1_MEYER FORTES unilineal descent.pdf

8/11/2019 1_MEYER FORTES unilineal descent.pdf

1/25

THE STRUCTURE F UNILINEALESCENT R O U P S ~

y MEYER FORTES

S IS well known, Africa has loomed large in Rritishfield research in the

A past twenty-five years. It is, indeed, largely due to the impact of ethno-graphic da ta from Africa that British anthropologists are now giving so muchattention to social organization, in the widest sense of t hat term. I n thispaper what I shall try to do is to sum up some positive contributions thatseem to me to have come out of the st udy of African social organization. Iwant t o ad d this. British anthropologists are well aware tha t their range ofinterests seems narrow in comparison with the wide and adventurous sweepof American anthropology. This has been due to no small extent to lack ofnumbers and there are signs that a change is on the way with the increase in

the number of professional anthropologists since the end of the war. At thesame time, I believe that the loss in diversity is amply balanced by the gainswe have derived from concentration on a limited set of problems?

Social anthropology has undoubtedly made great progress in the pasttwenty years. I would give pride of place to the accumulation of ethnographicdata obtained by trained observers. It means, curiously enough, that there isgoing to be more scope than ever for the armchair scholar in framing andtesting hypotheses with the help of reliable and detailed information. ForAfrica the advance from the stage of primitive anecdotage to that of scientific

description has been almost spectacular; and most of it has taken place since1930, as can be judged by comparing what we know today with the state ofAfrican ethnography as described by Dr. Edwin Smith in 1935. Mainlythrough Malinowskis influence we now have a respectable series of descrip-tive monographs on specific institutional complexes in particular African so-cieties. Studies like Evans-Pritchards on Zande witchcraft 1937), Schaperason Tswana law 1937) and Richards on Bemba economy 1939), to cite onlythree outstanding prewar examples, typify the advance made since 1930. Theyare significant not only for their wealth of carefully documented detail but

also for the evidence they give of the validity of the thesis, now so common-place, th at the customs and insti tut ions of a people can only be properlyunderstood in relation to one another and to the culture as a whole. Theyshow also what a powerful method of ethnographic discovery intensive field-work on functionalist lines can be.

The field-work of the past two decades has brought into clearer focus thecharacteristics of African socieites which distinguish them from the classicalsimple societies of, say, Australia, Melanesia or North America; and themark of this is easily seen in the thought and interests of Africanists. One of

these is the relatively great size, in terms both of territorial spread and ofnumbers, of many ethnographic units in Africa as compared with the classical

17


2/25

18 AMERICAN ANTHROPOLOGIST 155, 1953

simple societies. There are few truly isolated societies in Africa. Communica-tion takes place over wide geographical regions; and movements of groupsover long stretches of time, exactly like those that are known from our ownhistory, have spread languages, beliefs, customs, craft and food producingtechniques, and the network of trade and government, over large areas withbig populations, A tribe of ten thousand Tswana, two hundred thousandBemba or half a million Ashanti cannot run their social life on exactly thesame pattern as an Australian horde, which is, after all, basically a domesticgroup. In Africa one comes up against economics where in Australia or partsof North America one meets only housekeeping; one is confronted with govern-ment where in societies of smaller scale one meets social control; with or-ganized warfare, with complex legal institutions, with elaborate forms ofpublic worship and systems of belief comparable to the philosophical andtheological systems of literate civilizations. Even before its subjugation byEurope, Africa boasted big and wealthy towns. Certainly there was knowledgeof all this before professional anthropologists began to work in Africa. But i twas patchy and on the whole superficial. In particular, it lacked the explicitconceptualization and integral presentation tha t mark the kind of monographI have mentioned. That a belief in witchcraft occurred in many African cul-tures was known long ago. But the precise nature of the belief, and how itwas related to the notion of causation, the rules of moral conduct, the practiceof divination and the art of healing to form with them a coherent ideology fordaily living, was not understood till Evans-Pritchards book appeared. It wasknown, from the works of nineteenth century travellers and administrators,tha t many African societies had forms of government similar to what politicalphilosophers call the State. But there was little or no accurate informationabout the constitutional laws, the structure of administration, the machineryof justice, the sanctions of rank, the getting and spending of public revenues,and so forth, in any African state before Rattrays important studies inAshanti in the twenties Rat tray, 1929 and later). Rattrays description ofAfrican st ate structure has now been superseded. We have a pretty good ideaof how a monarchy was kept in power not only by ritual constraints andprerogatives, as in the case of the Divine Kingship of the Shilluk see Evans-Pritchard, 1948) but also by means of shrewd secular sanctions and institu-tions such as the control of public revenues and armed forces in Dahomey,described by Herskovits 1938); or the manipulation of a rank and classbased administration as in Nupe Nadel, 1942); or by means of both ri tualand secular institutions as has been so vividly described for the Swazi by Dr.Hilda Kuper 1947).

Of course, African customs and institutions often have significant resem-blances to those of the simpler peoples of other continents. Indeed it is justthese resemblances that make the distinctive features of African ethnology


3/25

FORTES] STRUCTURE OF U N I L I N E A L DESCENT GROUPS 19

stand out in proper theoretical perspective. Take the customs of avoidancebetween affines or between successive generations, known from many partsof the world. We are ap t to think of them, even with reference to such charac-teristically African cultures as those of the Southern Bantu cf. Hunter , 1936)as expressing specific interpersonal relationships. It is the more striking tofind among the Nyakyusa Wilson, 1951) that the whole scheme of localorganization in age villages turns on such avoidances. Moreover we can, inthis case, see sharply and writ large, how the avoidance between father-in-law and daughter-in-law is an aspect of the tension between successive genera-tions in a patrilineal kinship system.

Implicit and sometimes explicit comparison of African cultures with thoseof other areas is important in the recent history of field research in Africa.Seligmans pioneering researches in the Sudan were done against the back-ground of his experiences in New Guinea and among the Veddas cf. C . G.and B. 2. Seligman, 1932). More important, though, is the fact that the maintheoretical influence behind the field work of British anthropologists in Africain the middle twenties and the thirties was that of Malinowski. Now Malinow-skis Lfunctional heory is ordered to the concept of culture essentially in asense derived from Tylor and Frazer, and his empirical model was always theTrobrianders. It has taken twenty years for the Trobrianders to be placedin a proper comparative perspective in British social anthropology.

It is not, I think, a gross distortion to say that Malinowski thought ofculture fundamentally in terms of a utilitarian philosophy. The individualusing his culture to satisfy universal needs by attaining culturally definedends is central to his ethnographic work. It is in the real events of social life,in situations of work, ceremony, dance, dispute, that he saw the interconnec-tion of all aspects of culture. And this approach, crystallized in his formulafor the institution-the group, the universal need, the material basis, thelegal or mythical charter-has proved to be of the greatest value for theempirical task of field observation. Methodologically, i t might be describedas a form of clinical study. The net of enquiry is spread to bring in everythingthat actually happens in the context of observation. The assumption is thateverything in a peoples culture is meaningful, functional, in the here-and-now of its social existence. This is the cardinal precept for the anthropologicalstudy of a living culture. It is the basis of the rigorous observation and com-prehensive binding together of detail that marks good ethnographic field workof today. However we may now regard Malinowskis theories we cannot denyhim credit for showing us how intensive field work can and must be done.Tha t is, I believe, one of the major contributions made by social anthropologyto the social sciences, though it can probably only be satisfactorily used inhomogeneous and relatively stable societies or sections of societies.

