Maurice Bloch and Jonathan Parry Lienhardt, G. 1961. Divinity and experience: the religion of the Dinka. Oxford: Oxford University Press. Malinowski, B. 1948. 'Baloma; the spirits of dead', in Magic, science and religion. London: Faber and West. Morin, E. 1970. UHomme et la mort. Paris: Edition du Seuil. Ngubane, H. 1976. 'Some notions of "purity" and "impurity" among the Zulu', Africa, 46:274-83. Parry, J. P. 1980. 'Ghosts, greed and sin: the occupational identity of the Benares funeral priests', Man (n.s.) 15(1):88-111. 1981. 'Death and cosmogony in Kashi', Contributions to Indian Sociology, 15:337-65. Paulme, D. 1940. Organisation sociale des Dogons. (institut de droit compare: Etudes de sociologie et ďethnologie juridique, 32). Pina-Cabral, J. de 1980. 'Cults of death in Northwestern Portugal', fouurnal of the Anthropological Society of Oxford, 9(1):1-14. Raglan, Lord 1945. Death and rebirth. London: Watts & Co. Thomas, L-V. 1975. Anthropologie de la mort. Paris: Payot. Van Gennep, A. 1909. Les rites de passage. Paris: Emile Nourry. Wilson, G. 1939. 'Nyakyusa conventions of burial', Bantu Studies 13:1-31. 44 2 The dead akd the devils among the Bolivian Laymi OLIVIA HARRIS Introduction ... the dead go directly to Puquinapampa and Corapona. There they meet together and it is said that there they enjoy much feasting and conversation between the dead men and the dead women; and that whenn they leave there they go to another place where they endure much work, hunger, thirst and v cold, and when it is hot the heat is too great; and thus they bury them with their > food and drink. And they always take care to send them provisions to eat andi drink; and after six months they make another similar feast for the dead, and after a year another; but they do not take out the said deceased in a procession as they do in Chinchaysuyu, they leave him inside his cave and underground chamber and they call the town of the dead amayan marcapa (town of the ghosts). (Waman Puma, 1613 (1936:294) Waman Puma here relates the burial customs of Qullasuyu, which included what is today highland Bolivia, and formed the Southeastern quarter of the fourfold Inka state. His description, though separated by 350 years from the present day, offers illumination for the account that follows of mortuary rites in the Laymi ethnic group of northern Potosi, Bolivia. The Andean writer expresses a poignant1 contrast between the time of feasting and 'conversation' and the v subsequent hard labour and suffering experienced by the dead. While, he does not connect this polarity directly with calendrical rites, he does/ note that ceremonies for the dead are performed twice a year; thei divided experience of the deceased is echoed in a temporal arrangement that divides the year into two halves. Today too the Laymi divide their year into two contrasted halves, each marked by a feast of the dead; one half is a time for sorrow and hard work, while the other is dedicated to feasting, pleasure and rest from their labours. In other respects what we know of the mortuary practices of pristine Andean cultures stands in stark opposition to the present; the break was effected historically through the 'extirpation of idolatrous practices', the veritable military campaign waged against autochthonous Andean religion by the Christian priests and their heirs in the first century of European imperial rule.1 Many of the sixteenth-century 45 Olivia Harris Spanish 'chroniclers' give some account of the Inka practice ofu/ embalming, and describe the ceremonies for the royal mummies who / retained their own lands and servants, were fed and clothed daily, and/ were publicly honoured with periodic ceremonies. Other sources give '/ comparable information for the local level; each ethnic group apparently venerated its own ancestors, or at least those of the lords, often embalmed or in the form of an 'idol', and held sacred the tombs or caves where the relics and images were preserved.2 These ranked high' in the indigenous pantheon, and bestowed strength, good fortune and/ prosperity on those who worshipped them. Particularly significant for/ the context of this book is the information that mummies of the ancestors were identified closely with the land they once had worked; and with its continued fertility.3 • A priest of the Archbishopric of Lima in the early years of the seventeenth century wrote a manual on how to discover and eliminate idolatry; he was particularly concerned with the discovery that the u Andean population was disinterring the dead and taking them back to ,, their traditional resting places in the fields: ^ They pour chicha discreetly into the tomb so that the dead may drink, and they make a show of doing him honour, placing cooked meals and roasts upon the grave for him to eat. For this reason, it has been forbidden them to place anything of this sort upon a grave on All Saints Day. Their greatest abuse is to disinter the dead and remove their bodies to their machays, or burial places of their ancestors in the fields. In some localities they call these zamay, which means tomb of rest. At the time of death they cry out: 'Zamarcam', that is, 'Requieviť. On being asked why they do this, they say that this is cuyaspa, for the love they bear them. They say the dead lying in the/ church are in great torment and bound to the earth, whereas in the fields,/ because they are in the open air and not buried, they have more rest. (Arriaga,/ 1621 (1968):56) '* Arriaga's response to these pagan practices was a catalogue of subtle manoeuvres to outwit the Indians: in the last instance mummies were to be assembled in the village plaza and publicly burnt. The ashes should then be thrown in the river and thus obliterated from memory and experience. While practices for disposing of the dead differed throughout the1/ Andes,4 it seems clear that they were worshipped in a material form,, and that the images and relics of the dead lords enshrined memory/ wisdom, power and fertility. Small wonder then that so much/ missionary energy was directed towards this fundamental source of social existence. While I am here concerned with the present-day, attitudes towards death must be viewed through the optic of the past-both what is remembered and what is forgotten. Laymi beliefs about 46 The dead and devils among the Bolivian Laymi the dead today illuminate ttistorical knowledge, and are rendered z more intelligible through