The belief in survival after death is not a response of the individual to death. It is accepted as part of his heritage and must therefore be related to mortuary and other social customs. Social anthropologists who explain it entirely by the emotional comfort it gives to the dying person and his survivors 2 avoid many of the important problems involved. Typical statements such as that by believing in his immortality man "never has to come to an impasse and his hope will hurdle death itself," 3 or that in the "conflict and chaos" of death religion "standardises the comforting, the saving belief," 4 are very facile explanations. Unloved parents are buried with full ceremony; strangers have special funerals; people dying away from home may be buried by proxy. The emotions expressed at a funeral may have no relation to the emotions actually felt. Thus Miss Hunter tells us of the Mpondo that "women are not expected to make a fuss and weep after a child is buried ....., but the fact that they are composed does not mean that they do not feel the loss bitterly." 5 Mortuary customs impose on the survivors not only certain modes of behaviour but also the simulation of certain emotions which may or may not be felt. Thus the statements of Wallis and Malinowski may describe the emotional satisfaction derived on occasion from a funeral but they do not advance a sociological understanding of all mortuary customs and beliefs. I propose in the following analysis to indicate what seems to me some of the major problems, chiefly by reference to Junod's account of the Thonga of Portuguese East Africa.

According to the South Eastern Bantu it is either a man's shadow or breath which becomes his ghost after death. This ghost figures in the ritual beliefs and practices of these Bantu. Two other things are also made the subject of ritual. Firstly there is his body which must be disposed of in some way. Secondly there is what Radcliffe-Brown calls his social personality, "the sum of characteristics by which he has an effect upon the social life and therefore on the social sentiments of others. ...... A person occupies a definite position in society, has a certain share in the social life, and is one of the supports of the network of social relations." He goes on to say that "death constitutes a partial destruction of the social cohesion, the normal social life is disorganised, the social equilibrium is disturbed. After the death the society has to organise itself anew and reach a new condition of equilibrium." 6 I do not like the manner in which Radcliffe-Brown has described this readjustment of social relationships, but he makes clear that the mortuary customs as modes of social behaviour are concerned not with the dead man's own individual personality or his personal relations with his survivors. They deal with his social personality, the ideal position which he occupied in the society. The relation of these two personalities is another problem.

I shall later expand on this point. Now I wish to point out that Radcliffe-Brown also loses himself in the emotional interpretation when he says that "we may translate the above statement into terms of personal feeling by saying that the death removes a person who was the object of feelings of affection and attachment on the part of the others and is thus a direct offence against those sentiments in the survivors."

"Though the dead man has ceased to exist as a member of society, it is clear that he has by no means ceased to influence the society. On the contrary, he becomes the source of intense painful emotions." 7 I stress this point again because this type of explanation is frequent to-day. Ritual and mystical beliefs determine the emotional reaction of a particular people to a certain situation and therefore one cannot explain the reaction by the emotion expressed. Radcliffe-Brown's last quoted statement will not cover the rejoicing of the Thonga if a man is killed by lightning or the orgies of mourning which Shaka forced on his subjects after his mother's death.

Radcliffe-Brown has described the survival of the social personality; he has not properly elaborated his thesis. When a man dies his death breaks off a number of social relationships. During is life he was related to a large number of people in different ways and these relationships were determined by certain social rules and expressed in certain modes of behaviour. After his death these relationships are changed and new ones come into being; perhaps also the relationship between his surviving fellows are affected. These vary from one person to another. For different individuals a man to whom was owing a certain social attitude has become a ghost. One person may be the dead man's son, another is a maternal uncle, or a brother, or wife. The the death therefore each of these individuals is required to react differently. It must be remembered that a funeral as recorded by the ethnographer is an abstraction from different people's behaviour. He describes the funerals of headmen, women, children, witch-doctors, etc., always in relation to the dead person. 8 But for purposes of sociological interpretation a further description is required. For a Native life a funeral differs for each kinsman.