What I am concerned with in these remarks is the local history of British


4/25

2 4 M E R I C AN ANTHROPOLOGIST [55, 1953

social anthropology. We all know that Malinowskis functionalism was partof a wider movement; bu t this is not my su bject. T he pointI am leading up tois this . Malinowski ha d no sense for social organization, th ou gh parado xicallyenough his most valuab le specific hypotheses fall within th e fram e of referenceof social organization. This applies, for instance to his restatement of theDurkheim ian hyp othesis of th e function of m y th as the ch arte r of a n insti-tu tio n, to his remarkable analysis of th e configurationof social relations inthe matrilineal family, and to his development of the concept of reciprocity.Bu t he had no real understanding of kinship or political organization. T hu she never overlooked an opportunityof pouring scorn on what he called kin-ship algebra, as I can vouch for from personal experience. This prejudiceprevented him from completing his often promised book on kinship.It isbeautifully documented in theSexual Life of Savages p. 447). Kinship is tohim primarily a tissue of culturally conditioned emotional att itud es. So he ispuzzled by the extension of th e term for fath er to th e father s sisters son:an d being q uite unable t o think in wh at we should now call structura l term s,he comm its th e appalling methodological solecism of att rib ut ing i t to ananomaly of language. Malinowski was reacting against the preoccupationwith terminologies and with conjectural reconstructibns of extinct marriagerules which wasso widespread in th e early yea rs of this cen tury . I t is a m easureof th e progress ma de since 1929 th a t no on e tod ay com ing acrossso obviousa case of a C hoctaw type lineage terminology would ma ke M alinowskisblunder.

Malinowskis bias is th e more inst ructive because of th e debt we owe tohis genius. It is reflected in the field work directly inspired by him. We seethis in what I regard a s the most outstandin g contribution to African ethno g-raphy we have as yet ha d, Evans -Pritcha rds Zande book 1937).It is notab lethat he refers only incidentally and casually to the way witchcraft andoracles are tied up with Zande political organization. Firths studyof Tikopiakinship 1937) is a n exception for it s grasp of th e theory of social organ ization;bu t he still held th e view t h a t social organization is a n asp ect of cult ureofthe same modality as the others usually enumerated by Malinowski.I men-tion these two books because they mark important steps in the advance ofboth e thnography and theory; andI am not suggesting that they followawrong track. What I wa nt to stress is th at the y follow the trac k w hich leadsto cd lture as th e global concept subsum ing everything t h a t goes on in sociallife. A serious lim itation t o this point of viewis th at i t is bound to treat every-th ing in social life as of equal weight, all aspects a s of eq ua l significance.Th ere is no way of establishing an order of prio rit y where all in stit uti on s ar einterdependent, except by criteria that cannot be used in a synchronic study;and synchronic study is thesine qu non of functional research. There is, forinstance , the criterion of viab ility over a st re tc h of time which enablesus t o


5/25


6/25

AMERICAN ANTHROPOLOGIST 15.5 1953

tinction bot h to ethnographic discoveries of recent years and to th e cata lyt icinfluence exercised on their thought by Radcliffe-Brown since his return toEngland from Chicago in 1937. B u t the distinction had of course long beenimplicit in the work of earlier ethnologists. W e need only th ink of the co nt ra stbetween Lewis Morgan , whose idiom of th ou ght was in ter m s of a social sys-tem, and Tylor, who thought in termsof custom and often had recourse topsychological hypotheses. Rivers 1914) whose own work an d influence inEngland contribu ted significantly to th e development of th e idea of socialstruc ture, saw this.So did Low ie whosePrimitive Society 1921) is, I suppose,the first at te m pt a t a system atic analysis of what we should now call theprinciples of social structure in primitive society. What he brought out wasthe very obvious but fu nda m enta l fact t ha t closely similar, if not identical,forms of social relationship occur in w idely se pa ra te societies an d are expressedin varied custom.

By social organization or social structure, terms which they used inter-changeably, Rivers an d Lowie meant prim arily th e kinship, political and legalinstitutions of primitive peoples. And these, in fact, are the institutions withwhich British anthropologists are mainly concerned when they write aboutsocial structure. The advantag eof this term, as opposed to the more usualterm social organization is that it draws attention to the interconnectionan d interdependence, within a single system of all the different classes ofsocial relations found within a given society. This leads to questions beingasked ab ou t the na tu re of these interconnections an d th e forces behind th esystem a s a w hole.

What I want to stress is th at the sp ur to the current interest in structuralstudies in Britain comes in equal measure from field experience, especially inAfrica, and theory. Anybody who has tried to understand African religiousbeliefs and practices in the field knows, for example, that one cannot get farwith out a very tho rough knowledge of the kinship an d political organization.These studies have thus given new content to the familiar postulate that aliving culture is an integ rated un ity of some sor t. We can see more clearlytha n twenty years ago th at this is due not t o metaphysical qualities mysteri-ously diffused through it b ut to t he function of customs an d ins titu tio ns inexpressing, marking and ma intain ing social relations between personsand groups. It is this which underlies the consistencies between the customsan d institutions of a people th a t are comm only emphasized.A uni t must ,by definition, have a boundary. A cultu re, certa inly in most of A frica, andIve ntu re to believe in many o the r areas too as indeed Wissler long ago stressed),has no clear-cut boundaries. But a groupof people bound together withina single social structure have a boundary, though not necessarily one thatcoincides with a physical boundary or is impenetrable.1 would suggest thata culture is a unit y inso far as it is tied t o a bounded social structure. I n this


7/25

FORTES] S T R U C T U R E OF U N I L I N E A L DESCENT GROUPS 23

sense I would agree t h a t t he social st ru ct ur e is th e foundation of th e wholesocial life of any continuing society. Here again Rivers showed great insightwhen he sta te d 1911) th at the social stru ctu re is th e featu re of a peoples

social life which is most resistant to change.It is certainly a striking fac t th atth e family an d kinship institu tions of a continuing society in Africa displayrema rkable persistence in th e face of big changes in eve ryday h ab its, in ritu alcustoms an d belief, an d even in ma jor economic an d social goals. Th e Tsw anacf. Schap era, 1940 an d 1950) are a good instance. B ut we m ust be careful.

There is also plenty of evidence from em igran t groups, such as Chinese, Ea stIndians an d particularly t he Negro populations of th e New World cf. Hersko-vits, 1948, p. 542 ff. of the retention of religious a nd aesthetic cus tom s in theface of radical changes in struc tura l arrangem ents. Thisis a warning against

thinking of culture a nd social stru ctu re as mu tua lly exclusive. T he socialstru ctu re of a group does not exist without the custom ary norms a nd activitieswhich work through i t. We might safely conclude t h a t where st ru ct ur e persiststhere m ust be som e persistence of corresponding custom an d where c ustomsurvives there must be some structural basis for this. ButI think it would beagreed that though the customs of any continuing and stable society tend tobe consistent because they are tied to a coherent social structure, yet thereare im porta nt factors of auto nom y in custom. This has often been pointedou t ever since the facts of diffusion became know n. T he p a rt play ed by dis-

positional and psychogenetic factors in the con ten t an d action of custom isnow being clarified. A house is no t reducible to its foundations an d custom isnot reducible simply to a manifestation of social st ructure.