appreciating the extent of the catastrophe suffered by Andean cultures in the sixteenth century and by trying to [/ chart the complexity of the Catholicism which they espouse today. •/ The Laymi ethnic group, land and the State Laymi are Aymara speakers, though most also speak a second Andean language - Quechua. Their territory is divided into two radically contrasted ecological zones, the highland suni (11000-15000 ft) and temperate intermontane valleys known as likina (7000-10000 ft). The ethnic group as a whole numbers some 8000, of which about two-thirds lives in the treeless highlands, and the rest in the valleys. The staples of the suni are indigenous tubers, including a wealth of potato varieties, and Old World cereals; most households own flocks of sheep and llamas. In the likina (at a distance of several days' journey), a radically different ecology produces maize and squashes as staples, and goats and some sheep are raised. There is a sharp contrast' and a consciously-articulated sense of difference between these two/ 'tiers' of Laymi territory; nonetheless through the various historical V experiences of political and administrative separation the Laymi have / remained until the present a single integrated ethnic group, proclaim--' ing through the internal circulation of produce, their overwhelmingv endogamy, their intricate and highly-distinctive aesthetic in textile production, an identity that transcends separation and opposition.5 Any traditional agrarian society must be oriented towards the past. Land, the source of life, has been cleared, cultivated, improved and handed on from one generation to the next. For the Laymi, relations? with previous generations are represented through rights to land and inheritance of houses. Inheritance of fixed property is traced through^ agnatic links. Men inherit from their fathers and remember the»7 forebears through whose hands their land has passed. Usually the furthest back they recall is the 'original' landholder whose titles were recorded in the last general cadastre of land rights carried out in 1884.6 Agnatic lines of named forebears go back three or at most four generations where fixed property is involved. There is thus an agnatic ' bias in the returning of the dead; nonetheless the pattern of memoryv always includes four lines for any individual: father's, father'sv mother's, mother's father's, and mother's mother's. These cognatiô ancestors are remembered through the pouring of libations on many ritual occasions. In some rituals the emphasis is on the personal forebears of the individuals who have organised the ritual, for example 47 Olivia Harris in house-building ceremonies; on other occasions however all the dead of the locality (wisinu alma) are remembered as a group, for example feasts which confirm rights of access to community land. The important place given in ritual libations to the dead of the locality reflects the significance of the local land-group. As we shall see, all residents of a hamlet are implicated in the death of one of the vicinity, and celebrating the dead is the responsibility both of the kin1 and of the hamlet as a whole. But the contrast between the kin-group1 and the locality is not a sharp one in a context where most people marry close to home and where cognatic links are much emphasised. The ethnic group is overwhelmingly endogamous, and people circulate as spouses and as a source of labour throughout Laymi territory. While there are clear agnatic groups who own titles to, and work particular parcels, their identity as unique groups in opposition to other similar groups is diffused by the importance of the local land-group which usually embraces families from a variety of agnatic groups, as well as families with no direct access to land, who work land loaned by members of the vicinity. As such, demographic accumulation of particular groups in opposition to others is little emphasised. In Laymi culture land is paramount; humans must serve the land both directly by cultivating it and through worship of the telluric spirits. The dead too, as we shall see, direct their energies towards agriculture and animal husbandry. Large families are frowned on by1 Laymi because they upset the balance between humans and land, v When ritual libations are poured to celebrate natural increase, human fertility is never singled out as a primary value. Rather it is metaphorically transposed into the key of vegetation: the most common symboL of fertility is that of flowers. One ethnographer of the Titicaca region some 200 km to the north writing 35 years ago, spoke of the 'strong ancestralism' and geronto-cratic organisation of Aymara culture (La Barre, 1948:135).' But in Laymi society today there is little concentration of authority whetherv gerontocratic or not. Within the agnatic group authority is highly7 diffuse; inheritance is by ultimogeniture, and while land is mainly passed down agnatically, women too own property. Sexual parallelism and complementarity is fundamental to the way that authority is represented, such that it is multiple and shifting, rather than focused on a particular figure.8 Within the land-holding communities, all traditional political offices, both the local headman and the dual authorities for the ethnic group as a whole (segundas mayores), rotate annually amongst adult male landholders. Offices in the national peasant union, established in the aftermath of the 1953 48 The dead and devils among the Bolivian Laymi Agrarian Reform, do not rotate but are for the Laymi permanently elected positions; but the centres of power lie outside the ethnic group, controlled through the administrative system of the Bolivian state, careful surveillance by the army and manipulation of the union. The Laymi of today have forgotten the times when they embalmed their dead to negate the corrosive effects of time on the decomposition • of human flesh and memory. Control over the dead was a cornerstone y of Christian policy for shedding the light of true religion in the lives of ■ the heathens. The village church became the only acceptable place of1 burial; part of a wider colonial policy to concentrate the indigenous population into large villages (reducciones) in order to facilitate the administration of the new subject peoples of the Empire. The ancestors*., no longer exist in material form to which cult can be made. Nonethe-v less some trace lives on, particularly in the