative Function o f R itu al
w
O
312
Schism and Continuity in an A frican Society Nam e
Nyamunyaka Mandarmi Mateng’ineng’i Makisa Nyakamboya Nyawatwa Nyaluwema Manyosa Mangaleshi Mika Nyamuhemba Kaluswika Nyakamawu Lumiya Mésala Disa Nyawatwa Susanna Nyamatong’a
Village
R itual Status
Kamawu ^senior adept Nswanamatempa Lambakasa »• Lambakasa (daughter of Nya~ mumwema woman headman) Nsanji *♦ „ Nswanakudya • Chibwakata Mukanza Mukanza (mother from Shika) 99 Mbwambu (Ntambu Area) » N g’ombi „ Chibwakata ** Chibwakata »* Mpawu (Ntambu Area) ♦» Mukanza (mother from junior adept Kanyabu) ,, N g’ombi (daughter of Lumiya) *9 N g ’ombi «* Nswanakudya U 99 Mbimbi ,, 99
This list gives some idea o f the scale o f the ritual and o f hetero geneity o f com position o f the ritual assembly. T he 27 candi dates o f w hom 13 w ere males and 14 females, and 9 w ere adults and 18 children, came from 13 different villages and farms (14 i f Sandombu Farm is reckoned independent o f M ukanza). Seven o f these villages, including M ukanza, w ere in the same vicinage ; six belonged to other vicinages. T en o f the candi dates w ere resident in M ukanza V illage at the tim e o f the ritual, but four o f them belonged to the village lineage o f Shika. O n the other hand Kalusa’s and N to lolffs daughter Seliya Sakwim ba (see p. 272), resident at Shika but belonging to M ukanza V illage lineage, was made captive b y the adepts, and her younger sister came to the public dance. There w ere 18 m ale and 26 female adepts present, from 19 different villages and farms, 6 (1 man and 5 wom en) being resident at M ukanza Village. 12 o f these villages w ere in the same vicinage as M ukanza, but the remaining 7 came from 6 different vicinages. Thus in all there w ere 71 adepts and candidates (to whom jo in tly w e m ay apply the term congregation), drawn from 20 villages in seven different vicinages. In addition there was a large num ber o f attenders at public phases o f the ritual. A t the
The Politically Integrative Function o f R itu al
313
night dance im m ediately preceding the chasing and interrogation o f the candidates, I estimated that there w ere about four hundred people present. I noticed N dem bu from every part o f the region, som e m arried into the vicinage, som e visiting kin and affines, others w orking at the B om a as labourers or clerks, and others in transit from one part o f the region to another. B u t Chihamba w ould not be such an im portant cult i f there w ere not m any uninitiated persons to be excluded from its arcana. In the vicinage o f M ukanza V illage alone the follow ing headmen and farm heads had never been initiated : N sang’anyi, senior headman o f the K aw iku (see p. 210 et seq.), Chibw akata (see p. 45), M ukanza him self, M bim bi (who was caught by the adepts and paid a sum o f m oney to free him self, after having been chased once or tw ice, m uch to the amusement o f every one), M akum ela, M ulongesha (m other’s brother o f Headman N g ’om bi), M akayi (younger brother o f Chibw akata), Nsanji (younger brother o f M bim bi and Lambakasa), N shim ba (O vim bundu efficient at Tukuka rituals), N yam um w em a (a wom an headman w hose daughter was an adept), ajid Anderson M ulum bi. T he last, w hose w ife had entered the sacred enclosure as an adept, w en t to fetch her to cook fo r him . H e was attacked b y tw o adepts, Chisanji and Spider, and beaten up. B u t he managed to drag his w ife aw ay, shouting out that Chihamba was * rubbish \ E veryone was scandalized and said that in the past he w ould have had to pay a heavy fine, perhaps a slave or a gun, and w ould have been com pelled to becom e a candidate in which, role he w ould have been beaten for his offence. B ut nowadays such punishments are forbidden b y the Europeans. Anderson was an extrem e exam ple o f the grow in g class o f m odem sceptics. B ut the old headmen excluded from the cult felt ashamed o f their uninitiated state. M oreover, they knew very little o f what actually to o k place in the sacred enclosure. A ll o f them knew the nam e Kavula, but none knew w h o or w hat Kavula was o r w hat w ere his attributes. T he status o f candidate can on ly be measured against the ritual exclusion and ignorance o f highly respected and pow erful members o f the local com m unity such as these. This short analysis o f the social com position o f the ritual assembly at the Chihamba I attended at M ukanza V illage brings out in detail some o f the points made in the general account o f
3 14
Schism and Continuity in an African Society
N dem bu ritual. The allocation, o f ritual offices to members o f M ukanza V illage reflects the contem porary balance o f power w ithin the village. N yaluw em a, representing Sakazao w ho was ntondu (uninitiated), w ithin the village stood for majlabu lineage, as w ell as fo r Kafum bu, her matrilineal village. T he three most im portant fem ale candidates, N yam ukola, Zuliyana and Yana, belonged to m alabu lineage. Sandombu’s increasing influence was reflected b y his office in the ritual, and b y the fact that his w ife Zuliyana was one o f the principal patients. His w ife was thus put alm ost on a par w ith M ukanza’s w ife. Kasonda was not so happily placed in this ritual. H e was, in feet, nearly turned back from the entrance (mukoleku) to the sacred enclosure when he came w ith m e on the day o f chasing and was only adm itted on the paym ent o f quite a considerable sum (nyishinguy ‘ paym ent to attend an im portant phase o f a ritual o r to purchase ritual privileges *), and then on ly on account o f his position as m y henchman. I was m yself rated as a muyejir * candidate ’, although the on ly ordeal I had to undergo was to be chased several times b y the adepts. Kasonda claim ed that he had been a child-candidate years ago in his stepfather’s village in Sailunga Area, and that therefore he m ight enter the sacred enclosure. W ankie Soneka, likely successor o f Headman N g ’orabi, was nearly debarred from entering the enclosure b y furious women adepts, w ho resembled so m any Bacchantes, but someone remembered that his father Nswanakudya had carried him beyond the mukoleku years ago, when he was a child in arms and his father was still a pagan. W ankie tried to participate with the senior adepts in the construction o f the isoli and o f the repre sentation o f Kavula, but he was haled before an im prom ptu court w hich m et at the isoli or * central shrine *, the site o f Kavula. It consisted o f Lambakasa, Sandombu, M uchona and Sachinjungu, w ho sentenced him to be chased w ith a sym bolic yoke o f mudyi w ood round his neck, and to make obeisance to Kavula towards sunset. The interesting point here is that W ankie, a recusant Christian schoolmaster, was m aking a bid to be considered a leading elder in the vicinage and w ent to considerable lengths to be regarded as an adept in Chihamba. T he ritual had evidently not yet declined into a mere form ality but was still socially significant. Kasonda’s w ife M angaleshi was an adept, and he saw to it that
o © o 0 0
© © ©
© © 0
© © n?
© PLATE VII Sandombu makes a C h i ha m b a Shrine eor his wife Z uliyana In the early morning following the day of chasing, a personal shrine is made by senior adepts before the hut of each adult candidate. A chicken is sacrificed and its head and entrails are impaled on the shrine. The shrine is called hantonga and consists o f a cutting o f cassava (a food plant elaborately ritualized in Chihamba tied to a bunch o f twigs collected from various medicine trees, each of which has a symbolic meaning. Maize and beans are planted around its base and watered. In this photograph Sandombu has just finished praying to the spirit. He used a beer bottle I gave him to water the kantong'a, since ! was his ritual brother. When the food plants sprout, food taboos will be lifted.
© ©
© ©
■4
:Q m o e
c
P b
■ O b V> 0 iO ■ (; £;■ m:.
O
X; to
4
The P olitically Integrative Function o f R itu a l
315
four o f his children w ere captured b y the adepts and initiated. His elem entary fam ily was w ell represented in the ritual. His sister M anyosa w as an adept, though her husband Chikasa was not. It was rather pathetic to see Chikasa dress up in his hunter’s gear, although he was not goin g anyw here, and w alk along the path on w hich the candidates w ere being pursued, as though to dare the adepts to seize him . T h ey did not lay hands on him in fact, although they had m ocked old headman M bim bi (see p. 313), perhaps because they sensed the brittleness o f Chikasa’s sen-respect. N yaluw em a also acted as a lin k w ith K afum bu V illage lineage, to w hich she belonged b y descent. H er older sister Katendi was treated b y other adepts, m ale and fem ale, w ith the greatest respect. Kafcendfs eldest son, headman o f K afum bu V illage, was a Kavuta im personator, but he was unable to attend the per form ance. Som e Kafum bu children w ere initiated, and the ritual was generally reckoned by M ukanza people to have benefited the relations between the tw o villages, w h ich had been a little strained fo r some tim e. I was given the Chihamba adept-name o f Sam lozang’a N dum ba, w hich made m e a ritual namesake o f Headm an K afum bu. W e thus had the righ t and duty to exchange gifts w ith one another. In this w a y I became a social link betw een the tw o villages. Several Shika people, too, form ed part o f the congregation, and m any members o f that village attended the public dances and beer drinks. T h e hostility that had sprung up over Kalusa’s death-paym ent (see p. 269 et seq.) between the villages was to some extent reduced b y com m on participation in the ritual. T he m ajor social effects o f the Chiham ba ritual at M ukanza V illage m ay be sum m arized as follow s : (1) It reduced the hostility felt b y m any village members, especially o f n yach in tan g ’a lineage (including Sandombu, Kasonda, and C haw utong’i), against N yam ukola, the senior w om an o f the village, b y m aking her an object o f sym pathy and the focal personality o f a great public ritual. (2) It was an attem pt to close the breach opening between different factions w ithin the village b y establishing ties o f ritual collaboration betw een leading members o f each faction.
3x6
Schism and C ontinuity in an A frican Society
(3) & gave prestige to M ukanza Village in and outside the vicinage. (4) It re-established friendly relations w ith villages traditionally linked b y kinship and affinity w ith M ukanza, but which, for that very reason, w ere constantly quarrelling w ith Mukanza. (5) It drew m ore closely together lon g and recently established villages and farms in the vicinage : it w ill be remembered (pp. 285-6) that in 1952 Governm ent m oved m any o f the villages m entioned in this account o f Ckihamha into the vicinage in order to m ake their form er territory a Forest Reserve. (6) It made a dramatic restatement o f the values o f an indigenous society w hich w ere beginning to fall into abeyance. (7) It offered an alternative source o f prestige to those with lim ited authority in secular life— daughters w ere in some cases candidates and adepts but not their mothers, w ives but not their husbands, sons but n o t their fathers, nephews but not their mothers* brothers, commoners but not headmen, and so on. (8) It afforded opportunities to ambitious persons such as W ankie and Sandombu to enhance their prestige. Chihamba is only one, though a m ost im portant one, o f a dozen or m ore cults w hich resemble it broadly in form and social function. D ifferent aspects o f human suffering and misfortune are ritualized, and the rituals have a politically integrative role. I hope to establish, in a subsequent book, how , in the course o f a ritual, symbols and verbal behaviour are manipulated so as to discharge tensions in the social system and to reintegrate the members o f the ritual assembly into the disturbed social groups to w hich they belong— at all events for a short tim e after the ritual is over. Here. I .can o n ly state that the cultural structure o f an Mdembu ritual is consistent w ith its social function. R itual for N dem bu is closely, associated w ith .breaches, in social regularities and their redress. It is n o t so m uch a buttress .or. auxiliary o f secular social regularities as a means o f restating, Jtime.and. again, a group unity w hich transcends,.but to some extent-rests-^on and proceeds-out. o f,. the._mobiHty..-and_confHcts--o£ its -.component elements. I n conclusion,. it must. be.. stated...thatljthc„systein o f cults o f
© T he Politically Integrative Function o f R itu a l
o
3 17
affh&km . and-o£-.lifi^risisL .rituaJs_such as .Ckihamba, Mukanda or Wuyanga is to be h e ic fth e headman o f the sponsoring village sends a messenger (ntem esha)xothe ch ief or senior headman .o f the area in w hich-his villageJs^situated w ith a calabash -o fb e e r, to ‘ o b ta in fire ’ {nakutambulaJkesj)— in other words,, to o b tain permission from the senior to h o ld -th e ritual« T his^ actisthe^ pointofintersectionhetw eentw .ovirtuallyautonomoils systems, each o f w hich has the function o f .maintaining die highest_.commojnL Jvalues.-.of_the_whole. Ndembu-people«- T he ch ief and senior headmen,..it. .w ilt be. shown- in- the next chapter, enshrine, and em body in then tides, and ritual frmcrions-Ndembu history, and the link w ith M w antiyanvw a— they represent the historical unity o f the people. T h e x u lt as&ociations maintain com m on. values and reaffirm norms in relation alw ays to the contem poraxy situation, tOL.lQcal and transient hreaches in social relations^ to the here and now .