In communities such as those of the South-Eastern Bantu 9 where status is determined by consanguineous and affinal relationships every man and woman has his or her rôle in social life fixed by these relationships. On the death of, say, the head of a homestead, the relationships of his cognates and affines are altered. In addition the relationships of his relatives towards each other are affected and are reconstituted in the mortuary customs. For the widows this death means the loss of their husband and they must prepare to start a new life with one of his kinsmen; a brother or son may be succeeding to the headship of the homestead and the other male relatives becoming subordinate to him; his children lose their father, relatives of the widows their link with their affines, his uterine nephews perhaps their special connection with their mothers' kin. A different social situation exists for each of them. The ritual of death ceremonies provides them with a social reaction to this situation adapted to their special relationship to the dead man. Outside of their own ritual behaviour they may see the funeral as a whole: but in the readjustment of social relationships each participant has to observe his own prescribed rules of behaviour. At one funeral these will vary from relative to relative, for one individual they will vary from the funeral of one relative to another.

For a full understanding of mortuary customs we require detailed information on these prescribed rules of behaviour. Unfortunately it is not available. Thus Junod describes the part of various relatives in a long series of rites. He does not trace the connection in their behaviour through the whole series. Of course this behaviours is related to that of others at the same time but it is equally related to their own behaviour in preceding and succeeding ritual situations.

I must make one further point. When a man dies his relationships with his fellows are affected by mystical beliefs about the cause of his death; by the actual manner of his death; and by beliefs about death itself. All these elements combine in the complicated death ceremonies. I propose to consider these as they affect the readjustment and reconstitution of social relationships; the disposal of the body; reaction to the mystical causes and forces of death; the life of the ghost in the afterworld and the links of the afterworld with this world.


On the death of a family headman various kinship relationships are involved. There are his cognatic relationships with brothers and sons, especially as the death of a family head is often followed by the splitting up of the family which he has held together. Then there are his matrilineal relationships, and his relationship with his wives and their kinsmen. Finally one may consider on this plane his relationship with his dead ascendants, the ancestral spirits of whom he is to become one. In death ceremonies the old relationships are expressed and new one formally initiated and the ritual demands the expression of determined emotions and behaviour much of which is related to these changes in social relationships.

It is by no means certain that one could in the field work out coherently the ritual behaviour of each set of kinspeople. Many other ideas are represented in funeral ceremonies. The task is therefore still more difficult on the data available, but I shall set out suggestions which indicate that some such study might profitably be made.

To begin with one notices in all the South-Eastern Bantu tribes that all relatives must be informed of the death and if possible pay their respects to the dead. Those far away are informed by magic. The Thonga, for example, blow medicines in the direction of the absent person; Mrs. Kuper informs me that Swazi in Johannesburg use purifying medicines if they hear of the death of a relative at home, and that a Zulu in Johannesburg told her that he lost his job because he had not been informed of the death of his child and had not therefore had a chance to purify himself. The food of the homestead is taboo to relatives. They shave their heads and undergo strengthening medications. The essence of the funeral is that it brings together all the kin. I have noted that there is a tendency for families to split on the death of patriarchs. At their funerals this is shown in a ceremonial venting of grievances by all the relatives. The rites themselves stress the family unity, largely in sacrifices which affirm the bond of the survivors, the dead and their common ancestors, and in which the new head of the family first officiates as priest. The Thonga pray that they may live in peace; perhaps they wish that the surviving brothers may not quarrel. The death of the headman upsets the collective life of the homestead which is broken down and abandoned. The Thonga think this collective life is represented by marital sexual relations, and these are suspended for a time, when resumed in ritual intercourse. The purpose of this is stated to be "to cleanse the inheritance" which is primarily succession to the headship of the homestead.

Through all the rites run the expression of group unity. A funeral is the ceremony par excellence which all relatives must attend, lest they be suspected of causing the death by witchcraft. It is too the only ceremony when, if individuals are unavoidably absent, they are brought by proxy into the rites. Their belongings are purified for them and when they do return home they must first ritually eat of food before they can enter the homestead. When Thonga huntsmen return to their homes after a long trip to find that someone has died they and their guns are purified. In the rites of allegiance to the ancestors quarrels must be settled. On this occasion the group often reacts violently against the suspected cause of death (perhaps a wizard); in some Bantu tribes the men brandish their weapons against the spirits. Among the Zulu there is a ritual hunt which on the death of a chief may be turned into a war. In other words the unity of the kinship group or the tribe is additionally expressed in opposition to some enemy outside of it: a wizard (who does not probably in fact exist), the evil spirits, or a hostile tribe.