pre-Columbian burial chambers known as chullpas, found widely throughout the altiplano, and dating from the period preceding Inka hegemony (Hyslop, 1976). As Cieza de Leon commented in the early years of Spanish rule: The things which, to my mind, are most worthy of notice in the Collao, are the tombs oř the dead ... I was truly astonished to see how little they cared for having large and handsome houses for the living, while they bestowed so much care on the tombs where the dead were interred. (Cieza de Leon, 1553 (1947)) Near the Laymi village of Qalaqala there is a huge ruined chullpa whose adobe walls tower above the single-storey houses nearby. In other areas, bones are used for divination and agricultural magic (La Barre, 1948:137). Laymi chullpas appear to have been sacked of all their relics, but they are still places of ambiguous power. The dead today are * viewed with a repugnance which is never fully transcended. The t absence of monuments or shrines to the dead is perhaps an indication'»'' of their ambiguity. Just as the Laymi are subject to the workings of a y state over which they have little control, they are unable to harness l/ fully the potential power of the dead for the reproduction of their own < society. The process of death The dying are watched over by their kin and neighbours, who chew coca leaf and smoke cigarettes constantly as a prophylactic against the devils afflicting the sick person. When the person's soul has finally left>-" the body, a three-stranded string is spun counter-clockwise (the^ spindle is normally twisted clockwise) with which to tie the neck, hands and feet of the deceased and thus to prevent the ghost from -f 49 Olivia Harris escaping. In the Titicaca region the corpse is actually strangled, ostensibly to prevent the stench from escaping (Carter, 1968; Buechler, 1980). The Laymi, too, from the moment of death refer continually to the stinking state of the corpse, using always the same term of disgust/ (wali th"usqa) to a degree which suggests that it is a categorical rather^ than a descriptive statement. The kin then cover the corpse and lay it v on blankets; on top of it a hastily-made cross is placed, and some good-quality cloth, a hat and a bag for coca leaf. A candle, made from animal fat, is lit by the head and some piece of clothing closely associated with the death is thrown up on the roof. At the same time, the family go to notify close bilateral kin and ritual kin in the neighbourhood and these come to assist the bereaved. As soon as the corpse has been hastily prepared a grand mobilisation of* hamlet resources must be organised. First of all, quantities of coca leaf v must be acquired in order to protect all who come to mourn and keep vigil from the malignant ghost. In Laymi territory there are virtually no shops, thus amassing sufficient coca leaf usually involves visits to several households. In the suni it is by now common custom to drink cane alcohol through the vigil and the wake; this again must be acquired rapidly from whoever has some stored. Next the bereaved must go round the hamlet soliciting help with an offering of coca leaf. Men must be persuaded to go in search of firewood, women to bring í water; quantities of food must be assembled, an animal to slaughter is ^ chosen or if necessary bought, a man and woman must be persuaded to organise the preparation and distribution of food, drink and coca leaf, another man requested to dig the grave, and older men or women to wash the corpse. Once these basic provisions have been made, the close con-/ sanguines return to stay beside the body of the deceased until it is prepared for burial. They must keep a vigil for at least one night,/ chewing coca leaf, and if possible drinking cane alcohol, with constant offerings and libations to all deities, spirits, and souls of dead people.-The following day a funeral feast is prepared for the whole hamlet and ^ anybody else in the vicinity. To stay away deliberately from the feast/ endangers the whole community; on the other hand, little attempt isi made to notify kin in other communities, and it will be a matter of luck, whether they hear the news in time to attend. After the feast, of which the deceased takes an honoured - if token -share, the body is washed and clothed in a white habit and hood that have been hastily sewn together from cotton flour-sacks. On its feet are placed rough sandals made from llama hide, worn by all Laymi until the recent introduction of the more durable and now ubiquitous 50 The dead and devils among the Bolivian Laymi rubber-tyre footwear. Men, and sometimes women past childbearing age, not close consanguines, perform these tasks while women prepare a bundle for the ghost to take on its long journey to the land of the dead; it contains food, coca leaf and a few household and personal articles, and also money so that the ghost can buy food and a house when it arrives at its destination. During these final preparations young men play games in order to build a house for the soul's future life. One, played with sheep knuckle bones, resembles dice, while the other is a form of bowls played with large stones. While I found it impossible to grasp the principles of scoring, the games were noncompetitive and appeared to consist of reaching a round number and starting again. They provided a means of harnessing the energy of the young men to work for the future wellbeing of the deceased. The. occurrence of games of chance at funeral ceremonies has been noted by* various writers; their significance may lie in the repetitive elaborations of arbitrariness (for example in how the knuckle bone falls). In this way v culture is able to play with and reorder the unanswerable arbitrariness < of death.8 ' In the highlands, graveyards lie outside the nucleated hamlets, enclosed by high adobe walls and the single entrance blocked up with large stones. Where there is a burial ground there is often also within the settlement a small church or chapel, and the funeral party stops in the place of Christian worship to chew coca leaf, pour alcoholic libations and ask the protection of gods and saints for the soul now setting out on its long journey. In the graveyard a man not closely related to the deceased, again fortified by coca leaf and alcohol, has dug a grave three or four feet deep. Since graveyards are small and the digger is bound to unearth old bones, but those whose memory lives on