&
0 ©
a ©
© 0 ,© 1
©
© ©
© ©
© 0 0 ©
© O
© 0 © Cl
© ©
©
ii ■ i. } %) W'
( ) \p
CH APTER XI TH E CH IEFTAIN SH IP
• ? o
:n ■Ti., rf'\
•c if ■■
R O F E S S O R E V A N S-P R IT C H A R D , w riting o f the king ship am ong the? Shilluk,1 summarizes the king’s position as follow s : * T he king symbolizes a w hole society and must not be identified w ith any part o f it. He must be in the society and yet stand outside it and this is only possible i f his office is raised to a m ystical plane.’ In substance, this form ulation also holds good fo r the traditional position o f the N dem bu ch ief Kanongesha. It is difficult to obtain a clear and coherent picture o f the N dem bu chieftainship as it existed before the Europeans came. N dem bu territory has been divided between Britain and Portu gal ; and rival Kanongeshas in Rhodesia and A ngola have arisen, whose authority and functions have been considerably modified b y the different theories o f governm ent espoused by the respective colonial powers. Nevertheless, it is certain that Kanongesha was in the past a ritual rather than a political head, and was never able to exert effective political control over his senior headmen. From accounts I have collected from both Kanongeshas and from all the Rhodesian N dem bu senior headmen it w ould seem that the first Ndem bu did not all invade their present region in a com pact body, but m any came in separate groups led by senior headmen, and that, although they all recognized Kanongesha as their chief, each o f these headmen continued to maintain a con siderable degree o f autonom y. Y et they collaborated in national ritual and each o f them had a ritual function 2 to perform at the funerary and installation rites o f a Kanongesha. I f Kanongesha asked them to help him to send tribute to M w antiyanvw a they w ould usually com ply ‘ out o f respect \ but w ere Kanongesha to try to exact it by force I have been told he w ould have been fo rcibly opposed.
P
1 Evans-Pritchard, E. E-, T he D iv in e Kingship o f the S h illu k o f the N ilotic Sudan (1948), p. 36.
2 For example, at the installation, K ahungu placed die chiefly lion- and leopard-skins on Kanongesha’s chair, while Ikelenge removed them afterwards, in virtue o f His office as K alula. Mwininyilamba, as Ifu/ota, * the vanguardleader chose the site o f the new chief s capital village : and so on.
318
The Chieftainship
319
Already in the first h a lf o f the nineteenth century, before foe period o f slave-raiding and -trading on a large scale, foe N dem bu must have presented foe picture o f a scattered, foinly-spread, h igh ly m obile and locally autonomous people, similar to foe C h okw e, Bangala and other Angolan groups described b y Carvalho and C apello and Ivens. Y e t they possessed great pride and fierce consciousness o f their direct and exalted origin in the em pire o f M w antiyanvw a. This consciousness o f origin was epitom ized in foe office o f Kanongesha chieftainship. B u t that office was m uch m ore than an index o f tribal identity 4 it repre sented tribal rights to foe use o f land. A lth ough Kanongesha w ielded hardly m ore political pow er, in terms o f control over organized force, than any other im portant headman, foe ritual com ponent o f his status was bound up w ith N dem bu rights in the occupation and exploitation o f a large region over w hich N dem bu felt free to roam as they pleased. These rights had been reserved to them , according to tradition, b y foe mutual agreement o f a num ber o f free and equal Southern Lunda chiefs (who had conquered land both for themselves and also in order to extend foe political influence, i f not the direct role, o f their overlord M w antiyanvw a), to define and henceforward m utually to respect foe boundaries o f one another’s territories. Another party to foe agreement, according to one tradition, was foe Lwena ch ief Chinyam a, w ho also traced his descent from foe M w antiyanvw a dynasty. B ut Kanongesha reigned rather than ruled, and he reigned w ith m any o f foe attributes o f the D ivin e K in g as set forth b y EvansPritchard and in classical ethnography. In this capacity his ritual powers w ere lim ited b y and com bined w ith those held b y senior headmen o f foe autochthonous M bw ela people w ho made submission after long struggle to their Lunda conquerors. Pow er to confer and periodically to medicate the supreme sym bol o f chiefly status, foe lukanu bracelet, made from human genitalia and soaked in sacrificial blood at each installation, was vested in foe H um bu (a branch o f M bw ela) headman Kafwana, called Chivwikankanu, * foe one w ho invests w ith lukanu \ H e also had foe title o f Mama ya Kanongesha, * M other o f Kanongesha because he gave sym bolic birth to each new incum bent o f that office ; Kafw ana was also said to teach each new Kanongesha the
320
Schism and Continuity in an A frican Society
medicines o f sorcery w hich caused him to be feared b y his rivals and subordinates— yet another index o f deficient secular authority. The first lukanu was said to have been given b y M w antiyanvw a to the first Kanongesha, N kuba, and to have been a replica o f his ow n. W hen N kuba died, his younger brother Sakapenda and his son Kabanda reported his death to M w antiyanvw a. M wanti yanvw a made a new lukanu and asked Sakapenda to advance between a gauntlet o f warriors along a line o f leopard skins to receive the lukanu from his ow n hand. B ut Sakapenda, afraid o f treachery, w ould not m ove. Then Kabanda w alked forward fearlessly to* M w antiyanvw a w ho gave him the lukanu and said that he should succeed his father, for he was a brave man, not a cow ard like Sakapenda. Later Sakapenda is said to have bewitched Kabanda out o f jealousy, but then became reconciled, and Sakapenda and Kabanda jo in tly planted a muyombu tree to the spirit o f N kuba. Then Kabanda gave Sakapenda the office o f Ntete Mwenimajamu, * the one w ho looks after the nail-parings and the graves * o f the Kanongeshas and allocated him some territory over w hich he m ight preside. B ut his descendants were excluded from the succession from that day. It was about this time, it is said, that Kafwana, w ho had fought against Kabanda, subm itted to him and was given the office o f Chivwikankanu, ju st as Nsang’anyi, the K aw iku headman, became Chivwikankanu to Kabanda’s son, the first senior headman M ukang’ala (see p. 212). It w ould seem that in those early days Kanongesha’s office conferred a higher degree o f political power than it did at a later period, and that his political authority was set at the opposite pole to the ritual pow er possessed in different w ays b y Sakapenda, M wenim ajam u and Kafwana C h ivw i kankanu, in an opposed and com plem entary relationship. T he lukanu was ritually treated by Kafwana and hidden by him during interregna. The m ystical pow er o f the lukanu, and hence o f the Kanongesha-ship, came jo in tly from M w antiyanvw a and K afw ana : its em ploym ent for the benefit o f the land and the people was in the hands o f a succession o f individual in cumbents o f the chieftainship. Its origin in M w antiyanvw a sym bolized the historical unity o f the N dem bu people ; its m edication b y Kafwana sym bolized the unity o f the land— o f w hich K afw ana was the original ‘ ow ner \ The daily prayers addressed to it b y Kanongesha w ere for the fertility and continued
'0 5 © T he Chieftainship
3.21
health and strength o f the land, o f its animal and vegetable re sources, and o f the people. O n the negative side, its use by Kanongesha m cursing— he touched the earth w ith it w hile uttering a form ula— meant that the person or group cursed w ould becom e barren and their land infertile and devoid o f game. In the lukanu, finally, Lunda and M bw ela w ere united in the jo in t concept o f N dem bu land and folk. In a political system as fluid and m obile as that o f the Ndem bu it w ould be a mistake to lo o k for undue rigid ity in structure. Ndem bu say that the early Kanongeshas parcelled o u t specific tracts o f territory am ong the leading Lunda w h o had accom panied them from M w antiyanvw a. Later they gave further land to their sons. Thus Kabanda is said to have sent M ukang’ala to the Chitunta Plain area not far from the K aw iku Plain, to keep an eye on the K aw iku villages. O ther chiefs* sons w ho were given territories w ere K afw eku (see p. 13) in Rhodesia, M w eni Am beji (* ow n er o f the Zam bezi ') and M ulum bakanyi in Angola. A ch ief's son w ho was thus given a territory (mpata) o f his ow n was called the chief's Mwanawuta and was debarred there after from consideration as a candidate for his father's chieftain ship. U n til the second h a lf o f the nineteenth century succession to the Kanongesha-ship had been m odelled on succession to the M w antiyanvw a stool. T he first ch ief's sons b y his tw o principal w ives, supplied in the N dem bu system b y the matrilineages descended from the original pair o f w ives N yakaseya and N yachilesya, succeeded h im ; theoretically in order o f birth, hut actually according to the choice o f a sort o f * electoral college *, consisting o f senior headmen (excluding those called Mwanawuta), from all parts o f the N dem bu region. Succession passed along a line o f brothers, and then descended to the sons by the senior w ives o f the oldest brother. A fter this, theoretically, it should have passed to the sons o f the oldest son b y his senior w ives, but in practice, the sons o f form er chiefs w ould assert their claim to succeed. Succession wars w ere said to have fol low ed. In the reign o f Kanongesha Ishima W atuta Menjji (‘ w ell o f co o l w ater *), about 1890, the ch ief in council decided to change the m ode o f succession to m atriliny, to bring it into line w ith the custom follow ed in the villages. Ishima also established the new position o f C h ibw ika, the nom inated suc cessor o f the incum bent, and appointed his sister’s daughter's
0
0 ©
0
© © ©
© © 0 ©
© © ©
© ©
© @
€1 © ©
© ©
©
322
Schism and Continuity in an African Society
son, Chababa, as the first Chibw ika. B y this means he apparently hoped to divert against the Chibw ika the current o f intrigue against the incum bent unceasingly maintained b y hopeful suc cessors. Ishima apparently had no sons himself^ o r ow n sisters* sons. M uch pressure had been put on him b y his matrilineal kin to change the m ode o f succession. Since his death matrilineal kin o f form er Kanongeshas have succeeded Ishima both in Angola and Rhodesia. T oday, in M w inilunga D istrict, where there is also a Chibw ikaship parallel to the one in A ngola, the C hibw ika is nominated b y the Kanongesha, but the nom ination is made only after the senior headmen have been consulted, and no Kanongesha w ould dare to run counter to their wishes. There are a num ber o f villages belonging to the chiefly maternal descent-group in the D istrict, and there is a convention that the Chibw ika-ship, and hence the Kanongesha-ship, should circulate through them and not be m onopolized by a single village. Thus there should be no close relationship, between Kanongesha and Chibw ika. Kanongesha M ulum biV successor in Rhodesia, Kafuleji, came from K alw iji V illage in C hibw ika Area ; the present Kanon gesha came from Nyakansehila V illage in Kanongesha Area proper ; the present C hibw ika from W achikeka V illage in Chibw ika Area ; and it is thought that the next C hibw ika w ill be chosen from N swanakudya V illage (see pp. 207-10). The fact that chiefly villages are scattered through the Ndem bu region— there are some in Ikelenge and M w ininyilam ba Areas— and that each is situated in a different vicinage from the others, gives the w hole people an interest in the chieftainship. Thus, i f a member o f Nswanakudya V illage became Chibw ika (and thereafter Kanongesha), senior men in the vicinage, from such villages as Chibw akata, M ukanza, N g ’om bi and Nswanamatempa for example, w ould seek appointments at court as councillors, assessors, kapasus, road capitaos, etc., or w ould build in the vicin ity o f the new ch ief in order to obtain privileges as his kin and affines, givin g him in return their support and protection against his rivals and enemies. PatemaUy-lmked kin o f a ch ief are often singled out for special favour and provide him w ith his m ost trustw orthy support. In the capital villages o f Lunda chiefs where succession is matrilineal, one often finds a com pact block o f paternal kinsmen o f the ch ie f near the chief's ow n enclosure.