Of the behaviour of brothers and sons we are told so little that all one can infer is that the brothers who supported the dead man when he was alive have to give him their last office as gravediggers, and, where necessary, support the chief son, the heir, in his ritual observances. Part of the ceremonial is directly to institute the heir and exhort him to behave well. The brothers are magically cleansed and purified for, since death is infectious, they are likely to follow the dead. Even their spears are contaminated; they are cleansed with medicines and, among the Zulu, formally used again in a ritual hunt.

Of the behaviour of affines we are only told by Junod that parents-in-law mourn over their daughters, the new widows, but the first prayer at the grave requests that they will assist in the mourning. The ceremonies in which Thonga matrilineal relatives figures are significant. They have no official part in the burial or the great mourning ceremony; only at the end of the latter does one of the uterine nephews appear to cleanse the food of the deceased. That is, though the patrilineal kin cannot touch the food, they are formally allowed to after this nephew, a kinsman but not of the group, has cleansed it. But this rite is extraneous to the ordinary ritual: it is only in the final family rites that the uterine nephews have an important role. When the family unity is being ritually expressed these nephews sacrifice to the dead man. For some reason his first acceptance of a sacrifice is not from his cognates. And these uterine relatives desecrate the sacrifice by interrupting it and stealing the meat, which is in accord with their normal joking relationship with their maternal uncle when he was alive. Then they could snatch his food from him. Perhaps in this final family rite there is a tendency to assert the integrity of the patrilineal group against these friendly outsiders, or to give sublimated expression to any feelings of relief, hostility to the dead man, etc., the patrilineal relatives may feel. But one can add little to Junod's statement that the nephews can do this because they are "the darlings of the gods." Jacques however maintains that Junod is wrong in saying that bantunkulu are the uterine nephews; he says they are the patrilineal grandsons. Thus it would be the privileged descendants of the headman who treated them freely and lovingly who spoil the sacrifice, and perhaps they sacrifice to include the spirit to care for the descendants he loved.

Most of Junod's information is about the ritual behaviour of the widows. Whatever their feelings they are required to lament most and mourn longest. Their actions are the centre of the great mourning ceremony, which commences with their initiation to the status of widows 10 by those previously widowed. In this initiation the left groin of the widow is cut. The blood represents the husband and if it flows freely marital relationships are considered to have been good. The widow is exposed to medicine smoke and with her urine extinguishes the fire. Now, as a widow, for the last time she passes through her husband's hut. For the next five days the widows are treated by the magician and live outside the village. Some time later there are wild and lascivious dances when the widows, "uncovered by the death of their husband," posture lewdly before the men. 11 The widows do not participate in the ordinary sexual rites of cleansing, but each has to seduce a stranger into intercourse and break off the act ante seminem immissum. The man will die. The widow's intimate associations with the husband were through sex, and she has ritually to pass on the contamination of death to the stranger in intercourse. Then after a preliminary medication she can begin life with the man who inherits her.

Meanwhile the dead man is also moving into a new social status. At a Thonga burial he is introduced to the ancestors as "dead", and in the later rites he is himself sacrificed to. Among the Zulu an old man may be escorted on his way, before even he is dying, with a sacrifice; and when he dies there is little ceremonial as he has merely "gone home". For a time in Native parlance the spirit is thought to wander after burial in the veld, and then to appear in the form of a snake at a final ceremony some months later. Thereafter he has power, varying with his status on earth, to help or punish his own classifactory descendants. The ancestral spirits communicate with the living through diviners and dreams and by sending illness; they give fertility or barrenness, pestilence, rain, good crops.

I have attempted to trace through Junod's description the part played in funeral ceremonies by each set of relatives. Perhaps I may briefly indicate the relation of these parts within each Thonga rite. Before a man's death he is concerned not with the fate of his soul but with settling his affairs on earth. Immediately after his death his body is prepared for burial by his brothers while care is taken to inform all his relatives of the death. Sexual intercourse of married people is taboo. Everybody, and all the man's possessions are contaminated by the death: this notion explains much of the behaviour of readjustment. It seems to arise in the feeling that if one member of the family dies another will follow, and the closer the relationship the greater the danger. Not only are the ancestors invoked to help the family but the magician is called in to strengthen it. This seems to indicate that in Native belief death is ascribed almost always to a mystical force, which continues to operate now that it has its hold in the family and which must be ritually countered. After the death of children and old people whose death is taken as natural there are no such ceremonies. The body is buried reverently and the spirit commended to the ancestors.