must rest undisturbed; only those whose mortal remains bear no name can be viewed dispassionately and shoved aside to make way for newcomers. The corpse is carried on a blanket strung between two poles by teams of men, who run from the funeral house to the graveyard, stopping only at ritually-appointed places, one yoke replacing another as they run. On this final earthly journey the ghost desires urgently to escape;i by running with the corpse humans hope to outwit its intentions. In '' the highlands the nearest graveyard may lie some three or four hours'« walk from more distant hamlets; in the temperate lowland, distance is v" compounded with steep ascent, since graveyards in this lower zone r are located on mountain tops, unfenced-in but above the level of cultivation or habitation.10 To transport the corpse running up ther mountainside, after two days of mourning and a night of sleepless1 51 Olivia Harris vigil, would in itself be a defiance of mortality. The allusion seems however to be more precise; the teams of men are called by the Laymi by the same term {yunta) as is used of the yoke of bulls who open the earth for sowing. The semantic link is found most clearly in a ritual thatv takes place at the sowing season to celebrate the Pachamama (the Andean earth mother, in whom are incarnated both space and time), when pairs of young men are yoked to the plough and driven across the fields opening the first furrow; they are bulls, and like bulls they are strong and uncontrollable. The wilder they are, the harder they run; the more efficacious is the ritual, the more glorious their show of strength. Not only their strength, but also their paired duality under the plough makes of bulls a primary expression of the integral bond between humans and earth. The way that the corpse is taken to be placed in the newly-opened earth is thus explicitly reminiscent of a ritual which embodies the critical act of cultivation. This interpretation gains support from a comparable procedure in the Qullawaya region to the north. In the precipitous landscape of Kaata, Old World bulls and ploughs have not replaced the native Andean foot plough; there the manner in which the grave itself is dug, using the foot plough, bears striking resemblance to the way that fields are tilled and potatoes planted: in the words of the ethnographer 'the grave-digging was a relay race; they plowed and shoveled in pairs to the point of exhaustion and then were replaced by another pair' (Bastien, 1978:125). All those who accompany the corpse enter the graveyard for a final moment to wail and offer libations and coca leaf, and then household members, women of childbearing age and any children present must1 leave as three foot of earth is heaped over the corpse, topped by thorns i and a cross. In this final moment of dispatch the ghost tries desperately to seize another person for company; women, children and close kin v are the most vulnerable. After the burial more food is distributed to all who took part and they then return home. Immediate kin and household members remain in heavy mourning for eight days,/ wearing black, and abstaining from salt and chilli pepper in their food. The end of this first period of mourning is marked by washing the clothes and other possessions of the dead; libations are poured and the washers build in the stream-bed little houses and gardens for the deceased to inhabit and cultivate in the land of the dead. The mourners/ may cease to wear black; the dead person's property is allocated now>/ though it is not yet distributed. /' The participants in Laymi mortuary rites are first and foremost/ immediate kin and household members. After making the initiaK preparations they remain beside the corpse until the moment of burial./ 52 The dead and devils among the Bolivian Laymi The active tasks on the other band are the responsibility of those whose »„/ relationship with the deceased is less direct: affines and ritual kin play <-an important part, since they are less vulnerable to attack. The role of ^ in-marrying women, or men, is often ambivalent. Young in-marrying spouses appeared to stay apart from the central mourning group, while older people married away from their genealogical kin, more fully incorporated, are likely to join completely in mourning for their affines. The man who distributes coca leaf and sometimes alcohol to mourners is also responsible for urging them to weep and display their sense of loss; he himself is less likely to mourn. Close female kin! usually lead in wailing and keening, but men too show great emotion' and behave like the women. This is especially evident when alcohol is flowing; however the parallel behaviour of women and men in mourning is not simply the result of intoxication, but conforms to the overwhelming parallelism and complementarity between the sexes typical of Andean culture as a whole (Harris, 1978ö). In conformity-, with Laymi ritual practice overall, the participation of both sexes is í obligatory, but at any one time, there will be far more men than women £* in attendance. Few women join the vigil, and during the day too mosts women remain outside the house of mourning, helping in preparation * of the funeral feast, while the men sit beside the corpse itself. When the ¥' deceased was a woman, there is likely to be greater involvement of women as mourners, but in this case too it is men and postmenopausal women who mediate the dangerous world of ghosts because of women's great vulnerability to them.11 This vulnerability is stressed in various contexts. The implications are clear. In Andean culture women do not mediate between the dead and the living, representing simultaneously all bodily functions whether of birth or decay (unlike for example the account of Watson, this volume). Rather ■.