The C hieftainship
323
U terine kin o f the ch ief are also trusted, but not m atrilineal relatives from other chiefly lineages. W hen a ch ief dies the inhabitants o f his capital village often break up into sm all groups, each usually a m inim al m atrilineage o r a uterine sibling group, and found n ew villages some distance aw ay from the capital o f the new chief. Beneath Kanongesha in the traditional political system w ere the senior headmen o r ayilolu, w h o consisted o f the matrilineal descendants o f the first N dem bu leaders w h o entered the area and the sons o f chiefs w ho had been given an area (mpata). Chilolu is a term indicating relative social position w hen one person refers to another.1 In a village a chilolu is a person w ho does n ot belong to the village m atrilineal descent group. Thus N deleki (see p. 101) in M ukanza V illage m ay be described b y Kasonda or Sandombu as our ‘ chilolu . In the same w ay Kanongesha refers to headmen w h o do not belong to bis m aternal descent group or w ho are unrelated to him by paternal ties as his * ayilolu \ B y an extension o f meaning ayilolu, w hen used b y Kanongesha, cam e to mean im portant headmen w ithin his territory w ho m ight n ot enter into the succession to his chieftain ship. The term then came to he applied to senior headmen. Such headmen w ere not called * Sub-Chiefs * 2 am ong N dem bu, although the occupants o f sim ilar positions in fshinde’s area in Balovale D istrict w ere called anyanta yamutayi (‘ branch chiefs *).s T he term * mwanta *, * chief*, can be applied to the head o f any local unit o f w hatever span, qualified b y the nature o f the unit o f w hich he is head. Thus Kanongesha m ay be know n as mwanta wampata, ‘ ch ie f o f the country*. Senior headmen such as 1 C . M . N . W hite (persona! comm unication) holds that the basic meaning o f chilolu is * a headman upon w hom a chief relies fo r support, inform ation, etc.* The village usage I cite he considers to be a slang one and not the prim ary sense. H e w rites that * ayilolu are sometimes referred to as mhwambu jamutung*a> ** the nails that keep die country together ” *. B u t i f the village usage is a slang one it is very w idespread. W hen I collected census m aterial about m arriage, m y inform ants, i f they had not m arried kin (cross-cousins or * grandchildren *) w ould tell m e they had married. * a chilolu *, i.e. a person not o f their descent group. 2 See, however, page n , footnote x for discussion o f this point. 8 Thomas Chmyama, The Early History o f the Balovale Luttaa (1945), p. 8. I expressly raised this point with Ndembu informants in discussing Chinyama’s book.
324
Schism and Continuity in an African Society
Ikelenge, M wininyilam ba, Nyakaseya and M ukang’ala m ay also be so termed. Sometimes Kanongesha is distinguished from the others b y the adjective * g rea t’ {tueneni) w hile the others are ‘ small ’ (wanyanya). T he head o f a village is usually called mwanta wamukala (‘ ch ief o f the village \ pi. anyanta anyikala) whether he is a chilolu or not. O ther terms used for a village head are tnweni- or mwini-mukala (* ow ner o f the village *), and ntungi. B ut the political system o f the N dem bu is no simple pyramid o f authority, at the top o f w hich is Kanongesha, beneath him the ayiloluy and beneath them again the anyanta anyikala. It is true that according to custom toe ayilolu had the right to tribute from the headmen living in their areas and certain joints from game killed in their areas, h a lf o f w hich they w ere supposed to send to Kanongesha. It is also true that they w ere entitled to display certain insignia such as a short sword, a bead crown (chihangulu), a slit-gong (chikuvu), a xylophone (mudyimba), and a bangle. T hey also presided over courts to w hich w ere sum moned the best mahaku o f their areas and w hich served as appeal courts w ithin the areas. T h ey had m ore elaborate installation rituals than village headmen, and certain headmen gave them tributary w ives. T h ey possessed special m edicine not know n to ordinary village headmen, for protecting their capital villages from w itchcraft ; and they w ere held to kn ow m any sorcery techniques for punishing those w h o offended them . Neverthe less it was difficult for a chilolu to enforce his rights on proud and independent headmen w ho, like M ukanza Kandulu in C h ief Ikelenge’s area (see p. 105), w ould unhesitatingly quit his area i f he tried to gain his ends b y force or even b y the threat o f force. T he social com position o f a chilolu*s area was always changing. It m ight consist o f tw o or m ore vicinages, each composed o f villages belonging to various maternal descent groups. In each vicinage there m ight be one or m ore villages belonging to the chilolu*s maternal descent group. In his ow n vicinage w ould be found his closest matrilineal kin, a group belonging to his father’s maternal descent group, and a num ber o f ‘ strangers ’ (angeji, antu achengi) w hom he had given permission to settle in his vicinage, fit other vicinages (ayitungili) w ithin his area (mpata) m ight be found villages whose nuclei consisted o f his more remote matrilineal kin, others founded b y members o f his
The C hieftainship
325
predecessors capital village, notably paternal relatives o f his predecessor, and villages o f * strangers * {angeji). In vicinages other than his ow n w ithin his area one headman was usually recognized as the m oral leader, i f n o t as the political head. This w as usually the headman o f the village w hich had been established longest in the locality, and w h ich had m ost abandoned residential sites, gardens and graves in it. Its headman was called the mwenimbu,* and he usually, though n ot invariably, sponsored the boys' circum cision ritual w hen it Was decided to hold it in his vicinage. H e, rather than the chilolu o f the area, conferred or w ithheld permission to an incom ing group to build and garden in his vicinage. B u t he m ight not give such permission w ithout first consulting the other headmen and mahaku o f the vicinage. Bach mivanta watnukala had direct access to the chilolu o f the area, each paid him tribute (;tnulambu, or m ore com m only in the' plural, nyilambu) d ire ctly ; the mwenimbu o f a vicinage did not act as the chilolu*s agent. W h en the people o f a village elected a new headman the chilolu and n o t the mwenimbu, o f the vicinage confirm ed their choice,* although the mwenimbu, and indeed any elder o f the vicinage, m ight speak on beh alf o f any candidate in the election. T he givin g o f tribute was regarded as a m oral obligation rather than as a com pulsory matter— ultim ately as a recognition o f the historical origin and unity o f N dem bu in M w antiyanvw a. It is said— probably as a pious statem ent o f the norm — that when ever Kanongesha asked his ayilolu fo r tribute to send to M wanti yanvw a, and the latter asked their village headmen, it was never refused. T heir N dem bu-hood was an aspect o f their Lundahood, and both w ere m aintained b y w hat am ounted to a voluntary recognition o f their traditional leaders. I have spoken o f the N dem bu as being extrem ely individual istic, even anarchistic in their tribal character. This tendency, fostered b y the individualism o f their econom y, is strongly offset b y elaborate conventions eryom ing form al respect in the matter1*3 1 T h e prim ary sense o f mwenimbu is a person w h o lives at a villa g e and has his d om icile there. Thus a ch ild is a mwenimbu a t his ow n village. T he antonym o f mwenimbu is ngeji, * a stranger \ Ngeji, like the Larin hospes, means ' stranger * and * guest ’. 3 In th e past, I am told , a headm an w as given a piece o f cloth b y the chilolu as a token o f recogn ition .
:o :o n
I i O A) I. t;
Schism and Continuity in an African Society o f greeting and in behaviour at a ch ief’s court. Greetings between headmen express their relative status in terms o f criteria based on traditional history, and do not reflect their present circumstances. T he chiefbinship is a ritual and m oral structure w hich epitom izes the unity o f Ndem bu. It is a fixed emblem o f a m oral unity w hich is dynam ically i f indirectly maintained b y the cult associations. A lthough respect is shown to the ch ief and ayilolu in form al situations, N dem bu village headmen had no hesitation in fighting them in the past. It is said that the Angolan N dem bu headman Kapanga attacked Kanongesha Nkom esha in the nineteenth century to lo o t his capital village and take his people as slaves. M atem bu V illage attacked Ikelenge, one o f Kanongesha’s principal ayilolu, in alliance w ith several villages from the area o f M ukang’ala, another chilolu, but w ithout M ukang’ala’s participation in the rebellion. W hen the C h okw e invaded N dem bu territory, resistance was organized not b y Samuhang’a the Kanongesha, but b y Chipeng’e, a village headman. Chipeng’e later m arried Kanongesha’s sister and as ch ief’s brother-in-law acquired the title o f Sambanza. The ch ief was afraid o f him and fo r a tim e Chipeng’e established som ething like a dictatorship o ver part o f the Ndem bu area. I f tw o persons o r groups w ere engaged in a dispute one party w ould attem pt to invoke Chipeng V s aid b y means o f gifts. I f he was satisfied b y the presents or i f he thought the case o f the donor was just, he w ould force the other party to give compensation to his ‘ client *. B ut Chipeng’e was said to have respected custom and loved justice. O nce he became involved in a dispute w hich put him into opposition to the pow erful chilolu M w ininyilam ba, his ow n close matxilineal kinsman. 4M w ininyilam ba * was a famous historical title and Chipeng’e refused to figh t against its holder. In this he was bound by tradition w hen he m ight w ell have established a new dynasty o f his ow n. He w ould n ot attack either Kanon gesha or M w ininyilam ba because they represented the unity o f the people fo r w hom he had fought so w e ll against the slaveraiders. Y e t chieftainship and the possession o f a senior headmanship did give their holders a greater degree o f political control and econom ic advantage than an ordinary village headmanship. For this reason also such offices were struggled for and in the past
The Chieftainship
327
tenure o f them was typically short. Faction fights, poison, assassination, took their toll o f incumbents. Chieftainship and senior headmanship seem to have becom e exceptionally coveted positions in the period w hen the O vim bundu slave-traders made frequent visits to the N dem bu region. These traders set up posts near the ch ief’s and senior headmen’s capital villages, and made agreements w ith them to give them preference in the purchasing o f guns, d o th and other goods. T he traders even encouraged these pow erful N dem bu to raid their ow n village headmen to obtain slaves for sale. Perhaps this is one reason w h y die ch ief and senior headm en could not rally their people later against the C h okw e raiders w ho did n ot discriminate betw een senior head m en and other N dem bu in their depredations. E ven in these circumstances, how ever, the chitolu seems to have been little m ore than a village headman w rit large, unable to coerce the headman in his area to obey his w ill. M ost cases at law w ere settled in the vicinages. Self-help was resorted to in the event o f hom icide. A s fàr as I have been able to ascertain, the senior headman possessed o n ly tw o functions in w hich he acted as acknow ledged leader o f his area. R ecourse to his poison-oracle could be made b y any person nam ed as a sorcerer or w itch b y divination. I f the senior headman’s poison-oracle reversed the diviner’s decision the accused— no longer stigm atized as a sorcerer— had to pay a slave or a gun to the senior headman. T he senior headman’s other function also was o f a ritual nature and consisted in his officiating at the Musolu ritual, at w hich he was assisted b y the im portant elders o f his area, led b y his ‘ em blem -purifier ’ (Chivwikankanu)yto bring on belated rains and restore the fertility o f crops and animals. B oth these functions w ere ritual rather than political, although in consonance w ith the general orientation o f N dem bu society, the ritual had political aspects. In its general form , then, N dem bu society falls into that category o f societies o f w hich Professor Evans-Pritchard writes th a t1 * the political organization takes a ritual o r sym bolic form w hich in politics w ith a higher degree o f organization gives w a y . . . to centralized administration ’. 1 Evans-Pritchard, op. cit., p. 37.