In the succeeding rite the widows and grave-diggers (brothers?) are purified and strengthened by the magician; the former are initiated into the status of widows, and while the men hide they pass through the funeral hut for the last time. Then the men take the first steps towards the demolition of the hut. The dances, lewd for women, warlike for men, commence. The homestead, its inhabitants and their implements are medicinally cleaned.

Some time afterwards the married couples ritually resume sexual intercourse and impress their sexual secretions in the village square. After another interval the dead man is sacrificed to and his hut is crushed. There is a reunion which all relatives must attend and at which all grievances must be vented. The new headman is exhorted to rule well. The matrilineal relatives come to assist; they sacrifice and steal the sacrificial meat. Only the widows remain contaminated and they cleanse themselves in intercourse with a stranger.

The above analysis may be applied to the variation of death ceremonies with social status. One must note, however, that there is no mean for funeral rites and variation from it; a death creates a different social situation according to the status, or manner of death, of the deceased, and each funeral involves the participation of different persons behaving in prescribed ways.

In the burial of a Thonga child only the mother who alone has had real social relationships with it plays an important part, though the father digs the grave. Only the parents perform the sexual rites of purification: other villagers need not. If the child has been initiated its spirit is consigned to the ancestors. Old people are already so closely associated with death that they are sad to "go home" and their burial rites are few. A woman's death, unless she is the chief wife of a headman, affects only her own family. Strangers and friendless people, having no social relationship with the community, are buried without ceremony. In the descriptions of chiefs' funerals one finds that little attention is paid to their relations with their families. Apparently a chief's importance as head of is tribe overrides family relationships. Great care is taken to maintain peace in the country; otherwise noteworthy are the comparative magnificence of the funeral and the attendance of large numbers of tribesmen.

Warriors killed in battle and people killed by violence are not buried in the village lest they bring death with them. Lightning especially is a dread contaminating force, and the villagers must be cleansed by a special magician. The village is abandoned. There is no mourning lest Heaven which sent the lightning be offended. Certain criminals, such as wizards, have their bodies thrust out into the bush.

But the readjustment of social relationships does not explain the whole reaction to death. Most primitive funeral rites show no patient resignation to the death, but a determined opposition to it, expressed among the South-Eastern Bantu largely by the idea that it must be due to a mystical cause (sorcery or witchcraft). But here the strongest notion is that of the contaminating effect of the death which extends to relatives, material objects and food. Medicinal washings and protective rites, directed by a magician, and appeals to the ancestors and the dead man, are used against this contamination, as well as taboos to be observed for a period. These mourning taboos and reactions are a mode of expressing kinship relationships and are referred to as evidence in legal disputes. Some of the dead person's most personal effects are so closely related to him that they must be destroyed or buried with him. In communities where most deaths are ascribed to mystical cause the idea seems to be that now these forces have a footing in the group steps must be taken to expel them or to protect the survivors. To the Maulu of New Guinea the spirit is always evil and it must be driven away, and perhaps with it the cause of death. The South-Eastern Bantu hope the spirit will be beneficent (thus they appeal to it) and they have to keep it in the ancestral cult; therefore they protect themselves ritually against death, as can be seen in the custom of burying those who die by violence facing away from home, and they may in addition attack the cause of death by punishing the wizard who caused it. This last reaction is not, as among the Venda 12 and Azande, 13 essential: there is no mandate enjoining vengeance, and one cannot tell from the records how often it is done. Mrs. Kuper tells me that according to the Swazi the ancestral spirits may make a man very ill, but they never kill him.