-' the dead are represented here as antithetical to the living, and women' of childbearing age are correspondingly the category of people who t must be most rigorously separated from the ambiguous activities of the -dead (Harris, 1980). From the many rituals associated with the period after death, it is clear that the dead are a danger to the living; the spirit must be restrained in its own body (by tying the neck, arms and feet after death/ the use of thorns to prevent it escaping, and running to the graveyard), * and the living must protect themselves from it (through constant ► chewing of coca leaf, wailing, leaving the grave prior to burial, putting ;/ up thorns and knives to keep it from the house, taking care not to sleep í-or walk alone at night, and avoiding the graveyard at all times, ^ especially at night which is the time of ghosts). The dead are 'envious';" 53 Olivia Harris if they are not given their due whether of food, animal sacrifice, coca leaf, or display of sorrow from their kin they will haunt the abode of the living. Thus the major concern is to get rid of the ghost, and much of the ritual is performed with this aim. The corpse itself is polluting: we have already mentioned the frequent references to its fetid state, and all those who go in to pay their final respects, particularly those who actually prepare the corpse, must wash themselves in water in which cleansing herbs (ismillu) have been boiled. After the corpse has been taken away for burial the whole house is swept with a broom made of the same herbs, and then more of the water is sprinkled throughout. The sweepings, like all ritually-dangerous substances, must be taken to the river or stream for the devils to eat. But the ghost is from the start a more substantial threat than the ^ body; in death as in life, bodily functions scarcely pollute. Today Laymi v representation of death exemplifies the common duality that opposes what disintegrates to what survives. In previous epochs the preservedl mummies were objects of worship and power; today only the spirits of the dead survive, while respect for mortal remains is contingent on -human memory. However, while Christian teaching must have reinforced a dualistic understanding of death in the separation of body V and soul, it did not take full control of the spirits of the dead. As we shall see, Laymi beliefs about the afterlife imply either that they were given to understand that heaven was not for the likes of them, or more probably that the dead had other responsibilities which outweighed possible elysian pleasures. Feasting the dead and the devils While immediate mourning may be brought to a close after eight days, the deceased is still close to the world he or she has left behind. The v central moment of separation comes in the annual celebration of the V feast of All Saints on November 1st, followed by All Souls on the 2nd. v The month of November was dedicated to the dead also in the Inka calendar (Valcarcel, 1948:474), and there is evident continuity in the extraordinary complex of rituals practised today. The souls of the dead arrive the night before All Saints, and those >/ households where someone has died in the course of the previous year i must hold a feast for the soul. Previously they have brewed chicha and baked quantities of bread, much of it shaped as 'babies' (wawa in Aymara and Quechua) in the form of humans, animals, celestial and mythological beings.12 Wild flowers, heralding the spring and the onset of rain, are placed on the roof to welcome the dead {wayllura); a V 54 The dead and devils among the Bolivian Laymi llama is slaughtered in its honour; a table is prepared, typical of all offerings to telluric spirits and similar to the one at the wake itself, except that now a greater variety of luxury foods is offered to the spirits Close kin keep vigil through the night pouring libations for it and all" spirits of the dead, remembered and forgotten, and for all celestial2 beings, spirits and guardians of the community. The vigil is brought to a close with a ritual breakfast of maize porridge (qalapari). On the day of v All Saints itself, the bereaved family offers festive food to the entire" community. Where there is more than one soul to be feasted it is sol organised that everybody can go in succession from one feast to the / next. It is important that all should attend and celebrate together. After ■-' the feast, the clothes and personal property of the deceased are again ritually washed. In addition to the washing itself, all participants and the objects themselves must cross over the stream. The water signals ■>-an effective separation from the deceased; souls cannot cross water „. unaided.13 The property can then be distributed, and any old or worn i items will be burnt. The day of All Saints itself is a transition from the celebration of an individual ghost to a more collective ritual. Members of all the communities who use a particular graveyard converge on it. Graves of ť those who have died in the previous two years are covered by stepped « altars 7-8 feet high, known as escaleras (Sp. = ladder) on which thei souls will supposedly ascend into heaven. Their passage is assisted by1 prayers of all who go to the graveyard; those who can make long recitations in Latin are much in demand, but even a short prayer in Spanish is welcomed by those who preside over the 'ladders' of their dead kin. To this structure have been tied as many offerings as can bei crammed on: bread icons of the sun and moon at the top, together with»' a Bolivian five peso note, and below them all forms of delicacy - bread» babies, fruit, specially-prepared maize, chilli peppers, dried meat,«' coca leaf, cigarettes. All who pray are 'paid' with food, chicha, alcohol;» coca leaf, and after a few hours emotions have risen to a crescendo. When I attended this feast in the highlands, a few weeks after I arrived, the chaos was indescribable; it poured with rain and the walled-in graveyard was simply not large enough to contain all who had come to mourn and pray. All around people were wailing drunkenly and reciting prayers which grew shorter and shorter as the afternoon wore on, embracing each other and calling on the dead. The bread and food they earned was quickly hidden away in the folds of their clothes, since it could not be consumed until the feast of the dead was completed. When night fell more food was handed round, and drinking, singing and dancing continued in and around the grave- *•' 55 Olivia Harris yard until the following day. When the ladders had been dismantled and the graveyard unceremoniously abandoned, feasting and drinking continued for another day or two in the households of the deceased, and then the ghost was ritually despatched. On such occasions the family once again provides food for its journey and with this parting gift they shoo it away from their home. However even this somewhat dismissive act is recast as an offering from the earth to the participants. As they return they collect up bits and pieces - on the occasion I witnessed