C H A P T E R X II P O ST SC R IP T ook represents.an. attem pt. .to „co .m i^ e^ of jppological exam ination.. T h e .fnsfc_is_-a-..synchronic rof N dem bu village structure._T hesecQ ndJsan.experioiachronic nncror:s_o_cio]Qgy._.My„£p.ari^ has Jbeeo. th evilla g e,. m y ,unit_o£^rinae. ,the_social.drama._„I„have tried to m arry th a general to the.particuW „by_analysinga.series o f social dramas in the history o f.a sin g ie v illa g e . Tliat.village's membership_was organized. by. the sjxucturai-piiriciples, isolated dining-the synchronic study o f a num ber o f villages. B u t these principles w ere there interrelated in a um que vvay^ ^In^tiie social dramas .1 have tried to, show h ow in specific.sim ariom .certain principles, came, into conflict, and h ow attem pts.w ergjpiade to maintain the unity o f the ..disturb.ed_gtaup„despite_such„CQnflict. Thtee inain types o f conflict became jeisible..m .the_course_^ the social: dramas : (i) conflict between. prindples^iLorgam zation, receiving behavioural expression .as-xhoice..hetween_conflicting l o y a l t i e s (2) conihct.between_individuds..Qr^liques_for.pQ.wers presage, or w e a lth ;. and,.(3) conflic&-within..m ^ between selfish and-social-drives. Lha've tried to. handle these diflerent-kinds o f. conflict. w ithin a unitary, analytical fram ework not_only b y exam ining social relationships w ithin a convenient local system, but byi-examimng.„change5„and_br.ea.ches_jri those samei relationships.Jcxver_a.„period_Q£ tim e. U have looked for systematic interconnecrions_in_successive_.interactional. events w ithin a-single spatiaLsystem_I.-have.-attempted-to..shQ-W how the unique^.the haphazard, and the-arbitrary.are.subordinated to the^customary--.withim--a_..smgie^-i£.xhanging;,._spaticte.temporal system o f social-relations. I considered it necessary to take a single village as m y universe. My. main .aim..was_tfiLShow how the .general and the particular, the cyclical and_ the ..exceptional, the reglilar and the irregular, the norm al and. the deviant, are interrelated in a single social process. I^could. o n ly . do this by exam ining the vicissitudes in tim e o f a.social. groupjthat wAs in some-sense 4a going concern.’, a .social entity_ wit3x a certain 328
©
Postscript
o o
329
measure o f cohesion m d_contiiiuity,.iow3rds„.wHch its_meinbers felt-stron g . sentiments_..c>f loyalty. '$^ m „.m em bers-o£ a ..social group attach, a high value to its persistence, one is likely to find thaJLTedressive m achinery is.b ro u g h tin to .a ctio n .b yits forem os. membership.. resolve. crises...that .threajten_its-cr.ucial...relationshipst Crises are the overt, .expression .o f any o f the kin.ds. o f conflict just, .m entioned o r o f any com bination o f them __Each, kind o f conflict represents. _.a. challenge...to. . som e..norm,„..governing the behaviour expected between occupants o f .positions in the social structure _o.f the group. Redressive mechanisms, .re-establish norm s?; and jthey.. do so., b y scrutinizing,jdeviant behaviour o f all types. I D eviant behaviour—becomes....structurally, .relevant be h aviour'durin g the redressive phase o f the social drama. For it is against the background o f deviance that conform ity is assessed. T h e jaorm . derives... strength, and definition .condem nation o f itSLebreach in the public, situations_o f ritual and law . The deviantildie haphazardancL thecontingent.can. onIy.be recognized to Jbe^such w here consensus, as. to w hat is .typical, orthodox, regular, exists. A n d . v ic e . v.ersa. It is for this reason that I decided to use a single village as m y unit o f diachronic study, rather than to illustrate the synchronic analysis b y examples taken at random from a num ber o f villages. M ukanza V illage, the spatial system o f social relations chosen fo r detailed study, seemed exceptionally suitable from this point o f view . I have shown from num erical data h o w fissile and subject to fragm entation m ost N dem bu villages are. Ecological conditions, m odem political and econom ic changes, and con flicting principles o f social organization, aU m ilitate ,against the continuous gro w th o f large cohesive.social.units.. „..,Th.os.e..viUages whidbcdiave„managed.tQ,.persist.for.s.everal_g.eneratlons„.acquire great prestige, ..and_.their_.members..attach .a jiig h —value...to^their maintenance. M ukanza, as a long-established village, that had weathered m any crises threatening its unity, afforded an exam ple o f a group capable o f holding together in the teeth o f fissile and centrifugal tendencies o f all kinds. Since the disruptive trends here w ere confronted by pow erful reintegrative mechanisms, the social crises I observed and had described to me in the history o f this village had an intensity w hich threw into re lie f most o f the features o f N dem bu village structure. T he jEno-tto of-th is . hook,. ‘ G eneral Form s -have .their vitality
© m o
© ©
© 0 0
© © !
©
f
©
:
©
0
© © ■
©
i
©
i
®
j
©
© 'i © .
#
330
Schism and C ontinuity in an A frican Society
¿»„Particulars \ w ell summarizes.. tke...metko.
Postscript
331
the_stm cturaL ..form.. ,xj£_die.-ivideiL..Ndemlm-.soc3et}c. R itual attempts to redress crises in villages, but i£schism -is irrem ediable ritual ^restates in w hat are usually em oaonally^chargedckcurristances the highest com m on values o f N dem bu society. A sequel to this hook is in preparation w h ich w ill make N dem bu ritual its central topic. A series o f rituals w ill be analysed w ithin a single held o f social relations. Greater w eight w ill he given to the cultural aspects o f N dem bu ritual than in the present volum e, especially to the sym bolism .
BIBLIOGRAPHY OF WORKS CITED AND CONSULTED A l l a n , W ., Studies in African Land Usage in Northern Rhodesia, Rhodes*
Livingstone Paper No. 15, London : Oxford University Press (1949). r a p n e l l , C. G., LandHolding and Land-Usage among the Plateau Tonga o f Mazabuka District: A Reconnaissance Survey, Rhodes-Livingstone Paper No. 14, Cape Town : Oxford University Press (1948). (Tonga Report.) A r n o t , F. S., Garenganze, or Seven Years i f Pioneer Mission Work in Central Africa, London : Hawkins (1889). ----- B ihé and Garenganze, or Four Years' Further Work and Travel in Central Africa, London : Hawkins (1893). B a r n e s , J. A., Marriage in a Changing Society, Rhodes-Livingstone Paper No. 20, London: Oxford University Press (1951). ------4The Fort Jameson Ngoni *, Seven Tribes o f Central Africa (ed. by Colson, E., and Gluckman, M.), London : Published on behalf o f the Rhodes-Livingstone Institute by Oxford University Press (1951). Reprinted with amended title by Manchester University Press for the Institute for Social Research, University o f Zambia (1968). ------(with M it c h e l l , J. C., and G l u c k m a n , M.), * The Village Headman in British Central Africa Africa , xix. 2 (April 1949). B a u m a n n , H. v o n , Lunda : bei Bauern und Jägern in Inner-Angola, Ergebnisse der Angola-Expedition des Museums für Völkerkunde, Berlin : Würfel-Verlag (1935). B r s l s f o k d , W . V., Fishermen o f the Bangweulu Swamps, Rhodes-Livingstone Paper No. 12, Livingstone (1946). Reprinted by Manchester University Press for the Institute for African Studies, University o f Zambia (1971). B u c h n e r , M., 4Das Reich des Mwata Yaravo und seine Nachbarländer*, Deutsche Geographische Blätter (1883). B u r t o n , R. F ., The Lands o f Cazembe : Lacerda's Last Journey to Cazem be in 1798. Journey o f the Pombeiros, P . J . Baptiste and Amaro José, across Africa from Angola to Tette on the Zambesi, translated by B. A. Beadle, London : published together by John Murray (1873). C ameron, V. L., Across Africa, London : George Philip & Sons (1885). C a m p b e l l , D., Wandering in Central Africa, London : Seeley Service and Co. (1929). ----- In the Heart o f Bantuland, Philadelphia : Lippincott (1922). C a p e t to , H., and I v e n s , R., From Benguella to the Territory o f Yacca, translated by A. Elwes, London : Sampson Low (1832). C h i l d s , G. M., Umbundu Kinship and Character, London : Oxford University Press, for International African Institute (1949). C h in y a m a , T., The Early History o f the Balovale Lunda, African Literature A llant, W ., G l u c k m a n , M., P e t e r s , D. U., and T
Committee o f Northern Rhodesia, Lovedale Press (1945). C o l s o n , E., 4The Plateau Tonga o f Northern Rhodesia *, Seven Tribes o f
Central Africa (ed. by Colson, E,, and Gluckman, M.), London : Published on behalf o f the Rhodes-Livingstone Institute by Oxford 33-2
O
0 333
O
University Press (1951). Reprinted with amended tide by Manchester University Press for foe Institute for Social Research, University o f Zambia (1968). C o l s o n , E., ‘Rain Shrines o f foe Plateau Tonga o f Northern Rhodesia *, Africa, xviii. 3 (October 1948). ------‘ Residence and Village Stability among foe Plateau Tonga *, The Rhodes-Livingstone Journal, xii, London : Oxford University Press
0
Bibliography
t (i95i)-
0 0 ©
o
------* Social Control and Vengeance in Plateau Tonga Society Africa, xxiii. 3 (July 1953)C r a w f o r d , D., Thinking Black, London : Morgan and Scott (1912). C u n Nis o n , I. G . , A Social Study o f a Bantu People (Unpublished Thesis for foe degree o f D.Phil. at Oxford University) (195a). ------* Perpetual Kinship : a political institution o f foe Luapula peoples *, The Rkodes-Livingstone Journal, x x (1956). D ia s d b C a r v a l h o , H. A., Etknographia e Historia Tradicional dos Povos da Luttda : ExpediçSo Portuguesa ao Muatianvua, Lisbon : Imprensa Nacional (1890). E v a n s - P r it c h a r d , E. E., The D ivine Kingship o f the Shilluk o f the N ilotic Sudan, Cambridge : at the University Press (1948). F o r t e s , M., The Dynamics o f Clanship among the Tallensi, London : Oxford
©
----- The W eb o f Kinship among the Tallensi, London : Oxford University
©
University Press (1945).