There are, it seems to me, two aspects of survival after death: the journey of the spirit to and its continued existence in another sphere of living, and the relations of the dead with the living. Most people pay attention to both of these aspects. But it also seems that social attention when directed to the second emphatically in the form of an ancestral cult tends to exclude any emphasis on the first. If an ancestral cult may be defined as the belief in the continued inference of ancestral ghosts in the affairs of their living kin and continual ritual behaviour by the latter to the former, 14 I think it will be found in general that people with a well developed cult of tendence on the ghosts have a very inchoate picture of the afterlife, which will depend on individual imagination rather than cultural dogma. A number of problems are suggested by this: firstly, how is it possible to measure the intensity of ritual beliefs and behaviour for purposes of comparative study. Secondly, how frequently does the dead's paramount importance as related to their descendants exclude cultural attention from the afterlife, and how does this affect mourning customs and the disposal of corpses. Thirdly, why do particular societies have an ancestral cult: why, if it is a matter of extended kinship grouping do some societies with that form of grouping not have the cult, and how is the belief in active interference by the ancestors related to the social existence of gods and other spirits.

The first problem must be settled if social anthropology is to progress in the comparative study of ritual. Thus where an ancestral cult is concerned one would require to know how often and in what situations the people thought of and approached their ancestors, and as far as possible how they evaluated the cult in relation to other modes of behaviour, i.e. what satisfaction it gives them. Then the same questions would have to be answered about magic and the belief in gods. In view of my inability to judge these cults from existing data I can do little more than suggest some lines on which enquiry may be made.

How far does the concentrating of cultural attention on the relations of dead ancestors with their living descendants exclude attention from the afterlife? In some societies neither system is at all developed, but reference to others 15 shows that where one is important the other is always inchoate. The two are not found together. Thus Fijians, North American Indians, Egyptians, Romans, Greeks, have an elaborate mythology of the journey to the afterworld and what happens to the ghost there, unlike African tribes which, with their ancestral cult, have left the picturing of the afterlife to the individual - and it is a life more or less like that on earth. 16 The importance of the distinction is, firstly, that it may indicate a general tendency for cultural attention to concentrate on one only of several aspects of a phenomenon, at any rate in some particular situation. The other aspects will only be consciously evoked when some special situation occurs. Secondly it may be correlated with ritual practice. Thus in certain societies where there is no ancestral cult a large part of the mourning ceremonies are concentrated on the dead person and on helping his spirit on the perilous journey to the afterworld. 17 Among these societies the individual's desire to survive death might be of importance and it must be noted that Malinowski studied a society of this type. But the ancestor-worshipping Bantu are only concerned with instituting the dead man safely as an ancestral spirit and they have no mythology of his passage from this life and no developed notion about the soul. The body is carefully disposed of and the chief importance of the grave is that the spirit will be approached at it. Thus headmen of families who become the spirits really approached are buried in the cattle-kraal, the temple of the village, 18 or among the Thonga in the sacred groves. The graves of chiefs remain marked as sacred places, even sanctuaries. The Zulu and Thonga chiefs' hair and nail parings are preserved in sacred tribal symbols and often some attempt is made to mummify the bodies of chiefs. Their graves are more solid and the location of their graves is sought to be preserved by groves of trees and legends, so that the people can often return miles to sacrifice there. In addition their greater social importance is thus expressed. When a village is moved ritual is performed to transport the ancestors to the new site. Generally the form of the grave 19 (among the South-Eastern Bantu it is supposed to represent a hut with a square outside) seems to reflect the idea that after death the individual is not greatly changed. The graves of chiefs are always sacred; other graves only in the situation of sacrifice, otherwise probably only scant attention is paid to them, or to such places as the back of the great hut of the village where the ancestors dwell. Dr. Fortes has remarked that this is so of shrines among the Gold Coast Tallensi, a conclusion borne out by Dr. Evans-Pritchard's observations in the Sudan. The grave and the existence of the spirit are unimportant except when evoked in the ancestral cult: the main purpose of the burial rites is to restore kinship relationships among the survivors and of the dead man to his ancestors and descendants.