we gathered maize stalks, sheep dung and stones - which are distributed, as the property of the deceased, to enrich those who have paid their dues. I myself received a 'cow' and a 'grinding stone'. The timing of an individual death is arbitrary. In the feast of All i Saints, death is tamed, becomes cyclical, is transformed into a ritual in which all join together; after this celebration the bereaved may take off their mourning clothes. However, while all attend, the festival also-emphasises social divisions. Mourning of the dead is organised by \s their kin, who are united as a group by their need to propitiate the dead. Those who have prepared the ritual tables and 'ladders' use the opportunity to ask for prayers for all their other dead kin, thus reiterating the particularity of their own genealogical position. Again, the very disposition of the graveyard emphasises social distinction within the ethnic group; graves are arranged according to / group affiliation, both by moiety and sub-divisions within each-moiety. Each part of the graveyard is associated with a local section of these sub-groups. The division is not merely a spatial one; at the levels of moieties it finds expression in the ritual battle (tinku) which is fought at All Saints as at most feasts throughout the year. The composition of the units ranged against each other in a tinku differs according to the feast and who is present, but Laymi fighting at All Saints is particularly associated with the opposition of moieties within the ethnic group. These moieties are not primarily marriage classes, rather affiliation derives from landholding; Piatt argues for the neighbouring Macha that the ritual battles are in part assertions of rights to land, and that the blood spilt is necessary for the earth to bring forth the harvest (Piatt, n.d.). All Saints in the Andes is a spring festival, marking the time forv sowing and planting, and the start of the rains. The initiation of the'7 agricultural cycle is, then, marked by the 'socialisation' of the graveyard which at all other times is a place avoided with fear and / repugnance. In the attention paid to the souls of the dead, there is a':-" movement from obligatory mourning, to release from the ghosts and V 56 The dead and devils among the Bolivian Laymi festive celebration. In the calendrical cycle however, All Saints is a -s" moment of transition away from festivity since the season of rains, the v period of growth and maturation, is marked by taboos and restraint," and is a time of sadness. During this period contact with water and^ washing is restricted; small round mirrors, the festive decoration of the*"' young and unmarried, must not be worn. Above all, All Saints initiates * the time of waynus, melodies played on wooden flutes whose explicitly mournful tones pervade the whole season and attract rain. The end of the rains is celebrated at Carnival (February/March)* which is the most important feast of the Laymi year, singled out from all others, and marking both the First Fruits and the New Year. Its *-name in Aymara is Anata, the time of play; it is also known as the feast of the devils. Celebrations continue for up to ten days and embrace a wealth of different rituals. As the festival of First Fruits it is a time for <--rites of increase; armf uls of wild flowers are collected and placed on the' houses, the ritual altars, in the animal corrals and in the ritual bundles carried by individual sponsors of the feast. Plants from the maturing c crops are stolen from the fields and carried triumphantly as offerings ; for those who provide chicha and food for the festival. Special rites are; held for the flocks to encourage their continued increase. Carnival is a most dramatic proclamation of community. It is they' only feast of the year in which fighting is forbidden, and ritual battles are never fought.14 A highlight of the feast is the visits paid by people of all ages to villages and hamlets within, and sometimes beyond the confines of the ethnic group. It is the feast in which women enter most fully into the celebrations; on other occasions married women are often prevented from joining in the dancing and singing, but at Carnival they too stay up all night, and go visiting. Young people will make sure to visit other communities where girls from their own hamlet have married, thus reuniting briefly those whose lives have been separated by virilocal residence. Wherever they visit they offer plants from the fields, and sing and dance; in return they are given food, drink and coca leaf. These visits are marked vividly by the display of cloth and clothing. -J Girls and young men take their entire wardrobes and more, hanging what they cannot actually wear from cloths or ropes strung across their backs. This is personal display, but it also has the effect of reducing individuals to virtual anonymity beneath the weight of cloth. The anonymity is complete for those who wear what is known as sintapulla, cow-hide helmets worn by warriors and entirely covered over with brightly-coloured ribbons (sinta) which hang almost to the ground. It is virtually impossible to identify the wearer, and even sexual dis- 57 Olivia Harris tinctions are blurred. There is a deliberate creation of mystery, an attempt to avoid being recognised as long as possible.15 Another form in which individuality is suppressed is the shared " vulnerability of all to the devils who throng the world at this time.v During Carnival nobody should walk alone, whether by night or day, *' for they will surely be led astray or carried off by the devils. It isv virtually impossible to be alone during the week of major celebration. Even sleep is more or less abandoned, and certainly nobody would sleep on their own. The end of Carnival is signalled by the ritual dispatch of the male and female devils. Men and women in each hamlet impersonate them, wearing black goatskins over their clothes, and festooned with wild flowers and plants from the fields. The male(s) wear a cow-hide helmet, now divested of ribbons. The devils dance through the entire hamlet, visiting in turn each house; in each they are offered alcoholic drink. They are then accompanied out of inhabited space to a flat place16 where their disguises are torn off and the 'devils' of Carnival thus sent packing. In a dramatic finale the flutes that have provided music throughout the entire season of rains are piled up