The recent trend in British social anthropology springs, asI have said,primarily from field experience. Evans-P ritcha rds description of N ue r lineageorganization 1933-35), F irth s accou nt of Tikopia kinsh ip 1937) an d Fordesanalysis of clan and kin relations among th e Yak0 1938-39) are th e im po rtan tethnographic landmarks. A prominent feature in all three is the attentiongiven to the part played by descent rules and institutions in social organiza-

tion, and the recognition t ha t th ey belong as mu ch t o th e sphere of politicalorganization as to th a t of kinship. Following this lead, oth er stu de nt s ha vebeen making in tensive studies of the role of descent principles in Africanso-cieties where unilineal descent groups often co ns tit ut e th e genealogical basis ofsocial relations. Good ethnog raphy is bo th a continuous te st of existing hy-potheses and continuously creative of theory an d technique; a nd th is is ha p-pening so rapidly just at present that one can hardly keep pace withit. T h eyounger research workers to whose unpublished materialI shall be referringare developing struc tura l analysis int o a very effective technique a nd applying

it not only in Africa bu t also in India , New Guinea an d Indonesia.Seen against the backgroundI have sketched, there is no doubt that big

gains have been m ade in th e st ud y of social stru ctu re since th e nineteen-


8/25

24 AMERICAN ANTHROPOLOGIST [55 1953

twenties. This is well illustrated in recent investigations of unilineal descentgroups, bo th in Africa an d elsewhere cf. Eggan, 1950; Gough, 1950) b u tIwill deal mainly with th e African da ta . We ar e now ina position to formu latea num ber of con nected generalizations abo ut th e stru ctu re of t h e unilinealdescent group, and its place in the total social system which could not havebeen stated twenty years ago. It is moreover important to note that theyseem to hold for bo th patrilineal an d matrilineal groups. Som e of th e condi-tions governing th e emergence of such descent groups h av e recently b een dis-cussed by Forde 1947). He m akes the interesting suggestion th a t pove rty ofha bi ta t and of productive technology ten d to inhibit th e developm entofunilineal descent groups by limiting the scale and stability of settlement.Ta ki ng this in association with Lowies hypothes is of 1921 Lowie, 1921,p. 149)th a t the establishm ent of th e principle of un ilateral descent is ma inly du e toth e transmission of p rop erty rights a nd t he mode of residence afte r marriage,we have two sides of an hypothesis that deserves much further testing. Theground has been well cleared for this by M urdock 1949). F or it does seemth a t unilineal descent groups a re not of significance am on g peoples who livein small groups, depend on a rudim entary technology, a nd have little durableproperty. On the othe r hand, there is evidence th at they break down when amodern economic framewo rk with occup ational differentiation linked t oa widerange of specialized skills, to productive capital and to monetary media ofexchange is introduced Spoehr, 1947; Eg gan , 1950; Gough, 1950). W herethese groups are most in evidence is in the m iddle range of relatively hom o-geneous, pre-cap italistic economies in which the re is some degree of technologi-cal sophistication an d value is attac hed t o rights in durable property. Th eyma y be pastoral economies like the Nue r Eva ns-Pritch ard, 1940) an d theBeduin Peters, 1951), or agricultu ral economies like those of th e Yak0 Fo rde,1938, 1950), th e Tallensi Fo rtes, 1945, 1949) an d the G usii M aye r, 1949)-or if we look outside Africa, th e Tikopia Fi rth , 1937) an d the Hopi Eg gan ,1950) an d m any other peoples. Th e Na ya r of So uth In di a, classicallya testcase of kinship theories, ar e of pa rticular int ere st in th is connection, asarecent intensive field stud y by Dr . E .J. Miller and Dr. E. K. Gough shows.Tho ugh th e tota l economy of Sou th India was even formerlya very complexone, the Nayar themselves traditionally formeda caste of very limited occu-pational range. It is only during the past h undred years or so th at they havegradually entered other occupations than soldiering and passive landlordism.And with this change has come the breakdown previously mentioned in theirrigid matrilineal lineage organization. This does not imply that unilinealdescent groups ar e either historically or functionally th e prod uct of economicand property institutions alone. Other factors are undoubtedly involved.Th ere is the example of t he H aus a of N or th ern Nigeria, for instance, whohav e a rural economy of th e same type as th at of th e Tallensi, though techni-


9/25

FORTES] S T R U C T U R E OF UN ILIN EA L DESCENT GROUPS 25

cally more elaborate, and well developed property concepts; but they haveno unilineal descent groups. The socially significant genealogical groupingamong them is of the cognatic type based on the equal recognition of kinties on both sides, as among the Lozi and other Central African tribes Dry,1950; Colson and Gluckman, 1951). Nor can the Hausa arrangement beascribed to the local influence of Islam since the Cyrenaican Beduin havesharply defined patrilineal lineages Peters, 1951).

I have lingered a little on this problem to bring home a point which Ihave already referred to. It is the problem of assigning an order of relativeweight to the various factors involved in culture and in social organization,or alternatively of devising methods for describing and analyzing a configura-tion of factors so as to show precisely how they interact with one another.Much as we have learned from intensive field work in relation to this task,we shall learn even more, I believe, from such studies of local variations withina uniform culture region as Radcliffe-Browns 1930), Schaperas in Radcliffe-Brown and Forde, 1950) and Eggans 1950).

The most important feature of unilineal descent groups in Africa broughtinto focus by recent field research is their corporate organization. When wespeak of these groups as corporate units we do so in the sense given to theterm corporation long ago by Maine in his classical analysis of testamentarysuccession in early law Maine, 1866). We are reminded also of Max Weberssociological analysis of the corporate group as a general type of social forma-tion Weber, 1947), for in many important particulars these African descentgroups conform to Webers definition. British anthropologists now regularlyuse the term lineage for these descent groups. This helps both to stress thesignificance of descent in their structure and to distinguish them from wideroften dispersed divisions of society ordered to the notion of common-butnot demonstrable and often mythological-ancestry for which we find it use-ful to reserve the label clan.

The guiding ideas in the analysis of African lineage organization havecome mainly from Radcliffe-Browns formulation of the structural principlesfound in all kinship systems cf. Radcliffe-Brown, 1950). I am sure I am notalone in regarding these as among the most important generalizations as yetreached in the study of social structure. Lineage organization shows veryclearly how these principles work together in mutual dependence, so thatvarying weight of one or the other in relation to variations in the wider contextof social structure gives rise to variant arrangements on the basis of the samebroad ground-plan.

A lineage is a corporate group from the outside, tha t is in relation to otherdefined groups and associations. It might be described as a single legal per-sonality-one person as the Ashanti put it Fortes, 1950). Thus the waya lineage system works depends on the kind of legal institutions found in the


10/25


11/25

FORTES] STRUCTURE OF UNILINEAL DESCENT GROUPS 27

perpetuity in time. Where the lineage concept is highly developed, the lineageis thought t o exist as a perpetual corporation as long as any of its memberssurvive. This means, of course, not merely perpetual physical existence

ensured by the replacement of departed members. It means perpetual struc-tural existence, in a stable and homogeneous society; that is, the perpetualexercise of defined rights, duties, office and social tasks vested in the lineageas a corporate unit. The point is obvious but needs recalling as it throwslight on a widespread custom. We often find, in Africa and elsewhere, that aperson or descent group is attached to a patrilineal lineage through a femalemember of the lineage. Then if there is a danger t hat rights and offices vestedin the lineage may lapse through the extinction of the true line of descent,the attached line may by some jural fiction be permitted to assume them. Or

again, rather than let property or office go to another lineage by default ofproper succession within the owning lineage, a slave may be allowed to suc-ceed. In short, the aim is to preserve the existing scheme of social relationsas far as possible. As I shall mention presently, this idea is developed mostexplicitly among some Central African peoples.