The South-Eastern Bantu have a hierarchical society of ancestors manifesting active interest according to their status when alive in their descendants. 20 Mrs. Hoernlé points out that in all the South African tribes one of the strongest bonds holding individuals together is descent from a common ancestor. Their social organisation is built up largely on the basis of kinship for members of the same lineage tend to live in the same village or at least close to one another in the same area. The closest bond naturally is between members of the same family, i.e. between siblings and their descendants; the links between these people are not broken as the years go by, though since kinship is patrilineal and marriage patrilocal it is inevitable that a number of brothers can more easily maintain contact with each other than with their sisters who move off in different directions as they marry. Moreover inheritance is from father to sons or from brother to brother and then to the eldest brother's sons, and therefore a closer link is established in this line. Now many peoples have a kinship organisation of this type, and some of these are linked, in their smaller groupings, by a definite association with a specific area of land and through the possession of hero-gods, totems or badges. This does not apply to any of the smaller social groupings among the South-Eastern Bantu, though the tribe is intimately bound to its land. Therefore among the South-Eastern Bantu the links between living people are maintained by keeping in mind their actual common descent. Bryant says of the Zulu that "the preservation of pedigrees with them was absolutely necessary, because without it mutual relationships, with their appropriate terms of address, would never have been known." 21

This is obvious and easy for the individual family and even for the group descended from a common grandfather but as the ramifications of kinship become more complicated and the group larger "it needs a distinct effort and organisation of a high type to keep the links strong. It is obvious that the common ancestor must be remembered, and it is much easier to remember if there is a definite cult of remembrance, a definite ritual which brings those together who are of common kin so that they are perpetually reminded of their bonds." 22 This is done by sacrifices and commemoration, and it must be remembered that tendence on the ancestors is compulsory. A man may or may not believe in magic, but if the ancestors send him illness he must sacrifice, and there are certain periodic obligations to them which must be fulfilled, at birth, initiation, wedding, and death ceremonies. For a breach of custom they can punish and here their power is the same as that of Bantu elders to curse erring relatives. 23 Since, moreover, it is only these elders who have the right to sacrifice, their social position is strongly entrenched. A junior who has quarrelled with his senior must in the end, when he is in trouble, make amends. All the family should attend, and Blohm remarks that among the Xhosa even Christian converts had perforce to attend sacrifices lest they be expelled from their families. 24

Once the people linked by a common ancestor separate; they forget that ancestor as they lose their ties of communal life, and if they forget that ancestor they tend to drift apart 25 and perhaps to sacrifice to and approach a new founder. "A group therefore will be larger and stronger the longer the line of ancestors it can count, the further it can grope back into the past for its beginnings. Since the people are linked together through their ancestors the dead must not be lost to society. They represent the past life of the group, links binding together the living. The cult of the ancestors then gives life to the kindred which it would not otherwise possess." This may be seen in the manner in which the ancestors and their great deeds are praised at marriages, births, feasts, and funerals. For the dead are held on to and thought merely to be initiated into another part of society, the spiritual and unseen but potent part of it, and the people are perpetually concerned with maintaining contact with this unseen world.

Mrs. Hoernlé thus suggests possible answers to some of the questions I have posed. She points out that the ancestral cult is associated with a system of extended kinship where the kin are organised into a strong social group which tends to split up as it increases. The ancestral cult is a mechanism by which kinship bonds are affirmed (among the Thonga matrilineal as well as patrilineal bonds) and the hierarchy of society expressed. In this the ancestral cult is, like much ritual, a form of mnemonic, legally prescribed actions which vividly express social relationships.

I have spoken above of the patrilineal lineage among the South-Eastern Bantu as a strong social group. My recent researches in Zululand indicate that the Natives themselves do not think of this group, or of a clan, but think in terms of people with the same isibongo (clan-name). That is, we abstract the group. Mrs. Kuper confirms this of the Swazi. The Zulu have no word to describe a clan, as they have for political areas. But the Xhosa and Mpondo have such a word. However, my point here is that the ancestral cult is a mnemonic of kinship relationships, and since among the Thonga the matrilineal kin are important one finds among them that matrilineal ancestors are important.