together. Immediately the young men take up their charangos (small mandolins) and break into a radically different musical style (kirki) to which everybody dances round with an abandonment rarely seen on other occasions. Only through this final rite known as Tapakayu was I able to discover the identity of the Carnival devils. I had been recording the f wayňu melodies of the flutes, and later that evening began to play the! music back. I was immediately stopped by a horrified audience: wayňu, music belonged to the devils who were now safely dispatched on their way to the land of the dead, and would be drawn back if they heard the \s sound of the flutes. It thus emerged that the devils whose feast is \/ Carnival are in some form the spirits of the dead. At this season they^ are personified as a source of danger - they make people ill, lead therm/' astray and to their deaths if they remain alone instead of joining the collective celebration. But they are also celebrated as the abundance of natural increase, and are festooned with the wild and domesticated plant life they have helped to grow. It is they who make the crops flourish and reproduce the flocks. This association of danger and fertility is made in the person of spirits who are no longer individual, named ghosts but have a new identity. Nobody ever directly explained '-' that these devils were the ancestors. Only through music did it become v clear that the dead remain in the world of the living throughout» the season of rains. Having understood their transformed presence,l/ 58 The dead and devils among the Bolivian Laymi other symbolic statements became clear. For example in the temperate <•*' valleys glow-worms appear with the rains at All Saints and die at harvest time; their light in the night is that of souls, and makes known the presence of the dead in the world of the living at this season. Round mirrors, which are taboo as decoration throughout the rains, are used in neighbouring Macha to keep away the souls of the dead, for example placed at the entrance to the graveyard to prevent a death (Piatt, 1978:1097). According to the Macha, mirrors are 'the enemy of the soul of the dead'. When the dead are in residence, then, there must be no hint of wishing to have them leave. Other prohibitions express the1-same intent; for example it is considered harmful to the growing cropsu to bathe oneself during the rainy season. We have already noted that» water is antithetical to the souls. It is also prohibited to touch the w ancient chullpa burial chambers through the rainy season, or the dead «■ will send lightning or hail to destroy the crops." In the moment of death itself, explicit reference is made to the e process of cultivation; at All Saints, the deceased are embraced within y the agricultural cycle, and the process of bodily decay is virtually-/ ignored in favour of that of natural increase. The feasting of individual" ghosts at that time can be fully understood only in relation and in -' contrast to the celebration of Carnival, at which the cycle initiated at All>s/ Saints is brought to its completion.18 The integral relationship of theses two feasts was clear from the way that Laymi people talked about them. I was continually asked whether the two feasts were kept also in Tnkiltira' (England), while nobody expressed much curiosity as to what other festivals the English might celebrate. I was told that All Saints and Carnival were muntu intim (Sp. = mundo entern) i.e. that they were celebrated worldwide. Thus these feasts for the Laymi are universal: they unify and transcend social boundaries, while others-feasts are specific to particular groups or categories, and are thus forms i, of individuation and differentiation. V At All Saints, the spirit is first welcomed and mourned individually, celebrated during this season. Carnival brings seasonal reversal. Throughout the dry season^ collective consumption and feasting is enjoined. People travel to distant parts of Laymi territory and beyond. Through music, Carnival marks a complete discontinuity between the presence of the ghosts and their subsequent banishment to the land of the dead. While they have been there for the whole growing season, their presence is vividly dramatised during the final days before they leave. Flute music attracts rain; it is a form of dirge and thus will not cause offence to the dead whose co-operation is essential to bring the crops to fruition.19 In stark'-' contrast, the music of the dry season is joyful and celebratory. The5'' wayňu music of the rainy season is said to weep (q'asi) while the kirki of! the dry season is happy (kusisi). * The sharp opposition of emotions is indeed a common theme in Laymi ritual. It is found clearly in another rite celebrated for the individual soul; while the deceased is gradually caught up in the concentration of energies that produces the harvest, it is not immediately forgotten as an individual soul. In this further rite, after the first All Saints and Carnival, a life-size dummy of the deceased is made, wearing some of its former clothes. It is feasted and mourned, and then» with very little respect is seized by an affine, wife-taker to the bereaved' (tullqa); it is untied, dissolved and thrown by him out of the compound.1 While it is the duty of kin to weep on this occasion, it is correspondingly the duty of affines to laugh. In what is virtually an emotionali/ division of labour, the ambivalence that all must feel toward the recently-departed is enacted, and ends with the physical dissolution of the dead person's image.20 When All Saints comes round a second time for the individual soulX the rites of the previous year are repeated, but on a smaller scale and^ with less show of grief. After this a final ritual of dispatch must be** performed, known as misajant'aku (Mass for laying it to rest), in which a Mass is offered for the ghost and it is finally dispatched to the land of the dead as an individual. Only after this rite is a surviving spouse V 60 The dead and devils among the Bolivian Laymi allowed to remarry. The bereaved will continue to offer Masses and tob pour libations for the soul as long as it survives in living memory, but -it is no longer an individual threat. Its dangerous power is now harnessed to the great ceremonial cycle of production and fertility. The fate of the soul In the preceding account I have used indiscriminately the terms ghost, soul, and spirit to