But what marks a lineage out and maintains its ident ity in the face of thecontinuous replacement by death and birth of its members is the fact tha tit emerges most precisely in a complementary relationship with or in opposi-tion to like units. This was first precisely shown for the Nuer by Evans-

Pritchard and I was able to confirm the analysis among the Tallensi Fortes,1949). It is characteristic of all segmentary societies in Africa so far described,almost by definition. A recent and most interesting case is tha t of the Tivof Northern Nigeria P. J. Bohannan, 1951). This people were, until thearrival of the British, extending their terri tory rapidly by moving forward enmasse as their land became exhausted. Among them the maximal lineagesare identified by their relative positions in the total deployment of all thelineages and they maintain these positions by pushing against one another asthey all move slowly forward.

The presumed perpetuity of the lineage is what lineage genealogies con-ceptualize. If there is one thing all recent investigations are agreed upon it isthat lineage genealogies are not historically accurate. But they can be under-stood if they are seen to be the conceptualization of the existing lineagestructure viewed as continuing through time and therefore projected back-ward as pseudo-history. The most striking proof of this comes from Cyrenaica.The Beduin there have tribal genealogies going back no more than the four-teen generations or thereabouts which we so commonly find among AfricanNegro peoples; but as Peters points out , historical records show that they have

lived in Cyrenaica apparently in much the same way as now for a much longertime than the four to five hundred years implied in their genealogies. Dr. P. J.and Dr. L. Bohannan have actually observed the Tiv a t public moots rear-


12/25

28 A M E R I C A N A N T H R O P O L O G I S T [55 1953

ranging their lineage genealogies to bring them into line with changes in theexisting pattern of legal and political relations within and between lineages.A genealogy is, in fact, what Malinowski called a legal charter and not anhistorical record.

A society made up of corporate lineages is in danger of splitting into rivallineage factions. How is this counteracted in the interests of wider politicaluni ty? One way is to extend the lineage framework to the widest range withinwhich sanctions exist for preventing conflicts and disputes from ending infeud or warfare. The political unit is thought of then as the most inclusive,or maximal, lineage to which a person can belong, and it may be conceptualizedas embracing the whole tribal unit. This happens among the Gusii Mayer,1949) as well as among the Nuer, the Tiv and the Beduin; but with the lastthree the tribe is not the widest field within which sanctions against feud andwar prevail. A major lineage segment of the tribe is the defucto political unitby this definition.

Another way, widespread in West Africa but often associated with thepreviously mentioned structural arrangement, is for the common interest ofthe political community to be asserted periodically, as against the privateinterests of the component lineages, through religious institutions and sanc-tions. I found this to be the case among the Tallensi Fortes, 1940) and thesame principle applies to the Yak0 Forde, 1950 b)) and the Ibo Forde andJones, 1950). I believe it will be shown to hold for many peoples of the WesternSudan among whom ancestor worship and the veneration of the earth are thebasis of religious custom. The politically integrative functions of ritual insti-tutions have been described for many parts of the world. What recent Africanethnography adds is detailed descriptive data from which further insight intothe symbolism used and into the reasons why political authority tends to beinvested with ritual meaning and expression can be gained. A notable instanceis Dr . Kupers 1947) account of the Swazi kingship.

As the Swazi data indicate, ritual institutions are also used to supportpolitical authority and to affirm the highest common interests in Africansocieties with more complex political structures than those of segmentarysocieties. This has long been known, ever since the Divine Kingship of theShilluk cf. Evans-Pritchard, 1948) brought inspiration to Sir James Frazer.But these ritual institutions do not free the individual to have friendly andco-operative relations with other individuals irrespective of allegiance tocorporate groups. If such relations were impossible in a society it could hardlyavoid splitting into antagonistic fractions in spite of public ritual sanctions,or else it would be in a chronic state of factional conflict under the surface. Itis not surprising therefore to find that great value is attached to widely spread-ing bonds of personal kinship, as among the Tallensi Fortes, 1949). The recentfield studies 1 have quoted all confirm the tremendous importance of the web


13/25

FORTES] STRUCTURE OF UN ILI NE AL DESCENT GROUPS 29

of kinship as a coun terweight to t he tend ency of unilineal descent grouping toharden social barriers. Or to p u t i t slightly differently, it seems th a t w here th eunilineal descent group is rigorously st ruc tur ed within th e tota l social systemthere we ar e likely to find kinship used to define and sanction a personal fieldof social relations for each indiv idual.I will come back to this point in a mo-ment. A fu rth er point to whichI will refer again is this. We ar e learnin g fromconsiderations such as thoseI have just m entioned, to think of social stru ctu rein term s of levels of organization in th e m anne r first explicitly followed in th epresentation of field d a ta by W arner 1937). We can investigate the t ota l socialstru ctu re of a given comm unity a t th e level of local organization, a t th a t ofkinship, a t the level of corporate group structure and government, an d a tth a t of rit ua l institutions . We see these levels ar e related t o different collec-tive interests, which are perhaps connected in some sortof hierarchy. And oneof t he problems of analysis and exposition is to perceive an d st a te th e factth a t all levels of str uc tu re are simultaneously involved in eve ry social relation-ship and a ctivity . This restatement of what is commonly me ant b y t he con-cept of integra tion ha s the ad va ntag e of suggesting how the different modesof social relationship distinguished in an y society ar e interlocked with oneanother. It helps to m ake clear also how ce rtain basic principles of socialorganization can be generalized throughout the whole struct ureof a primitivesociety, as for instance t he segm entary principle among the Nu er an d the T al-lensi.

Thi s way of thinking a bo ut the problem of social integ rationhas been use-ful in recent s tud ies of African politica l organ ization. S tu dy of t he unilinealdescent group a s a par t of a t ota l social system m eans in fact stu dy ing its func-tions in th e widest framew ork of social str uc tu re , th a t of the political organiza-tion. A comm on an d perhaps general feat ure of po litical organization in Africais th a t it is built u p in a series of laye rs,so to speak, so arranged t h a t the prin-ciple of checks an d balances is necessarily m obilized in political activit ies. T heidea is used in a variety of ways bu t w hat i t comes to in general is th a t th emembers of the society are distributed in different, nonidentical schem es ofallegiance and mutual dependence in relation to administrative, juridical andritual institutions. It would take too long to enumerate all the peoples forwhom we now have sufficient data to show this in detail. But the Lozi ofNorthern Rhodesia Gluckman, 1951) are of such particular theore ticalinterest in this connection th ata word m ust be said abou t them. T h e corporatedescent group is not found among th em . Inste ad th eir political organizationis based on what Maine called the corporation sole. This is a title carryingpolitical office backed by ritual sanctions and symbols to which subjects,lands, jurisdiction, and representative status, belong. But every adult isbound to a num ber of titles for dif fe rent legal an d social purposes in such away th at w ha t is one allegiance group with respect to one titleis split up with


14/25

30 A M E R I C A N A N T H R O P O L O G I S T [SS 1953

reference to other titles. Thus the only all-inclusive allegiance is that of allthe nation to the kingship, which is identified with the State and the countryas a whole. A social structure of such a kind, knit together moreover by awidely ramifying network of bilateral kinship ties between persons, is well for-tified against internal disruption. It should be added that the notion of thecorporation sole is found among many Central African peoples. It appears, infact, to be a jural inst itution of the same generality in any of these societiesas corporate groups are in others, since it is significant a t all levels of socialstructure. A good example is the Bemba cf. Richards, 1936, 1940b) amongwhom it is seen in the custom of positional inheritance of sta tus, rank,political office and ritual duty, as I will explain later.