In many societies there is a cult of the dead which is not a cult of the ancestors, since no tendence is paid to a line of ascendants. In others the spirits of the dead, especially of rulers, are important, but not the spirits of dead kinsmen as such. In modern spiritualism we are afforded a chance to examine the growth of a ghostcult and see whether it is affected by social groupings. As far as I can judge from spiritualist conversations 26 these deal mainly with the fate of those at the seance (the living of course) or the contact of the spirits with the earth, and very little, despite the theory of seven dimensions, with the life of the spirits themselves. And then, it seems to me, it is only by forcing a polite interest that people ask questions about that life, to get the same answer as the Bantu tribesman gives - a world better than, but more or less like, our own. Here however dogma is not yet clearly formulated. The springs of genuine spiritualism are mainly the desire to get into touch with dead relatives, but one cannot do so directly. One must do it through one's own, and the medium's, spirit guides. Of them we each have two or three. At the seances I attended there were two mediums: the first, a Christian woman, was "guided" by three spirits, those of an American missionary, a Hindu, and a Turkish girl who died at the age of two; the second medium, a naturalised South African Jew, had as his guides an American Negro vaudeville singer and a Lithuanian cantor - of a New York synagogue. The guides mentioned by Bradley are equally exotic. The "mediums" of the Thonga and other primitive tribes are also "possessed" by foreign spirits; but the ordinary tribesmen (and Zulu mediums) receive guidance from their dead kin. In spiritualism we are all accompanied by foreigners who will call up our relatives, i.e. where the important social kinship group is the family, the attendant spirits are not kinsmen: and similarly in the Andamans 27 a man sees-not his own dead ancestors, but just ghosts in the forest. It may be that in the growth of spiritualism one can see how social morphology may determine dogma, in that people are guided by stranger-ghosts, unlike the Bantu tribesmen whose kindred-ghosts advise them with dreams and omens, though their wishes may have to be interpreted by a skilled "medium" who acts with certain techniques and occasionally with stranger-ghosts. In an appendix on the ancestor cult Junod makes a similar point in comparing the Thonga and Ila with the Akamba, among whom the largest kinship unit is the family, and the power of the ancestors is "reduced to a minimum; though they are very much feared they are thought to be mortal." 28

Mrs. Hoernlé suggests another line of enquiry when she says that many peoples who have a kinship organisation similar in type to that of the South-Eastern Bantu, but no ancestral cult, maintain their association by reference to a specific area of land, or by possession of hero-gods, totems or badges. The point is that in all these cases ritual is a means of expressing the structure of the group; and a tendency to do this runs through all South-Eastern Bantu ritual. It may be noted that the South-Eastern Bantu lineage relates to a specific area of land only in so far as its ancestors are buried there 29 and when migrating it takes ritual steps to carry its ancestors with it. 30 In many communities the kinship group seems to maintain itself by owing a particular area of land. Thus for example there are Scottish clan demesnes (and names), the Iroquois long houses 31, Algonkian hunting areas, 32 and Hottentot waterholes. 33 Radcliffe-Brown has stressed that one of the main features of totemism is its relation to this type of social grouping, 34 generally with no ancestral cult. The several kinds of ritual mechanism for maintaining cohesion exist side by side in some communities; but one or other tends to be dominant.

But once this rough correlation is established it is not yet clear why any society should have one ritual rather than the other. An explanation of this can probably be offered only in historical terms. However, if other culture traits which accompany each of these rituals be investigated, it seems that the clan association with land is found with one or two sets of environmental factors. Firstly it is found as among the Hottentots where land is plentiful but water limited, and ownership of waterholes is important; and this is also what constitutes the clan bond in Australian tribes where totemism is associated with local totemic spots and reincarnation, rather than with descent groups. Secondly Scottish clans and old English families are associated with land in a country where it is all taken up. The real totemic cult (an active ritual, not mere prohibitions) is found in hunting communities and it may be noted that hunters such as the Andamanese, Bushmen and Algonkian who have no such cult, have no clan grouping. With all these societies one may contrast the South-Eastern Bantu, pastoral-agriculturalists with spacious land and organised in lineages.

I have now to consider the relation of the ancestral cult to other ritual practices. As far as South-Eastern Bantu magic is concerned it is related in that the two practices are used together, and in many instances the power of medicines is increased by the ancestors, from whom knowledge of the medicines was probably obtained. This is particularly noticeable of socially important magic. Other magic may be bought, sold and taught outside the kinship group, but in general in their magic the substance used is of greatest importance: the spell is included in a prayer to the ancestors which gives the magic its social character and which guarantees the right and power of the magician to use it.