Press (1949). F o r t e s , M., and E v a n s -P r ït c h a r d , £. E. (editors), African Political Systems, London : Oxford University Press (1940). F r a n k e n b e r g , R., Kinship and Community in a Welsh Border Village (Unpub lished Thesis for the degree o f Ph.D. at the University o f Manchester) (1954). G a n n , L., * The End o f the Slave Trade in British Central Africa \ The RhodesLivingstone Journal, xvi (1954). G l u c k m a n , M-, Rituals o f Rebellion in South-East Africa, Manchester : Man chester University Press (1954). ------ ‘ The Role o f the Sexes in W iko Circumcision Ceremonies \ in Social Structure : Essays Presented to A . R . Radclijfe-Brown (ed. by M. Fortes), Oxford : The Clarendon Press (1940). ----- 4Political Institutions ’, in The Institutions o f Primitive Society, Oxford : Blackwell (1954). ------ (with M it c h e l l , J. C., and B a r n e s , J. A.), ‘ The Village Headman in British Central Africa’, Africa, xix. 2 (April 1949). ------ The Judicial Process among the Barotse o f Northern Rhodesia, Manchester : Manchester University Press (1955). ------Custom and Conflict in Africa, Oxford : Blackwell (1955). ------ * Kinship and Marriage among the Lozi o f Northern Rhodesia“and foe Zulu o f Natal * in African Systems o f Kinship and Marriage (ed. by A . R. Radclifte-Brown and C . D. Forde), London : Oxford Uni versity Press (1950). H a r d in g , C., In Remotest Barotseland, London : Hurst and Blackett (1904).
a*
© G O
© © © © ©
©
© © ©
© © @
O © 0
© ©
©
O
X.
;{r‘r
334
Schism and Continuity in an African Society
H il t o n - S im p s o n , M. R , Land and Peoples o f the Kasai, London : Constable
(1895). S ir H. H., George G renfell and the Congo, London : Hutchinson & Co. (1908). L iv in g s t o n e , D., M issionary Travels and Researches in South A frica , London : John Murray (1857). ------T he Z a m bezi and its Tributaries, London : John Murray (1865).
J
■ O-
:P
o h n sto n ,
M cC
I,
-G ( ;
K.■' f'\ fe
b e"
A.:
f ;; '1 ■t i' ;
ÎJ
p
X ±:
M . , T he Southern Lunda and Related Peoples {Northern Rhodesia, A ngola, Belgian Congo ) (Ethnographic Survey o f Africa), London :
u llo ch ,
Commercial Aid Printing Service Ltd., for the International African Institute (1951). M a g y a r , L a s z l o , Reisen in S iid A frika in den fahren 1849-1857, Pest a n d L e ip z ig (1859). M a r w ic k , M . G., * The Social Context of Cewa Witch Beliefs \ A frica, xxii. 2 and 3 (April and July 1952). M e l l a n d , F. H., In W itchbound A frica, London ; Seeley Service (1923). M it c h e l l , J. C., * The Yao o f Southern Nyasaland \ in Seven Tribes o f Central A frica (ed. by Colson, E., and Gluckman, M.), London : Published on behalf o f the Rhodes-Livingstone Institute by Oxford University Press (1951). Reprinted with amended title by Manchester University Press for the Institute for Social Research, University o f Zambia (1968). ------{with G l u c k m a n , M-, and B a r n es , J. A.), * The Village Headman in British Central Africa \ A frica, xix. 2 (April 1949). ------ (with B a r n e s , J. A.), T he Lamba Village : Report o f a Social Survey. Communications from the School o f African Studies, New Series No. 24, University o f Cape Town (1950). ------‘ A Note on the African Conception o f Causality \ T h e Nyasaland Journal, v. 2 (July 1952). M o n t e ir o , J. J., A ngola and the R iver Congo, London : Macmillan (1875). P o g g e , P ., Itn R eiche des M wata Yamvo, B e rlin : Dietrich Reimer (1 8 8 0 ). P r é v il l e , A. d e , L es Sociétés africaines, Paris : Firmin-Didot (1894). R ic h a r d s , A. L., ‘ Mother-right among the Central Bantu \ Essays Presented to C. G. Seligm an, ed. E. E. Evans-Pritchard, R. Firth, B. Malinowski, and I. Schapera, London : Kegal Paul (1933). ------ Bemba Marriage and Modern Economic Conditions, Rhodes-Livingstone Paper No. 4 (1940). ------ * The Bemba o f North-Eastern Rhodesia *, in Seven Tribes o f Central A frica (ed. by Colson, E., and. Gluckman, M.), London: Published on behalf o f the Rhodes-Livingstone Institute by Oxford University Press (1951). Reprinted with amended title by Manchester University Press for the Institute for Social Research, University o f Zambia (1968). S e r p a P i n t o , A. A. d a R o c h a d e . H ow I Crossed A frica, London : Sampson Low (1881). S in g l e t o n - F is h e r , W ., ‘ Burning the Bush for Game \ A frican Studies, vii. 1 (1948). ------ * Black Magic Feuds *, African Studies, viii. r (1949).
Bibliography St . H
335
G ib b o n s , A., Exploration and H unting in Central A frica, London : Methuen (1898). ------ A frica from South to North through Marotseland, London : John Lane (1904). T il s l e y , G. E. D a n Crawford . M issionary and Pioneer in Central A frica , London : Oliphants (1929). T o r b a y , E. {with J o y c e , T . A.), ‘ O n the Ethnology o f the South-West Congo Free State %Journal o f the R oya l Anthropological Institute (1907). ------ Xjes Pays et les Populations du Bassin du K assai et du Kw ango O rientale . Annales du Musée du Congo Belge, it. 2 (1922}. ------ Cam p and Tram p in A frican W ilds » London : Seeley Service and Co. (* 9 * 3 )• T r a p n e l l , C . G., and C l o t h ie r , J., Ecological Survey o f N orth-W est Rhodesia , Lusaka : Government Printers (1937). T u r n e r , E. L. B. (w ith T u r n e r , V. W .), ‘ Money Economy among the Mwinilunga Ndembu : a Study o f Some Individual Cash Budgets \ T he Rhodes-Livingstone Journal, xviii (1955). T urner, V. W ., ‘ Lunda Rites and Ceremonies’, Occasional Papers o f the Rhodes-Livingstone M useum , New Series, No. 10 {1953}. ;------* The Spatial Separation o f Adjacent Genealogical Generations in Ndembu Village Structure *, A frica, xxv. 2 (April 1955). V erhuepen, E., Baluba et Bdluhaisés dc Katanga , Anvers : L’Avenir Belge (1936). * ' ' W a t s o n , W ., ‘ The Kaonde Village’, T h e Rhodes-Livingstone Journal, xv (1 9 5 4 ). W h i t e , C. M. N., ‘ The Ornithology o f the Kaonde-Lunda Province, Part I. Climate and Ecology *, Ibis (January 1945). ----- - * Witchcraft, Divination and Magic among the Balovale Tribes A frica , xviii. 2 (1948). ------* Material Culture o f the Lunda-Luvale *, Occasional Papers o f the R hodesLivingstone M useum , New Series, No. 3 (1948). ------* Stratification and Modem Changes in an Ancestral Cult A frica, xix. 4 (*949)------* The Balovale Peoples and their Historical Background *, T h e RhodesLivingstone Journal, viii (1949). W il s o n , G., ‘ The Nyakyusa o f South-Western Tanganyika ’, in Seven Tribes o f C entralA frica (ed. by Colson, E.,and Gluckman, M.), London : Published on behalf o f the Rhodes-Livingstone Institute by Oxford University Press (1951). Reprinted with amended title by Manchester University Press for the Institute for Social Research, University o f Zambia (1968). W ilson, M. H., ‘ Witch Beliefs and Social Structure *, Am erican Journal o f Sociology, Ivi. 4 (1951). il l
IN I abolition o f slavery, 64a,, xSo-x, 183,
193-4
accretion,, 197 addphic succession, 87-9,193-200 ; in Mukanza, 99 — group o f male uterine parallel cousins, 68-9, 73 adepts, 293 ; in Chiham ba, 3x1—12 adjacent genealogical generations, 307, 3 3 5 » 337, 240 adjacent siblings, 209 adult males, residence o f 66 affinal connections, 281—6 affines, 235, 237, 240 African P o litica l Systems» 287» 291, 302 age-seniority, 89 agriculture, 20-x Akosa, in., 13 * Allan, W ., on carrying capacity o f land, z x ; on Lamba village size, 39
alliance o f alternate generations, 80-x, 106,167,199 ; in Mukanza, 106-7 — between lineage segments, 109 ambition for headmanship, 99, 218 ancestor cult, 173, 195—6 ; associated with the bush, 173 — shrines, 35,173; Tale and Nderabu, 173 -3 ; see muyomhu Anderson Mulumbi, 160, 164 avunculocal, definition of, xviiin-
ry
if
Bangala, founded by a hunter, 26; village size, 42 Barnes, J. A,, I28n.; on Ngoni village size, 39 Bashi-lange riamba cult, 26 Baumann, H. von, on Mbala and Chokwe hunters, 25—6 ; on Chokwe villages, 42 beer as payment at work-parties, 22 ; for rituals, 1x6, 118, 294 Belgian Congo, x ; trade with, 9 beliefs regarding the dead, 173 Bemba village size, 39 ; dans, 8 6 ; uxoxilocal marriage, 76-7 Bibiana, 107, 277 Biscuit, 89, 217, 230 * black liver 98 33
blindness, XJ9 blood-brotherhood, 314, 2 8 6 ; for Chiham ba, 307 Bottn, 8 boundary o f Northern Rhodesia, s-a ; arbitration, 7 boys, residence o f 53-4, 64-7 breach, o f norms, X13, 115, 120-x» 1*3-4 • o f social relations, 91, X87 ; caused by natural calamity, 138-40, X42 Brelsford, W . V., on Unga village size, 39 British Government, and position, o f chiefs, xtn., 1$ ; and Nyakaseya chiefdom, 3 ; and social change, aix-13, 2x8-20 ; and suppression o f slavery, X92-3 ; and witch craft, 1x4. ; and changes in lineage depth, *29 British South Africa Company, ad ministration o f 7 ; and abolition o f slavery, 64». brother-in-law, conflict between hus band and, 189, 198 brothers, 249-50 ; conflict among, 79 ; separating to found adjacent settlements, 206 ; division of skills between, 250 ; gardens o f 22 brother-sister tie, 78,251-3 ; aku, 235, 251-2 budgets, 23 Campbell, D., on colonial expansion o f Lunda, 5n. candidates, 293 ; for Chihamba, 3103x1 Capello, N., and Ivens, R ., on Luunda capital, 4—5 ; on Jinga hunters, 2$ ; on Chokwe and Bangala villages, 41—2 capitals, 322—3 ; Musutnba, 4—5 carrying capacity o f land, 21, 45-d cash crops, 9, 23 cash economy, 9, 51, 13 S~6 cassava, varieties, 20; and hunting, ad-7 ; ^ d descent and residence, xviii-xix causes o f misfortune, 142-3
33»
Index
census, village, 37 centralized political system o f Luunda, 4 -5
Chawutong’i, 101-2 Chayangoma, 159, 269 ekettda, see work-parties chtbinda, chiyang'a, types ofhunter, 28-9 Chibinda liunga, 25 Chibwakata, 12 , 150; in law-case, 160—1 — Village, history of movements, 45 ; fission, 49 ; marriages in vicinage, 381-4 Chibwika, n , 321-2 chief, meaning o f mwania, 323-4; representative of national unity, 326; character, influencing ap pointment o f 13 ; kin as officials, 322-3, 2x2-13 ; senior wife, Nyakaseya, xx ; lukamt bracelet, 12, 319-21 — officials o f Chwwikankanu , emblem purifier, 12-13, 212-13, 320, 327 ; Ifwota, scout, zz, 318m ; Kaluta , ix, 3xStu ; Kam banji, war-leader, 12, 45, 160 ; Kabuttg'tt, 3 1 8n.; Ntete Mwenimajatnu, 320 chiefdoms, population, 17-18 ; boun daries, 14, 57-8 chiefly villages, 207-8, 322 chiefs* courts, 15-16, 163, 189, 193 Chihatnba, see ritual Chikang’a Village, 211 Chikasa, 164»., 165, 315 Chikimbu, 154, 242 children, residence o f 65 Childs, G. M., on Ovimbundu trade, 4 chtlolu, senior headman, 323—7