What is the main methodological contribution of these studies? I n myview it is the approach from the angle of political organization to what aretraditionally thought of as kinship groups and institut ions that has beenspecially fruitful. By regarding lineages and statuses from the point of viewof the total social system and not from t hat of an hypothetical E G O werealize that consanguinity and affinity, real or putative, are not sufficient inthemselves to Bring about these structural arrangements. We see that descentis fundamentally a jural concept as Radcliffe-Brown argued in one of hismost important papers 1935) ; we see its significance, as the connecting linkbetween the external, th at is political and legal, aspect of what we have calledunilineal descent groups, and the internal or domestic aspect. It is in thelatter context that kinship carries maximum weight, first, as the source oftitle to membership of the groups or to specific jural sta tus, with all th a t thismeans in rights over and toward persons and property, and second as thebasis of the social relations among the persons who are identified with oneanother in the corporate group. In theory, membership of a corporate legalor political group need not stem from kinship, as Weber has made clear. Inprimitive society, however, if it is not based on kinship i t seems generally topresume some formal procedure of incorporation with ritual initiation, So-called secret societies in West Africa seem to be corporate organizations ofthis nature. Why descent rather than locality or some other principle formsthe basis of these corporate groups is a question tha t needs more study. Itwill be remembered that Radcliffe-Brown 1935) related succession rules tothe need for unequivocal discrimination of rights in rem and in personam.Perhaps it is most closely connected with the fact that rights over the repro-ductive powers of women are easily regulated by a descent group system.But I believe tha t something deeper than this is involved; for in a homogeneoussociety there is nothing which could so precisely and incontrovertibly fix onesplace in society as ones parentage.

Looking a t i t from without, we ignore the internal structure of the unilinealgroup. But African lineages are not monolithic units; and knowledge of their


15/25

FORTES] S T R U C T U R E OF U N I L I N E A L DESCENT GROUPS 31

internal differentiation has been much advanced by the researches I havementioned. T he dynamic character of lineage structure can be seen most easilyin the balance that is reached between its external relations and its internalstructure. Ideally, in most lineage-based societies the lineage tends to bethought of as a perpetual unit, expanding like a balloon but never growingnew parts. In fact , of course, as Forde 1938) and Evans-Pritchard 1940)have so clearly shown, fission and accretion are processes inherent in lineagestructure. However, it is a common experience to find an informant who re-fuses to admit tha t his lineage or even his branch of a greater lineage did notat one time exist. Myth and legend, believed, naturally, to be true history,are quickly cited to prove the contrary. But investigation shows that thestretch of time, or rather of duration, with which perpetuity is equated variesaccording to the count of generations needed to conceptualize the internalstructure of the lineage and link it on to an absolute, usually mythologicalorigin for the whole social system in a first founder.

This is connected with the fact that an African lineage is never, accordingto our present knowledge, internally undifferentiated. It is always segmentedand is in process of continuous further segmentation a t any given time.Among some of the peoples I have mentioned e.g. the Tallensi and probablythe Ibo) t he internal segmentation of a lineage is quite rigorous and the processof further segmentation has an almost mechanical precision. The generalrule is that every segment is, in form, a replica of every other segment andof the whole lineage. But the segments are, as a rule, hierarchically organizedby fixed steps of greater and greater inclusiveness, each step being defined bygenealogical reference. It is perhaps hardly necessary to mention again thatwhen we talk of lineage structure we are really concerned, from a particularanalytical angle, with the organization of jural , economic, and ritual activities.The point here is that lineage segmentation corresponds to gradation in theinstitutional norms and activities in which the total lineage organization isactualized. So we find t ha t the greater the time depth tha t is attri buted to thelineage system as a whole, the more elaborate is its internal segmentation.As I have already mentioned, lineage systems in Africa, when most elaborate,seem to have a maximal time depth of around fourteen putative generations.More common though is a count of five or six generations of named ancestorsbetween living adults and a quasi-mythological founder. We can as yet onlyguess at the conditions tha t lie behind these limits of genealogical depth inlineage structure. The facts themselves are nevertheless of great comparativeinterest. As I have previously remarked, these genealogies obviously do notrepresent a true record of all the ancestors of a group. To explain this by thelimitations and fallibility of oral tradit ion is merely to evade the problem.In structural terms the answer seems to lie in the spread or span Fortes,1945) of internal segmentation of the lineage, and this apparently has inherent


16/25

32 AMERICAN ANTHROPOLOGIST [ 5 5 1953

limits. As I interpret the evidence we have, these limits are set by the condi-tion of stability in the social stru ctu re which it is one of the chief functionsof lineage systems to maintain . T he segm entary spread found in a givenlineage system is tha t which makes for maximum sta bility ; an d in a stablesocial system it is kept at a particular spread by continual internal adjust-ments which are conceptualized by clipping, patching an d telescoping genealo-gies to fit. Just w hat th e optimum spreadof lineage segmentation in a particu-la r society tends to be depends presumably on extra-lineage factors of politicalan d economic organization of th e kind referred to b y F orde 1947).

It is when we consider the lineage from within that kinship becomesdecisive. For lineage segmentation followsa model laid down in the parentalfamily. It is indeed generally thought of as th e perpetuation, through th e ruleof the jural u ni ty of t he descent line andof the sibling grou p cf. Radcliffe-Brown, 1951),of the social relations that constitute the parental family.Sowe find a lineage segment conceptualized as a sibling group in symmetricalrelationship w ith segmentsof a like order. It will be a paternal sibling groupwhere descent is patrilineal and a maternal one where it is matrilineal. Pro-gressive orders of inclusiveness ar e formula ted a s a succession of generations;an d the actu al process of segm entation is seen as the equiv alent of the divisionbetween siblings in the paren tal family. W ith this goes the use of kinshipterminology and the application of k inship norms in the regulation of i nt ra -lineage affairs.

As a corporate group, a lineage exhibitsa struc ture of auth ority, an d it isobvious from w ha tI have said why th is is aligned with th e generation ladder.We find, as a general rule, t h a t no t only th e lineage b u t also every segmentofit h as a head, b y succession or election, who m anages its affairs with th e adv iceof his co-members. H e may n o t have legal sanctions b y mean sof which toenforce his authority in internal affairs; but he holds his position by consentof all his fellow members, an d he is backed by moral sanc tions comm onlycouched in religious concepts. H e is th e tru ste e for the whole group of th eproperty an d othe r productive resources vested in it. H e ha s a decisive juralrole also in th e disposal of rights over th e fertility of th e women in th e group.H e is likely to be th e representative of th e whole group in political a nd legalrelations with other groups, with political authorities, and in communalritual. T he effect may be to m ake him p u t t h e interests of his lineage abovethose of th e com munity if there is conflict with th e latter . T hi s is quite clearlyrecognized by some peoples. Among the Ashanti for instance, every chiefshipis vested in a matrilineal lineage. Bu t once a chief has been installed his con sti-tutiona l position is defined a s holding an office th at belongs to th e wholecomm unity no t to an y one lineage. T h e man is, ideally,so merged in th e officeth a t he virtually ceases to be a m ember of h is lineage, which always ha s anindependent head for its corporate affairs cf. Busia,19.50 .