There are other spirits and spiritual beings besides ancestral ghosts. Among the South-Eastern Bantu they are of two kinds, what one may call for want of better terms "demons" and "gods". The demons are evil spirits: souls not properly buried (a sanction enforcing observance of mortuary customs); the spirits of living and dead which have fallen prey to wizards; and certain sprites which exist in their own right but may be used by wizards. These demons are held responsible for ill which is not attributed directly to wizards, or to the vengeance of the ancestors among whom female spirits are capriciously evil. In other words these evil spirits and sprites are used to explain certain misfortunes. What misfortunes I cannot on the data available detail, though they are sometimes suspected of causing certain illnesses. But here in the South-Eastern Bantu ritual system the action of wizards, demons and ghosts are interwoven in complicated ways and to which of them attention will be directed by the diviner in any situation it is, on the evidence available, impossible to say.

The South-Eastern Bantu have very incomplete conceptions of "gods". They may be said to have two of these gods, Heaven (representing storms and certain other dread phenomena) and a "First Man" who initiated the life of men. The Zulu have in addition a certain goddess, Nomkubulwana, who figures in hoeculture 35. The First Man (Zulu Unkulunkulu) does not figure at all in ritual practice; he is one of Seligman's "otiose High Gods"; he is merely a final reference in mythology to explain such facts as creation and death. Heaven appears in ritual chiefly as a vast destructive power, more or less personified, in situations of tribal catastrophe such as drought, floods, locust invasions, blight, and among the Thonga in the birth of twins. In ordinary life Heaven is referred to when a man is killed by lightning, suddenly and inexplicably. Thus Heaven is evoked in South-Eastern Bantu consciousness rarely and in overwhelming situations. 36 Usually attention is directed to ancestors and wizards. And to deal with Heaven there are therefore special ritual techniques conducted by magical specialists of high standing.

Viewing the problem comparatively one notices that the ancestral cult is seldom found with a deistic ritual. In Hebrew history the religion of Jahweh had continually to struggle with the tendency to approach the dead. 37 Generally the idea of "High God(s)" seems to be found in simple undifferentiated communities such as the Bushmen, Hottentots, Andamanese and Australians, and in communities with a highly developed centralised organisation, such as the Egyptians, Romans and certain Bantu. 38 In fact in these latter communities the development of the political organisation seems to be accompanied, where there is not already belief in a High God, by the growth of the god. Where the belief exists the head of the state becomes defender or representative of the god. In South-Eastern Bantu ritual the ancestors of the chief are almost gods already; in Hebrew history the development of the state sees the growth of Jahweh; in Roman history the emperor becomes deified. In such communities one tends to find the divine king, as Seligman recorded of the Shilluk. Throughout Africa the political development of the tribe shows this tendency to deify the chief's ancestors and ultimately the chief himself, as might have happened to Shaka if he had not been assassinated. Certainly one may observe it of the Ashanti, Baganda and Rozwi chiefs. 39

To sum up the comparisons as applied to the South-Eastern Bantu, one finds that social morphology affects death customs and the belief in immortality in that under their mode of life and social organisation they have an ancestral cult which is a means of maintaining kinship groupings; it is accompanied by inchoate ideas about the existence of the soul in the living and after death, and the future existence of the ghost is reflected in mortuary customs. The ancestors of the chief are tending to become gods; other gods are on the whole unimportant.

In South-Eastern Bantu society then the ancestral cult is the chief aspect of immorality (in the sense of continued existence for a time after death, not eternal imperishability), and it is related to the lineage; it has little social importance for individuals as compensation for death. Emotional value in that way it may have. However when Manyibane the Thonga lay dying his care was not for his soul but to settle his affairs on earth. Yet death is not conceived to be on the natural order. In no society is that admitted. Death myths express both the desire for life and the recognition of death as a fact. Death is unnatural as the mythical causes of it show for it is introduced to man as the result of negligence or evil, and usually the cause is very minute, e.g. a chameleon's pausing to eat berries, the eating of an apple, a bird's singing. It is always an upsetting of previously existing conditions in which dying had no meaning for man. 40


(I am grateful to Mrs. Hoernlé and Father Bryant for allowing me the use of their manuscripts.)

Go to the main Ancestors in Africa page Go to the main Experience Rich Anthropology page Go to the CSAC Anthropology Pages - online