Chinema, 100-1, 167, 278 Chinyama, T., 323 Chipendipendi, 178, 195 Chipeng’e, 7, 44, 189, 326 Chiutvikankanu, emblem purifier, see chiefs’ officials choice o f residence, 61—2, 87 Chokwe slave raids, 7 ; hunters, 25m, 26 ; villages, 41-2 Christian Mission in Many Lands, 8, 9 circumcision, see ritual dans, 85-6 classificatory kin, proportion o f in v illag es 7 2 -5
classificatory kin, sisters’ sons’ settle ments, 206-7 ; brothers, rivalry for headmanship, 199-200 ; cross cousin marriage, 199, 255 cleavage between mattilineages, 100, 143- 4, 146-8 ; between matricentric families, 79 ; between factions, xo6 climate, 18-19 Colson, E., on Tonga village size, 39 ; on instability o f Tonga villages, 6l compensation for homicide, X89, 191 compensatory structural devices, 224, 226 composite village with ‘ attached’ lineage, 170, 210-11 concept o f the social drama, 91-4 conflict, theory of, xvii, xxii—xxiii, 8993 ; three main types, 89-90, 328 ; handling of, in Ndembu society, 90 ; pressed into service o f group unity, 127, 129 ; spreading in situations o f crisis, 123, 127—8 — structural, between, principles or norms, xvii, 123-5, 288, 328 ; of principles within and between villages, 287 ; between principles o f unity o f raatricentric family, virilocal marriage, and classifi catory adelphic co-residence, 69 ; and of unity of matricentric family and unity o f matrilineage, 79 ; between familial and lineal principles, 237-8 ; between lineages and between adjacent genealogical generations, 167, 198 ; between matrilineages, 144- 5 ; between mother’s brother and sister’s son, 93 ; between male parallel cousins, 93, 251 ; among brothers, 79 ; between brother and sister over children, 352-3 ; between husband and his wife’s matrilineal group, 260 ; between husband and brother-inlaw for wife, 198, 204 ; between role of father and that of uterine brother, xix ; in father-son rela tionship, 241 ; between co-wivés, 28X — interpersonal, between a husband and wife, 218, 157 ; between a
Index husband and brother-in-law for wife, 19a, 198 ; between a sonin-law and mother-in-law, 240 ; between Mukanza Village mem bers and a tailor, 163 conflict, causes of, over succession, 93— 115 passim; over succession within the matrilineage, 89-91,157,163 ; over the allegiance o f a slave lineage, 181 ; over goats, 132, 163 ; over meat, 32m, 95, 148, 151-2, 251, 308; over beer, 116-18 conjugal status, 279, 281-3 conquests o f Ndembu, 5-6 conservatism o f Ndembu, 40 consumption, 23-5, 31-2 cooking, 24 co-parents-in-law, 273 Copperbelt, 134-6 co-wives, 281 Crawford, D., 7 crisis, following breach, 91, 240 critical density o f population, 21 crops, 20 cross-cousins, 253-5; marriage of,
176» 199, 255
cross-cutting affiliations, xxiii, 273 Crown Land, 20 cult associations, 293-8 ; affiliations, 296-7 cursing, and sorcery beliefs, 118, 120-3 death payments, mpepi, 263-5, 267-74 decentralized political system o f Ndembu, 5, 14-15, 200, 324 demographic data, 37-60 passim dependent *kin group, ntang'a, 234-6 Dias de Carvalho, H. A., on Northern Lunda traditions, 2, 25 ; political system o f Luunda, 4 ; departure o f Kanongesha, from Luunda, 2 ; settlements o f grass huts, 41 District Commissioner, 17-18, 155-6, 182, 212-13 divination, 112-13, 119, 142, 143,180, 299 ; poison test, 271 division o f labour between sexes, 21-3 division o f land after Lunda conquest,
5-6
divorce, ratios, 62 ; and virilocality, 63, 72-3 ; o f sick spouses, 263-5 doctors, 293-4
339
Barth cult o f TaUensi, 174 eating customs, 23-4, 201 ecological factors a ffectin g residential instability, xviii-xLx, 45-6, 173—4 ecology, xg-20 economic co-operation, 33 egalitarianism, 104, 189 elementary family, 10, 24, 2x4—15 emigration of Ndembu from Northern Rhodesia, 7 employment, 136 ; in the pedicle, 8-9 English speaking, advantages o f 103 European centres, 7-9 Evans-Pritchard, E. E., 318, 327 * exceptional persons * within the structure, 184-6, 232 executors, i2pn. exogamy-, 224, 247 expulsion o f sorcerer, 1 19 ; of witch, 148 extended family, 236, 256 ex-wife’s husband, 274 familiars, men’s, itornba, 95-8, m , 122, 180, 245 ; k a toteji , xo8 -— women’s, tuyebela, 144—5, 148-50, 153, 159 ; kahwehwi, 145, 151 family, elementary, 10, 24, 2x4-15 ; matricentric, xix, 68, 75-6, 79, 84-5, 222-3 ; extended, 236, 238, 256 farms, 10, 42, 133-5 * two types o f 34 ; spatial arrangement, 36 ; distribution o f 36-7; mag nitude, 38 ; farm heads’ occupa tions, 36 ; and parish system, 41 ; contiguous with villages, 50,220 ; superceding villages, 2x7; Sandombu’s form, 133 father-son relationship, 241 ; in hunt ing, 29 ----- daughter relationship, 242 father’s sister, 237, 243 feuds and village movement, 46 fights, x6o, 181, 182 finger -millet, 20-2 ; planting after village move, 47 Fisher, Dr. W ., 8,105 Fisher, ffolliot, 8 fission o f villages, 48-51, 87-8; defined, 169; various types o f 169-70, 204-7 I between lineage segments, 169 ; increase in rate o f
5o-ï ; considered a calamity, 177 ; renewed relations after, 176-8, 184» * 9 7 fission of Mukanza Village, 178-82 ; o f Kafumbu Village, 182 ; of Yimbwendi Village, 182 ; further possibility o f in Mukanza Village, ïo i , 164-7, 3<>7 ; o f Shika Village, 207-10 ; o f Nsang’anyi Village and offshoots, 210-18 ; o f Chibwakata Village, 49 ; o f Kawiku villages, 49-50 following, 15,134 ; an embarrassment to cash earners, 135 ; Sandombu’s» 99 »153 - 4 »*64; Kasonda’s, 164-5 ; Yimbwendi’s, 181, 188 food, 23-4, 31-2 Forest Reserve, 20, 216, 285 form and process, xviii Fortes, M., on concept of social system, xrii ; on Tallensi, 82, 84, 170-2 fragmentation o f villages, 45, 169 Frankenberg, R ., on role o f the stranger, 147 funeral rites, 1x9, 267—71 ; and village movement, 46 game, 19,29 ; control, 19-20 ; drives, 29-30 ; resources as factor in village size, 43, and movement, 46 Gann, L., on Angolan slave trade, 7 ; on introduction o f guns, 28 gardens, 20-1 ; the social organiza tion o f gardening, 21-3 genealogical recall, 82-4 genealogies, 37, 69-70 generation separation, 73-4, 236-7, 245 ; principle of, 175-6 ; effect within the village, 175-6, 199 ; and the extended family, 236-7 geographical location of Ndembu, x ghosts, nyisaîu, X41, 143, 145 Gideon, 148, X53, 165, 265 girls* initiation, see ritual, girls* puberty girls, residence o f 53-4, 64-5 Gluckman, M., on cross-cutring alli ances, xarii-xxiii, 288 ; on role o f sexes, 298 ; on Lozi judges, 122 ; on divorce in patrilineal societies, 223 ; on exclusion from ritual, 2970. goats, disputes over, 132, 163 Government T a x Register, 74
grandparent-grandchild relationship, 80, 244-6 ; marriage, intravillage, 79-80, X67 ; marriage, inter-village, 176 grass huts, hunters*, 29 ; early dwel lings, 35, 41-2 graves, 173 graveyards, ownership o f 2x7 greetings, 326 growth o f small villages, 42-3 grudges, factors in sorcery/witch craft, 527 guns, introduction o f 28m Harding, C., on Lunda ‘ fortress *, 41 headmen, and their kin, 70-6 ; per sonality o f 52, 84, 200-2, 217; privileges o f 129a. ; longest established headman, mwettimbu, 217, 285, 325; polygyny o f 279-83 ; as hunters, 32, 202-3 ; situation after death o f 178 hereditary titles of villages, 104-5 Millwood Farm, 8-9 historical development o f Ndembu villages, 43-4 history and traditions, 2-18, 212-13, 318-27 hoe agriculture, 20, 22 hospital, 8, Ï4X-2 hospitality, of headman, 200-2, 281 ; between members o f hunters’ cult, 29 ; between Lwena dans, 85-6 ; Sandombu’s, 99, 136 Humbtî, 4 hunters, training o f 29 ; aristocratic rank o f 25-6 ; and headmanship, 32, 202-3 Î ideal personality o f 202 ; regarded as sorcerers, 32 hunting, a masculine occupation, 25-8 ; social organization o f 28-31 ; and the conflict between the sexes, 27-8 ; and descent and residence, xviii, 27-8 ; and span o f noatrilineages, 227 ; and village size, 42, 173
htmtsmanship, cults o f see ritual husband, and wifb-beating, 118, 120 ; as witchcraft familiar, 145 ; in conflict with brother-in-law for wife, 192, 198 husband and wife, co-operation in work-parties, 22
© Index hut building, 36 hut forms, 35 hut owner, definition of, 70 Ifw ota, see chiefs’ officials ihaku, see law-men
Ikayi, 132,154 Ikelenge, Chief, 7» 11» 13 « . J visits Mwantiyanvwa, 3a. — Area, 8-9 ; many farms in, 37 — capital, 9 Ikubi, death of, 142, 148, 269 ilornba, see familiars « immigration from Angola, 8-9 incest, 252 ; as index o f village fission, in Nswanakudya, 86, 208-9, in Mukanza, 181, 195-6 increase in population, 50-1 indigenous social system, survival o f 16-17 Indirect Rule, 11 individualism o f Ndembu, 22-3, 30, 101, 325 individual mobility, through villages, 51-6 ; through chiefiioms, 56-8 inheritance, 129m in-laws, avoidance of, 235 ; o f an adjacent genealogical generation, to e g o , aku, 235-43, 273 ; o f e g o ’ s own genealogical genera tion, mashaku, 235 installation o f chief, 319 intercalary role after fission, xSx-2, 184-6 intermarriage, x09-10 ; between Mu kanza and its seceding groups, 184 intrigue in Mukanza, 131-3 Ishinde, chiefdom o f in, ; departure from Luunda, 2 Itota, 286, 311 ivum u, 248-9 ; see matrilineages Jinga hunters, 25 John Kambanji, secedes, 214 ; sorcery o f 2x4 joking relationships, 245-6, 248, 253 judicial institutions, 15—16 jural mechanisms and ritual, 122-8 Kabompo villages, 41 Kabonzu, founding ancestor, 105, 195
34*
Kafumbu secedes, 178-82 ; personality o f 180 ; a hunter, 203 ; named as a sorcerer, x8o, 183 — Village, composition of, 198 Kafwana, 12, 320 Kahali Chandenda, in. Ikelenge Area, 105 ; returns to Mukanza, 105, 180, 188 ; supporters o f 180 ; defects as headman, ixx, 184 ; death o f 95-8 Kahumpu, 190-1, 263, 273 kahwehwi, see familiars Kajata, 100 ; death o f 138 Kakoma, Chief 12 K alula, see chiefs’ officials Kalusa, 101, 150 ; death o f 267-72 Kamawu, 2x7, 230 Kam banji, see chiefs’ officials Kangombe, 7 Kanongesha, senior chief of the Ndembu, xx-xxi, xx, 15 — history o f : migrations, 2, 5-6 ; apportionment o f land, 6 ; history o f chieftainship, 321-2 ; succes sion, 321—3 ; two rival incum bents, 1—2, 14, 318 ; his twelve senior headmen, 5-6, X2,183 — functions of : judicial, 15-16 ; ritual, 16, 318—21 ; office repre senting tribal rights to land, 319 — chiefs : Nkuba, 320 ; Kabanda, 320-1 ; Nkomesha, 326 ; Ishima Watuta Menji, 321-2 ; Mulumbi, 14 ; SakayoU Chifuwu, 14 Kanyombu, death of, 138 ; vengeance medicine o f 141, 144 Kaonde, 3 ; village size, 39 ; clans, 86 Kasonda, in succession to headmanship, 102 ; ambitions, 103, 276 ; abilities, 103 ; builds a brick house, 138 ; plans to found farm, 155 » 164—5 ; rivalry with Sandombu,"102-3 ; defends Mukanza Village, 150 ; favours the name Mukanza, x56 ; accusedofbewitching Kanyombu, 138-42 ; handles sorcery accusations, 145-6 ; his dependent kin group, 234 ; mari tal ties, 276 ; ties with Shika Village, 265-6, 273 Katendi, X84, 198, 315 Katiki, X33, 157 kavum bi, self-respect, 201
0
© O ©
©
© © © © o © 0
o © ©
© 0 ©
© €f
G
© © ©
©
C G
©
© ©
©
ip
r f
S3 O •ô •p.