17/25


18/25

34 A M E R I C A N A N T H R O P O L O G I S T [55 1953

brought out however by an example. A Tiv may claim to be living with aparticular group of relatives for purely personal reasons of convenience oraffection. Investigation shows th at he has in fact made a choice of where t o

live within a str ict ly limited range of nonlineage kin. What purports to be avoluntary act freely motivated in fact presupposes a st ructural scheme of in-dividuation. This is one of the instances which show how it is possible andfeasible t o move from the structural frame of reference to another, here tha tof the social psychologist, without confusing data and aims.

Most far-reaching in its effects on lineage structure is the use of the ruleof complementary filiation to build double unilineal systems and some strikinginstances of th is are found in Africa. One of the most developed systems ofth is type is that of the Yako; and Fordes excellent analysis of how this works

Forde, 1950) shows that it is much more than a device for classifying kin. Itis a principle of social organization t ha t enters into all social relations and isexpressed in all important institutions. There is the division of property, forinstance, into the kind that is tied to the patrilineal lineage and the kind thatpasses to matrilineal kin. The division is between fixed and, in theory, per-petual productive resources, in this case farm land, with which goes residencerights, o n the one hand, and on the other, movable and consumable propertylike livestock and cash. There is a similar polarity in religious cult and in thepolitical office and authority linked with cult, the legally somewhat weaker

matrilineal line being ritually somewhat stronger than the patrilineal line.This balance between ritual and secular control is extended to the fertilityof the women. An analogous double descent system has been described forsome Nuba Hill tribes by Nadel 1950) and its occurrence among the Hererois now classical in ethnology. The arrangement works the other way round,t o o in Africa, as among the Ashanti, though in their case the balance is farmore heavily weighted on the side of the matrilineal lineage than on that ofthe jurally inferior and noncorporate paternal line.

These and other instances lead to the generalization that complementary

filiation is not merely a constant element in the pattern of family relationshipsbut comes into action a t all levels of social structure in African societies. Itappears th at there is a tendency for interests, rights and loyalties to be dividedon broadly complementary lines, into those that have the sanction of law orother public institut ions for the enforcement of good conduct, and thosethat rely on religion, morality, conscience and sentiment for due observance.Where corporate descent groups exist the former seem to be generally tied tothe descent group, the lat ter to the complementary line of filiation.

If we ask where this principle of social structure springs from we must

look to the tensions inherent in the st ructure of the parental family. Thesetensions are the result of the direction given to individual lives by the totalsocial structure but they also provide the models for the working of tha t


19/25

FORTES] STRUCTURE OF UN ILIN EA L DESCENT GROUPS 35

structure. We now have plenty of evidence to show how the tensions tha tseem normally to arise between spouses, between successive generations andbetween siblings find expression in custom and belief. In a homogeneous societythey are apt to be generalized over wide areas of the social structure. Theythen evoke controls like the Nyakyusa separation of successive generations ofmales in age villages that are built into the total social structure by the de-vice of handing over political power to each successive generation as it reachesmaturity Wilson, 1951). Or this problem may be dealt with on the leveI ofritual and moral symbolism by separating parent and first born child of thesame sex by taboos that eliminate open rivalry, as among the Tallensi, theNuer, the Hausa and other peoples.

Thus by viewing the descent group as a continuing process through timewe see how it binds the parental family, its growing point, by a series of stepsinto the widest framework of social structure. This enables us to visualize asocial system as an integrated unity at a given time and over a stretch oftime in relation to the process of social reproduction and in a more rigorousway than does a global concept of culture.

I do want to make clear, though, that we do not think of a lineage as beingjust a collection of people held together by the accident of birth. A descentgroup is an arrangement of persons t hat serves the attainment of legitimatesocial and personal ends. These include the gaining of a livelihood, the settingup of a family and the preservation of health and well-being as among themost important. I have several times remarked on the connection generallyfound between lineage structure and the ownership of the most valued pro-ductive property of the society, whether it be land or cattle or even themonopoly of a craft like blacksmithing. It is of great interest, for instance, t ofind Dr. Richards att ributing t he absence of a lineage organization among theBemba to their lack of heritable right in land or livestock Richards, 1950).A similar connection is found between lineage organization and the controlover reproductive resources and relations as is evident from the common occur-rence of exogamy as a criterion of lineage differentiation. And since citizenshipis derived from lineage membership and legal status depends on it, politicaland religious office of necessity vests in lineages. W e must expect to find andwe do find that the most important religious and magical concepts and insti-tutions of a lineage based society are tied into the lineage structure servingboth as the necessary symbolical representation of the social system and asits regulating values. This is a complicated subject about which much moreneeds to be known. Cults of gods and of ancestors, beliefs of a totemic nature,and purely magical customs and practices, some or all are associated withlineage organization among the peoples previously quoted. What appears tohappen is that every significant structural differentiation has its specificritual symbolism, so that one can, as it were, read off from the scheme of


20/25

36 -4MERICZ IN A NTIIROPOLOGIST 15.5 1953

ritua l differentiation th e pa tte rn of s tru ctu ral differentiation an d th e con-figuration of norms of conduct th at goes wi th it. The re is, to p u t it simply,asegmentation of ritual allegiance corresponding to the segmentation of gene-alogical grouping. Locality, filiation, descent, individuation, are thus sym-bolized.

Reference to locality reminds us of Kroebers careful argum en t of1938 infavor of the prio rity of the local rela tionshipsof residence ove r those of descen tin determining the line that is legally superior. A lineage cannot easily act asa corpo rate group if its members can never g et to geth er for th e conduct oftheir affairs. It is not surprising therefore to find that the lineage in Africansocieties is generally locally anchored; but it is not necessarily territoriallycompact or exclusive. A compact nucleus may be enough to act as the localcenter for a group that is widely dispersed.I think i t would be agreed thatlineage and locality are independently variable an d how the y inte ract de pend son other factors in the social structure.As I interpret the evidence, local tiesar e of secondary significance,puce Kroeber, for local ties do not appear togive rise to stru ctur al bonds in and of themselves. T he re mu st be commonpolitical or kinship or economic or ritual interests for structural bonds toemerge. Again sp atial dispersion does not immediately p u t a n end to lineageties or to the ramifying kin ties found in cognatic system s like th at of t h eLozi. For legal status, property, office and cult act centripetally to hold dis-persed lineages together and to bind scattered kindred. This is important inth e dynam ic patt ern of lineage organization for i t contains w ithin itself th esprings of disi nteg ritio n, a t th e corporate level in the rule of se gm ent atio n,a t t he individual level in th e ruleof complementary filiation.

As I have suggested before, it seems that corporate descent groups canexist only in more or less homogeneous societies. Just what we mean by ahomogeneous society is still ra the r vague th ou gh we all use th e ter m lavishly.T h e working definitionI m ak e use of is th a t a homogeneous society is ideallyone in which an y person in t h e sense given t o thi s term by Radcliffe-Brown inhis recent 1950) essay, can be sub stitu ted for any o the r person of th e samecategory without bringing about changes in the social structure. This impliesth at any two persons of t he sam e category have th e same body of custom aryusages and beliefs.I relate th is tent ati ve definition to the rule of sibling equiva-lence, so t ha t I would say that, considered with respect to their achievablelife histories, in a homogeneous society all men are brothers and all womensisters.

Societies based on unilineal descent groups are not the best in which tosee wh at the notion of social substituta bility means. For t h a t it is bet ter toconsider societies in which descent still takes primacy over all other criteriaof association an d classification of persons in t he regulation of social life b u tdoes no t serve a s the con stitut ive principle of corporate g roup organ ization.