I'.';
D
m
X) X.)
I) e
X :o X) Kj
m {, p x,
X %>
15
34 2,
In d e x
Kawiku, i i, 18, 218 ; the name, 3-4 ; origins in Nsang’anyi, 210 ; vil lage mobility, 45 ; village fission, 49-50, 183 ; and. JSfdembu, 4, 160,163,273 ; intermarriage with Ndembu, 262 Kayineha, 181 ; his friendship with Mukanza, 184-6, 232 Kazem.be Mutanda, Chief 3a. ; de parture from Luunda, 2 ; chief* dom of, in. Kimberley-brick houses, 10, 35-6, 45, 133 , 1 3 « kin, categories of, 70-2 ; absence of, 108, 152, 199 kingship, Shilluk, 3x8 kinship and affinal connections between headmen in vicinage, 279-87 kinsman, kawusoku, 234-5 ku-swanika ijitia, see ritual, succession to a name labour migration, 17 Lamba village size, 39-40 ; village structure, 61 ; clans, 86 ; crosscousin marriage, 255 land, utilization, 20-1, 46 ; allocation, 217 law-cases, village : Nyamukola j/. Sandombu (slander), 157-61 ; Sandombu v. Chibwakata (slander), 160 ; Mukanza Kabinda v. Shika (death payment), 270-2 — chief’s court : Anderson Mulumbi v . Mukanza Kabinda (loss of goat), 163 law-men, tnakaku, 16, 159 ; Kasonda, 103 ; Kahali Chandenda, i n ; Chayangoma, 159,269 ; Ng’ombi, 201, 271 ; o f vicinage, 159-60, 270-1 ; of senior headmen, 324 law, skill in, 103, 201 life crisis rituals, 292 Line, 136 ; makes a farm, 136-8 ; marital ties, 186, 278 lineages, xix, 82-6 ; unstable struc ture of, 79, 174-6, 198-200 ; and individual mobility, 176-7 ; compared - with Tale lineages, 170-3, 222 — depth of, 82-3, 172 ; factors limiting depth, 227-30 ; and village size. 82-3
lineages, attached, 170, 210-11, 222 lineal and familial principles in con flict, 237, 247 literacy, 103 Livingstone, E>., on limited power of Lunda chiefs, 5 ; on stockaded villages, 41 local autonomy o f villages, 6 local matrilineage as an ideal goal, xx, 84. 199 long-established and recently estab lished villages, 74-6 ; and lineage depth, 82-4 Luba hunters, 25 Luchazi village size, 39 lukanu bracelet of chieftainship, 319321 ; theft of, 14 Lukolwe, 6 Lunda, see Southern Lunda Lunda of Luapula, clans, 86 Luunda, the name, 2n .; centralized political system, 4—5 ; Lunda origins in, 2—3 ; clans of Luunda peoples, 85 Lwena, chiefs, 5-6 ; village size, 39 ; clans, 84-5 McCulloch, M., 2n. maize, 20, 35 Makumela, 216-18 Malinowski, B. M., on the ‘ mytho logical charter 98-9 Manyosa, 101-2 ; succeeds to Nyamwaha’s name, 1151-20; accuses maiabu lineage, 141 ; mourns Kalusa, 267; marriage, 164«., 165 ; marital ties, 274-5, 277 marital status, see conjugal status marriage, spatial range o f 275, 280; between chiefdoms, 260 ; brittle ness o f 62, 78 ; payments, 265—7 ; service for in-laws, 23 ; tradition o f marriage* between villages, 261-3 masked dancer, iktshi> 157 Matempa secedes, 212-13 ; fission in Matempa village, 216-17 matriceritric family, xix, 68, 75-6, 82, 84 - 5 , 222-3 ; as unit of secession, 79
matrilineages, major, minor and mini mal, definition of, Son.-, as units of secession, 205-6
Index matrilineal kin, akwamama, 236 ; ex tending beyond village, 86-7 matriliny, and village residence, 6476 ; linking villages in different
vicinages, 272—3
matrilocal, definition of, xviiin., ssn. Mbala hunters, 25 Mbwela, 3, 6, 319 meat, distribution and consumption, 31-2 ; drying of, 29 ; division of, 30, 31-2 ; conflicts over, 32m, 95, 148, 15 1 -2 , 308 medicines, 294 men and women, different roles in agriculture, 24-5 ; different pat terns of mobility, 52-6 menstruation and huntsmanship, 27 Messengers, 208 minor lineage, 195, 206 minors, residence of, 65 missions, 8-9, 140 Mitchell, J. C., 77 ; and Bames, on Lamba village structure, 61 mobility, of villages, xx, 6, 44-8 : o f individuals, 51-8 model, of village residence, 63-4 ; of cult affiliations in a vicinage, 296-7 mode o f affliction, 293-6 moral community of males, 243—4 mother-child bond, influence on resi dence, 62—3 ; and effects o f generation separation, 67 -----daughter relationship, 243 ; -«on relationship, 65—7, 207 — witchcraft of, 243 mother-in-law, avoidance of, 235 mother’s brother, 237, 241-2 ; duty towards, 93 movement o f a village, precipitates fission, x8o-i tnpepi, see death payments Mukanda, see ritual Mukang’ala, Chief, n „ I3n., 98, 115, 211-12 ; withdrawal o f recogni tion from, 12 ; Mukang’ala ibala, 44 ; chiefdom, 10-11,17-18 Mukanza Kabinda, succeeds to the name o f Mukanza, 106 ; made headman, 98, 115 ; divorces Nyatungejx, 265 ; pivotal posi tion, xoz ; defects as headman, X03 ; illness o f 138-42 ; infirm state, 154-5 ; appoints a successor.
343
XS5 ; marriages, 109-10, 263; marital ties, within Mukanza Village, 239 ; longest established headman in vicinage, 217 Mukanza Kandulu, 105-6 ; death by sorcery, 180 — Village, chosen for study, 94 ; movements, 45 ; history of, 105-7 ; sources of disruption in, 239 ; marriages of members, 262-7, 274-9 ; connections with vicinage, 279-86 muku, see in-laws murder, 191, 240 Musokantanda, Chief in., 3n. ; de parture from Ltxunda, 2 ; chief dom o f 2, 6 M usolu , see ritual Mutanda School, 114-15 muyombu, ancestor shrine, 119, 126-7, X73 . 195 » 2.92 Mwanawuta, the tide, 13m ; Ikatu, secedes, 211-12, 216 — village, nuclear group o f 213 ; fission in, 213-16 Mwantiyanvwa, Chief, 2-3 land of, 2n. ; capital, 4-5 ; and origins o f Lunda, 2 ; tribute to, 6 ; Ianvo Noeji, 2 ; Mbaka, 3m mwenimbu, see headmen, longest estab lished Mwinilunga District, x Mwininyilamba, Chief 3m, 11, 13 ; see also chiefs’ officials, Ifwota mythological charter, 98-9 namesakes, tnajinda, 187 national unity, preserved by ritual, 289 : represented by Mwantiyamwa, 317 Native Authority, 14, 15-16 Native Trust Land, 20 natural misfortune, and ritual, X28 ; and projection o f responsibility on to witches and sorcerers, 109, 126, 128, 152 Nayar o f Malabar, 68n. Ndeleki, 101—2 Ndembu, the name, 1 ; region, loca tion o f 3 ; classification under administration, 20 ; migrations, 5n. ; and Tale social organization compared, 170-7
344
Index
necrophagy, 3211., I$1-2 neighbourhoods, see vicinages N g’ombi, 3,00-2, 270-1 ; Village, links in vicinage, 281-4 Ngoni village size, 39 iSTkang’a, see ritual N kula, see ritual Nsang’anyi, office o f Chivwikankanu , xi-12, 212-13 ; head o f the Kawiku, 153, 191, 2io, 218 ; applies for office o f chief, 213 ; character of, 218 ; law-man in Kahumpu case, 191 — Village, attached lineage in, 2x0-11 ; fission, in, 210-X8 nsewu, see marriage payments Nswanakudya, authority in vicinage, 217; law-man, 160, 271; * older brother * o f Shika, 271 — Village, founding of, 208-10 ; links with vicinage, 282-4. Nswanamatempa, succeeds, 216 ; a law-man, 270; sorcery reputa tion of, 217 — Village, fission in, 2x6-17 Ntambtt Lukoxxkesha, Chief, 12, 13, 276 ttteng'a, see dependent kin group
Ntololu, accuses Nyamuwang’a o f witchcraft, 148 ; keeps customs o f widowhood, 270 ; in dispute over death payment, 271 numerical analysis, xvh-xviii; data for, 37 Nyakaseya, Chief, 3n„ xi ; attempt to abolish chieftainship o f 3 Nyakatolu, 6 , 7 Nyakyusa, village size, 39 Nyaluhana Village, history o f move ments, 44-5 ; its vicinage, 47-8 Nyaluwema, a link between Kafumbu and Mukanza, 184; a ritual adept, 295-6, 314-15 Nyamukola, intercalary role in Mukanza Village, 100,167,307-9; marriage of no ; accuses Kasonda, 140, X44; in situation o f strain preceding Chihamba ritual, 307-9 Nyamuwang’a, origins, 150-1 ; divorced from William Matorokoshi, JJ3, 266-7 i a diligent gardener, 151 ; quarrel with
Chinema, joi ; accused o f witch craft, 148-54» 191-7, 269 ; ex pelled from Mukanza Village, 148 ; executrix at Kaluga’s death, 270 ; marital ties, 278 Nyarioli, 116, 144—5, 300 Nyatungeji’s divorce, 263-5, 272 Nyawatwa, 208 Nyawunyumbi's children in important role, 186, 197; marital ties, 278 oath, form o f 252 older and younger siblings, 348-9; older brother, 249-50 old women, 152 Ovimbundu, founded by a hunter, 25 ; slave traders, 5, 7 , 13m, 327 ; and introduction o f guns, 28n. parallel cousins, 250-x patients, 293-4 patriiateral kin, 71-2, 75 patrilineal, kin, akuratata, 236 ; succes sion o f chiefs, 2o8n., 321 patriiocal, definition o f xviiin., 550.; children, xix-xx ; residence, 6X-3,
65
peace, period o f 6 Pearson, death o f 187 pedicle area, 8-9 perpetual relationship, 13 personality, influence o f within social structure, 94 ; Ndeleki, 101—12 ; Kayineha, 184-6 ; Kamawu and Biscuit, 230 ■— factors in choice of headman, 200-3 poison oracle, x6, 189, 27X, 327 pole-and-mud huts, 35 political system, 14—16 polygyny, 279-83
population, of chiefdoms, 17; per settlement, 38 ; increases in
pedicle, 9 Portuguese, Luunda trade with, 5 ; rebellion under, 9 prayers, at ancestor shrines, 1x9-20, 126-7, 292 ; before initiation, 112 Priville, A. de, on cassava-growing by hunters, 26 primary and dassificatory kin, 63 primary sisters* sons’ setdements, 206-7
If @ Index principles
of
village
organization,