21/25

FORTES] S T R U C T U R E OF U N I L I N E A L D E S C E N T G R O U P S 37

Central Africa provides some admirable instances cf. Richards, 1950; Colsonand Gluckman, 1951). Among the Bemba, the Tonga, the Lozi and many oftheir neighbors, as I have already remarked, the social struc ture must be

thought of a s a system of interconnected politico-legal statuses symbolizedand sanctioned by ritual and not as a collection of people organized in self-perpetuating descent units. The stabili ty of the society over time is preservedby pe rpetuating the sta tus system. Thus when a person dies his sta tus is kep talive by being taken up by an heir; and this heir is selected on the basis ofdescent rules. A t any given time an individual may be the holder of a clusterof sta tuses; bu t these may be distributed among several persons on his deathin a manner analogous to the widespread African custom by which a mansinherited estate goes to his lineage heir and his self-acquired property to his

personal heir. Ideally, therefore, the network of sta tuses remains stable an dperpetual though their holders come and go. Ritual symbols define andsanction the key positions in the system. What it represents, in fact, is thegeneralization throughout a whole society of the notion of the corporation soleas tied to descent bu t no t to a corporate group. Descent and filiation have thefunction of selecting individuals for social positions an d roles-in other words,for the exercise of particular rights and obligations-just as in cross cousinmarriage they serve to select egos spouse.

The concept of the person as an assemblage of statuses has been the

sta rting point of some interesting enquiries. A generalization of long standingis th at a married person always has two mutually antagonistic kinship statuses,that of spouse and paren t in one family context and th a t of child and siblingin another cf. Warner, 1937). This is very conspicuous in an exogamouslineage system; and the tensions resulting from this condition, connected asthey are with the rule of complementary filiation, have wide consequences.A common rule of social structure reflected in avoidance customs is that thesetwo statuses must not be confounded. Furthermore, each status can be re-garded as a compound of separable rights and obligations. Th us a problem

th at has to be solved in every matrilineal society is how to reconcile the rightsoyer a womans procreative powers rights in genetricem as Laura Bohannanhas called them in her paper of 1949) which remain vested in her brother orher lineage, with those over her domestic and sexual services rights in uxoremcf. L. Bohannan, loc. c i t . which pass to her husband. Among the Yao ofNyassaland, as Dr. Clyde Mitchell has shown 1950), this problem underliesthe process of lineage segmentation. Brothers struggle against one anotheror sisters sons against mothers brothers) for th e control of the ir sisters

procreative powers and this leads to fission in the minimal lineage. I t is of

great significance that such a split is commonly precipitated by accusationsof withcraf t against the brother from whose control the sisters are withdrawn.By contrast, where rights over a womans child-bearing powers are held by her


22/25

38 AM ER IC AN ANTIIROPOLOGZST [55 1953

husbands patrilineal lineage the conflicts related to this critical interestoccur between the wives of a lineage segment; and among the Zulu and Xhosaspeaking tribes of South Africa these lead to witchcraft accusations betweenco-wives cf. Hunter , 1936). As Laura Bohannans paper shows, many wide-spread customs and institutions connected with marriage and parenthood,such as the levirate and the sororate, wife-taking by women, exchange mar-riage as practiced by the Tiv, and ghost marriage as found among the NuerEvans-Pritchard, 1951) have structural significance not hitherto appreciated

if they are regarded from the point of view I have indicated.But one thing must be emphasized. This method of analysis does no t ex-

plain why in one society certain kinds of interpersonal conflict are sociallyprojected in witchcraft beliefs whereas in another they may be projected interms of a belief in punitive spirits. It makes clear why a funeral ceremony isnecessary and why it is organized in a particular way in the interest of main-taining a stable and coherent social system. It does not explain why theritual performed in the funeral ceremonies of one people uses materials, ideasand dramat izations of a different kind from those used by another people. Inshort, it brings us nearer than we were thirty years ago to understanding themachinery by which norms are made effective, not only in a particular primi-tive society but in a type of primitive society. It does not explain how thenorms come to be what they in fact are in a particular society.

In this connection, however, it is worth drawing attention to certainnorms that have long been recognized to have a critical value in social organiza-tion. Marriage regulations, incest prohibitions and the laws of homicide andwarfare are the most important. Analysis of lineage structure has revealed anaspect of these norms which is of great theoretical interest. It is now fairlyevident th at these are not absolute rules of conduct which men are apt tobreak through an outburst of unruly instinct or rebellious self-assertion, ashas commonly been thought . They are relatively obligatory in accordance withthe structural relations of the parties. T he Beduin of Cyrenaica regard homi-cide within the minimal agnatic lineage, even under extreme provocation, asa grave sin, whereas slaying a member of a different tribal segment is an ad-mirable deed of valor. The Tallensi consider sex relations with a near sisterof the same lineage as incest but tacitly ignore the act if the parties are verydistant lineage kin. Among the Tiv, the Nuer, the Gusii and other tr ibes thelineage range within which the rule of exogamy holds is variable and can bechanged by a ceremony that makes formally prohibited marriages legitimateand so brings marriage prohibitions into line with changes in the segmentarystructure of the lineage. In this way previously exogamous units are split intointermarrying units . I n all the societies mentioned, and others as well, an ac tof self-help tha t leads to negotiations if the parties belong to closely related


23/25

FORTES] STRUCTURE OF U N I L I N E A L DESCENT GROUPS 39

lineages might lead to war if they are members of independent-though notnecessarily geographically far apart-lineages. Such observations are indica-tions of the flexibility of primitive social structures. They give a clue to theway in which internal adjustments are made from time to time in those struc-tures, either in response to changing pressures from without or through themomentum of their own development. They suggest how such societies canremain stable in the long run without being rigid. But this verges on specula-tion.

The contributions to African ethnography mentioned in this paper areonly a small and arbitrary selection from a truly vast amount of new workthat is now going on in several countries. My aim has been to suggest how thiswork links up with a theoretical approach that is much in evidence amongBritish social anthropologists. It is perhaps needless to add that this approachis also being actively applied by American, French, Belgian and Dut ch anthro-pologists concerned with the problems of social organization. What I wish toconvey by the example of current studies of unilineal descent group structureis that we have, in my belief, got to a point where a number of connectedgeneralizations of wide validity can be made about this type of social group.This is an advance I associate with the structural frame of reference. I wishto suggest that this frame of reference gives us procedures of investigationand analysis by which a social system can be apprehended as a unity madeof parts and processes that are linked to one another by a limited pumber ofprinciples of wide validity in homogeneous and relatively stable societies. Ithas enabled us to set up hypotheses about the nature of these principles thathave the merit of being related directly to the ethnographic material now soabundantly a t hand and of being susceptible of test ing by fur ther field obser-vation. It cannot be denied, I think, that we have here a number of positivecontributions of real importance to social science.

UNIVERSITY F CAMBRIDGECAMBRIDGE, NGLAND

N O T E S

Editorial note: This paper was presented by Professor Fo rtes a t the Symposium on th ePositive C ontributionsof Social Anthropology, held a t the 50th a nnu al meetingsof the Ameri-can Anthropological Association in Chicago, November 15-17,1951. Professor Fortes particip a-tion in the symposium w as made possible by th e generosity of th e W enner-Gren Fou ndatio n forAnthropological Research, Inc.

2 This w as written beforeI saw the discussion between Dr. M urdock an d Professor Fir th onthe limitations of British social anthropology in the October-December 1951 nu m be r Vol. 53,No. 4 t . 1) of the American Anthropologist.

8 In the bibliography th at follows, references marked byan asterisk are cited by permissionof th e auth or.


24/25

46 AMERICAN ANTHROPOLOGIST 155, 19.53

BIBLIOGRAPHY

BATESON, . 1937, Naven.BOIIANNAN,AURA,949, Dahomean Marriage: a revaluation.Africa 19.4.- 1951, A Comparative S t d y of Social D ifferentiation i n Primitive Society. D.Phi1. thesis,

BOHANNAN,. J., 1951, Political and Economic Aspects of Land Tenure and Settlement Patterns

BUSIA,K . A., 1951, The Position of the Chief i n the Modern Political Sys temof Ashanti.DRY,P. D. L., 1950, The Social Structure o fa Llama Village. B.Sc. thesis, Un iversityof Oxford.)*EGGAN,., 1937, Cheyenne and Arapaho K inship Systems, inSocial Organisation of North Ameri-

- 1950, Social Organization o j the Western Pueblos.EVANS-PRITCHARD,. E., 1933-35, Th e Nuer: tribe a nd clan.Sudan Notes and Records Volume

1937, Witchcraft Oracles and Magic among the Azande.

University of Oxford.)*

among the Tiv of Central Niger ia. D.Phi1. thesis, University of Oxford.)*

can Tribes.

XVI, Part 1, Volume XVII, Part 1, Volume XV III, Part 1 .

1940 a), The Nuer.1940 b), Th e Political Systemof the Nuer, in African Political Systems.

1951, Kinship and Marriage among the Nuer.1948, The Divine Kingshipof the Shilluk of the Nilotic Sudan Frazer Lecture.

FIRTH,R., 1936, We the Tikopia.FORDE,C. DARYLL,938, Fission and Accretion in the Pa trilineal Clans of a Semi-B antu Com -

- 1939, Kinship in Umor: Double Unilateral Organization in a Semi-Bantu Society,Ameri-

- 1947, The Anthropological Approach in Social Science, inThe Advancementof Science

195O a), Double Descent amongthe Yako, in Radcliffe-Brown and Forde,1950._ _ - 1950 b), W ard Organisation among the YakoAfrica 20. 4.- 1951, The Yoruba Speaking Peoplesof South-Western Nigeria. Ethnographic Survey of

FORDE, . DARYLLnd G. I. JONES, 1950, Th e Ibo and Ibibio-Speaking Peoples of South E aste rn

FORTES, ., 1945, The Dynamicsof Clanship among the Tallensi.- 1949, The Web of Kinship among the Tallensi.- 1949, Time an d Social Structure: an A shanti Case Study , inSocial Structure: studies pre-

- 1950, Kinship and Marriage among th e Ashanti, in Radcliffe-Brown an d F orde,1950.FORTES,M., and E. E:. EVANS-PRITCHARD940, edit.) African Political Systems.GLUCKMAN,., 1950, Kinship and M arriage among the Lozi of N orthe rn Rhodesia and th e Zulu

--_ 1951, Th e Lozi of Barotseland in N orth W estern Rhodesia, inSeven Tribes of British

HERSKOVICS~M.., 1938, Dahomey.--- 1948, Man and His Works.HUNTER,MONICA,936, Reaction to Conquest.COUGH, . K., 1950, Kinship among the Nayar of the Malabar Coast of India. D. Phil. thesis,

KUPER,HILDA,1947, A n Ajrican Aristocracy.

munity. Journal of the Royal Anthropological Institute Volume 68.

can Anthropol. Volume 41.

Volume IV.

Africa. Pt . IV.

Nigeria, Ethnographic Survey of Africa Western Ajrica Part Z II .

sented t A . R . Radclije-BrownEd. by M. Fortes.

of Na tal, in Radcliffe-Brown an d Forde,1950.

Central Africa edited by E. Colson andM . Gluckman.

University of Cambridge.)*

1950, Kinship among the Swazi, in Radc liffe-Brown and Forde,1950.


25/25

FORTES] S T R U C T U R EOF U N I L I N E A L D ES C EN T G RO UP S 41

KROEBER, . L., 1938, Basic and Secondary Pa tte rn s of Social Structure .Journal of the Royal

LOWIE, R., 1921, Primitive Society.

MALINOWSKI,.,1929, The Sexual Lifeof Savages.

MAINE,SIR HENR Y, 866, Ancient Law.MAYER, ., 1949, The Lineage Principle inGusii Society. International African Institute Memo-

randum XX IV .MITCHELL,. CLYDE,950, Social Organisation of the Yao of Southern Nyasaland. D.Phi1. thesis,

University of Oxford.)*- 1951, The Yao of Southern Nyasaland, inSeven Tribes of British Central Africa edited

by E. Colson and M . G luckman.MURDOCK,. P., 1949, Social Structwe.NADEL,. F., 1942, A Black Byzantium.- 1950, Dual Descent in the Nuba Hills,in Radcliffe-Brown and Forde,1950.

PETERS,E. L., 1951, The Sociologyof the Beduin of Cyrenaica. D.Phi1. thesis, Universityof Ox

RADCLIFFE-BROWN,. R., 1930-31, Social Organisation of Australian Tribes,Oceania 1.- 1935, Patrilineal and Matrilineal Succession.Iowa Law Review Vol. X X .2.I 1950, Introduction t oAfrican Systems of K insh ipa d Marriage.RADCLIFFE-BROWN,. R., and C. DARYLLORDE edit.), 1950, African Systems of Kinsh ip and

RATTRAY, . S., 1929, Ashanti Law and Constitution.RICHARDS, I. 1936, Mother Right in Central Africa, inEssays presented t C. G. Seligman.- 1939, Land Labour and Diet i n Northern Rhodesia.- 1940 a), Bemba Marriage and Modern Economic Conditions,Rhodes-Livingstone

- 1940 b), Th e Political System of the Bem ba, inAfrican Political Systems.- 1950, Some Types of Family Structure among the Central Bantu, in Radcliffe-Brown

RIVERS,W. H. R., 1911, Presidential address,British Association or the Advancementof S c i w

- 1914, Kilzship and Social Organisation.SCHAPERA,. 1940, Married Life i n an African Tribe.- 1950, Kinship and Marriage among th e Tswana, in Radcliffe-Brown and Forde,1950.SELIGMAN,. G . and B. Z. SELIGGWN,932, Pagan Tribes of the Nilotic Sudan.SMITH, E . W., 1935, Africa: wha t do we know of it?Journal of the Royal Anthropological Ins titute

SPOEHR,A., 1947, Changing Kinship Systems.Anthropological Series Chicago Notural H istory

- 1950, Observations on the S tudy of Kinship.American Anthropologist Vol. 52.WARNER,W . L., 1937, A Black Civilization.WEBER,MAX,1947, The Theoryof Social and Economic Organisation ranslated by A. R . Hudson

WILSON,MONICA,950, Nyakyusa Kinship, in Radcliffe-Brown and Forde,1950.1951 a), Nyakyusa Age Villages,Jownal of the Royal Anthropological Institute Vol. 791951 b), Good Company: A Studyof Ny aky wa Age Villages.

Anthropological Institute Volume 68

ford.)*

Marriage.

Institute Papers No. 3

and Forde, 1950.

Section H.

Volume 65.

Mweum Vol. 33, No. 4.

and Talcott Parsons.

Date post:	03-Jun-2018
Category:	Documents
Upload:	hugo-prudente
View:	232 times
Download:	0 times

1_MEYER FORTES unilineal descent.pdf

Documents