339-40
processional form o f social drama, xvii, 91 projection of responsibility for natural misfortune on to witches and sorcerers, 126, 128 puberty ritual, girls’, see ritual Public Works Department, 132 putative maternal descent links, 172, X75 -
X9S-6
©
345
ritual, funeral, C hipenjtt 129^,269,271; Mudyileji, 119, 270 ; M ung'ong'i, 298 ; Chiw ilat 298 — other : circumcision, Mukanda, 27, X73 . X87, 243-4 ; girls’ puberty. Nkang’a, 53, 245, 266, 292 ; *mentioning the womb ku-tena ivtttnu, 54 ; rain-making, Musolu, 16, 298, 327; huntsmanship, Wuyang’a, VVubinda, 28a., 30, 298 ; Nkula, 116, 293-5 twin ceremony, Wubwang'u, 143, 300 ; Isoma, 299 ; cleansing from incest, ku-disola, 181, 196, 208 roads, 36, 211 road-work, 132 Sailunga, in., 13 Sakazao, and succession to headmanship, xoo—2, 137-8 ; appointed successor, IS5 ; character of, X3 <5- 7 . i 4 s ; marital ties, 277 Sakeji School, 8 Sakutoha purchases freedom, 192 Saluvaji, 197-8 Samlozang’a, x82 Samunuma, secession of, 182, 198 Sandombu, settles in Mukanza Village, xo6; bewitches Kahali Chandenda, 95-8 ; is excluded from headmanship, 98, X01-9 ; sorcery, 95ff. ; character of, 99 ; ambition 99 ; lack of kin, 107-8 ; sterility, 107, 109-xo ; marriage to Zuliyana, xxo, 116, 120, 123 ; be witches Nyamwaha, 118 ; is expelled, 119 ,12 1; returns, 1x9xao ; period o f intrigue, 132-3 ; in dispute over goat, 132; in employment, 132; builds brick house, 133 ; reckoned innocent o f Kanyombu’s death, 146 ; gener osity of, 114-15, 146-7 ; invites Nyamuwang'a to his farm, 148150, 153—4 ; acquires following, 153-4 t succeeds to the name o f Kahali, 155-6 ; accuses Zuliyana and Nyamukola o f witchcraft, 159 ; in law-case, 159-61 ; im proved position, 164 ; his manip ulation. o f principles of social affiliation, 230-x ; defends Nyamuwang’a from the imputation o f
©
© @
© 0
© 0
©
©
0
©
©
© © ©
© © © © © © O
©
G
©
© © G
Index
346 witchcraft,
269;
marital ties,
276-7
scapegoats, witches and sorcerers as, 109, 1x5, 151-2
scattering o f kin, 174
schools, 8-9, 114 seceding groups, lineage span, 205 ; leaders of, 205 secession, reasons given for, 177; instability o f groups after, 182 ; of a hunter, 32 —- units of, 169-70, 221-2 ; mamlineage, 100,178-80 ; matricentric family, 79 ; uterine sibling group, 85 ; elementary family, xo, 214-15 senior adjacent generation, challenge to, 1x3-14 senior chiefs, 11 senior headmen, in tradition, 5 -7 ; as Native Authority Sub-Chiefs, 11—16 ; ayilolu, 14-15, 323-7 ; insignia, 13m, 324 ; courts, 324 senior women in seceding groups, 205 separation of adjacent generations, xx ; spatial, 36, 73-4; and village unity, 67 ; in Mukanza, 99 settlements, size of, 37-8 sewing machines, 36, 135 sexual excess, wuvumbi, 151 Shika Village, fission in,. 207-xo ; homicide at, 190-1 ; headman in law-case, 270 ; intermarried with Mukanza Village, 152-3, 262-7 siblings, 25X-3 ; headman’s, propor tion of in village, 73 ; by the same mother, 67 ; junior, paid as slaves, 189 ; group of, 248-9 Singleton-Fisher, W., on game drives,
30
sister’s son, duty to unde, 1x3 ; hiving off, 87-9, 204-5 — child, paid as slave, 189 Six secedes, 273 »275 slave raids, and village size, 40-1, 43-4* 220, 228-9 slavery, and social change, 187-97 — abolition of, 640-, 180-x, 183, 192-4 ; and village conflicts, 183, x88, 193-4 slaves, 70 ; payment of, 189 ; duties of, 190; rights of, 193-3 ; assimilation into kinship system, 189-90; status of inherited
matnlineally, .190 ; large lineages of, 190 ; may purchase freedom, 19°» 192 ; in a seceding group, 170, 178-82, 1 9 3 -4 ; interlinking lineages by marriage, 199 ; pun ishment of, 193 slave trade, 7, 327 social change, 17, 24, 133-6, 218-21 social composition, of vicinages, 47, 172; of settlements, 71 ; of longestablished and recently estab lished villages, 75 ; of a typical recently established village, 63-4 ; o f seceding groups, see secession social drama, concept of, xvii, 91-3, 161, 230-2 ; as tool of analysis, xvii, 92—3, 230; in conjunction with numerical material, xviixviii, 92-3, 23 x ; as index or vehicle of change, 162; and interim periods, 131 socially procreative link, brothersister tie, 78 social organization of gardening, 21—3 ; o f hunting, 28-31 social system, concept o f as held of tension, rarii soils, 19 Sondash, secedes, 2x4, 225 sorcerers, hunters as, 32 ; persons of mixed tribal origin as, 150 — Sandombu, 1x1-15, 118-19; Kasonda, 138-42 ; Kahali Chandenda, 95, ixx, 184; Kafumbu, 180, 183 ; Yimbwendi, 181, 183, 198 ; Nswanamatempa, 217 sorcery, disqualifies persons from headmanship,98,115 ; accusations after natural misfortune, 125-6, 142-3 ; and grudges, 127 ; and commensality, 201 ; when regarded as justifiable, 143 ; per formed by hired specialist, 1x2—13 Southern Lunda, in., s-6 ; chiefs, 3 spirits, ancestor, akishi, 146 ; affliction by, 119, 140 - 3. -4 6 , 293-6, 309 stabilizing factors in villages, 45 starting a village, 87 sterility, 107-8, xx6 stores, trading, 8 strangers, categories o f, 70; propor tion in villages, 72, 75 ; villages of, 324-5 ; role of, 147
Index stream gardens, 20-2 structural amnesia, 84, 176, 295 structural principles governing resi dence, 226-7 Sub-Chiefs, 11—16 subsistence, mode of, 32-3 succession, to headmanship, 87-90; rule of, xo3-»4; history of, in Mukanza Village, 104—7 ; struggles, 93-1x5 passim ; to chieftainship, 321 ; see also adelphic succession — to a name, see ritual Swazi ritual, xxi symbols o f masculinity, 27 synchronic and diachronic analysts, 328 tailors, 36, 43, 135 Tallensi, xx-xxi, 84, 170-4, 177 taxation, introduction of, 7 temperature, 18-19 temptation to rebel, X23-4 * ten taxpayers rule ’, 41 Tonga, village size, 39 ; clans, 86 topography o f village, 35-7 Torday, B., on Bashi-lange, pakassero and Chokwe hunters, 26 traders, 9, 36, 43, 135 traditional'Villages, 10 Trapnell, C. G-, and Clothier, J., 18-19 travellers’ use o f kinship ties, 176 tree felling, 20-2 tribal land and international bound aries, 1-2 tribute, to chiefs, 6, 15, 212, 325 ; to headmen, X29n. ; to Dr. Fisher, 8 Turner, V. W-, on Lunda rites, 2950. tuyehela, see familiars
347
uxorilocal phase o f marriage, agricul tural duties in, 23 ; modifying residential pattern, 62 values of Ndembu society, xxi vegetation, 19 vengeance medicine, 141, 145 Verhulpen, 35., on date o f lanvo Noeji, 2 ; on slave trade, X6-X7 vicinages, composition of, xx, 44-8, 206-7; linked by their com ponent villages, 176-7 — o f Nyaluhana, 47-8 ; o f Mukanza, 279-86 village, topography, 35-7 ; registered and unregistered, 37-8 ; distribu tion, 47 — magnitude, 37—44; and village structure, 256-7; and hunting, 42,173 ; and lineage depth, 82-5 ; and slave raids, 40-4 — mobility, 6, 44-8; history o f movements o f Nyaluhana Village, 44-5 ; movements o f Chibwakata Village, 45 ; movements o f Mukanza Village, 45 — affiliation, 64 ; typical residential pattern, 63-4 — continuity, 238, 261 ; recently established and long-established, 74“ 5
— o f the village, 145, 147
— fragmentation, 45, 169 — amalgamation, 182 — see also fission villages, nyikala, 34 ; in Central Africa, 61 ; traditional, 10; ancient titles, 84, S04-5, 183 virilocal, definition of, xviiin. — marriage, xviii-xx, 256-9 ; and matriliny, 222-7; and hunting and cassava-growing, xviii, 28 ; a cause o f conflict, x v iii; and adelphic relations, 68-9, 85
uterine sibling group, 85, 198-200, 249 ; definition of, 204 ; as unit of secession, 85, 205-6 ; factors in persistence of, 224-6 — brothers, as core of village, xix, 68 — kin, and inheritance, X29n. uxorilocal, definition of, xviiin.; mar riage, and lack o f co-operation, among men, 76-7
wage-earners, 133-6 Watson, W., on Kaonde village size, 39 ; on Kaonde village structure, 61 "White, C. M. N., on history of Southern Lunda, 3 ; on SubChiefs, 1 in .; onjudicial functions
Unga village size, 39 unity o f all Ndembu, in ritual, 173, 288-93, 301-3 ; in Chief Mwantiyanvwa, 317
Index
348
o f chiefs u n d er N a tiv e A u th o rity , 15—16 ; o n size o f v illag es, 40 ; o n ckilolu, 323m ; o n kavutnbi,
work-parties, yenda, 22, 157
20m.
Yana, ritual history of, 299-300 ; marriage of, 277 Yao, 76 Yimbwendi, ambition to be headman, x8o, x88 ; sorcery of, x8i, 183, 198 ; a hunter, 203 ; secedes, 178-82 ; strained relations with Mukanza Village, 187
* whiteness *, 107, 120
widows’ residence, 62
William Matorokoshi, 153, 266-7 Wilson, G., on Nyakyusa village size, 39
Wilson, M. H., on witches, 32m witchcraft, accusations, 102, 159 ; o f women, 14 4 -5 l of Nyamuwang’a, 148-53 ; see also familiars wives, duty to prepare food, 24 women, struggle for control over, xix ; continuity of village through. 238
Zuliyana, Sandombu’s jealousy over, n o ; childlessness of, 109 ; beaten by Sandombu, n 6 -i8 , X20 ; accused of witchcraft, 157-9
Î1
o
0 ©
© O
© © © ■ © ©.
©
© © © © O
© ©
©
©
O
© íl ©
© © ©