refer to the essence of a dead person which survives bodily decay. In doing this I have followed the apparent lack of consistency with which Laymi themselves think about the afterlife. Nonetheless there is a certain linguistic conformity, in that many of the terms used to talk about the recently-dead and the ceremonial addressed to them are of Spanish origin. The word used to refer to the corpse and the ever-present envious ghost of the recently-dead is the Spanish term for the non-corporeal soul (alma). What is in Catholic > of the soul are in their very nature vague and indefinite: we should not make them too clear cut' (Hertz, 1960:34). For example, the moment of death is when the ghost leaves the body, but one of the first ritual acts i after death is intended to tie the ghost to the body. Again the ghost sets out immediately on its journey, equipped with food and other necessities; yet it also stays to haunt the living. Until the final despatching in the misa jant'aku it is both present and departed. One 61 Olivia Harris Laymi man tried to resolve this conundrum in the same way as Tschopik (1951), by suggesting that one of the 'shades' (ch'iwu) goes try the land of the dead while another remains in the hamlet.22 The shades, in tune with Laymi belief that the number three signifies ritual;/ completeness, are conceived as tripartite; this man suggested in the'-'' interests of consistency that the third shade went after death to v 'heaven', a neat solution which also resolved the ambiguities ofv Catholic and indigenous visions. It was nonetheless contradicted byv' many other people. While the spirit of the dead is tripartite, to attempt./, a clear-cut division of functions would be to do violence to a meaning«/] I that is intrinsically shifting and multivocal. If Laymi believe that the v1 non-corporeal element of individuals is tripartite, it is not because the three parts go in different directions when they are liberated from the body, but more likely because they do not share Western < preoccupations with the transcendent unity of the individual subject. v Concerning the land of the dead itself there is greater consistency, but also contradictory elements. This place, known as Tacna, is held to lie on the other side of the sea, which the souls of the dead must cross on the nose, or in the ear, of a black dog.23 In reality the town of Tacna lies near the coast in the extreme south of Peru and is situated on the Laymi side of the sea. The ghosts are thought to travel the same roads v as those used by living Laymi; indeed the highlanders are not freed v from the dead for a while after Carnival, since all the ghosts travelling; back to Tacna from the valley region must pass on their way through the highlands. As well as lying on the further side of the ocean, Tacna-is said to be underneath, that is, a sort of underworld reached byv crossing the water. Underneath (manq"a) is in some contexts'41 synonymous with the land of the dead, and is the place where the sun goes when it is night on earth. This place is an inversion not only as regards diurnal alternation, but also of the seasons, for while the Laymi live through the rain it is dry beneath, and while on earth it is winter, the dead live through the season of agricultural growth.24 Descriptions of Tacna varied; there was some disagreement as to^' whether one could remarry there, or whether one remained for ever/ with one's first spouse of life on earth. But all agreed that the main» occupation of the dead is the cultivation of red chilli pepper (aji). To say v/ that somebody has gone to cultivate chilli pepper is a commons metaphor of death.25 This hot relish is a potent symbol in many^ contexts; also red as a colour is closely associated with the dead and mortuary ritual throughout the Andes. An old woman told me of how she once nearly died and went right to Tacna. It was a large town, but 62 The dead and devils among the Bolivian Laymi the buildings were very low, only a few feet high, and everything was red. The children all had flowers in their hair. It seems that when the souls of the dead are dispatched to Tacna at the end of Carnival they go to work, to cultivate a crop whose significance to Laymi culture is paramount, both in codes of food consumption and in magic. It is however a crop that the Laymi are no longer able to cultivate since they lost control early in the colonial epoch of the hot valley-lands they had formerly worked.26 The theme of seasonal renewal here receives an additional twist: through the rainsv when the living toil and cultivate in sadness, the dead among them are t-feasted and respected. Conversely when the living gather the harvest* and rest, the dead return to work. While there is no indication in» Waman Puma's account of mortuary beliefs, cited above, of a cyclical movement, his contrast between the two states enjoyed by the dead captures the stark opposition in Laymi cosmology. The world of the dead inverts not only day and night, summer and winter, but also productive activity and collective enjoyment; the dead produce a crop, that complements the subsistence-orientation of Laymi economy today - a crop which, according to the traditional patterns of Andean exploitation of the environment, the Laymi should be able to cultivate for themselves. While I have talked of the dead as an undifferentiated category, the description applies primarily to married adults. It is they -both women and men - who have achieved full social identity; this bears emphasising since so many accounts of mortuary rites apparently assume that it is only men who die. Laymi who die single encounter similar experiences in the afterlife, but their lack of completion must be made good by burying with them a domestic fowl as companion for the future: a hen for a male and a cock for a female. Married souls, as I have said, travel on the known paths to Tacna; the unmarried go instead over the thorn bushes, and by night when human beings do not under normal circumstances travel. Since there is today no class differentia-^ tion within the ethnic group, this distinction in marital status is one of ' the few points of difference in how the dead are treated. Children however experience a different fate. In general there is no < mourning beyond the kin-group when a child dies, and the feast» offered at All Saints is on a smaller scale, but children who die are not simply a smaller version of adults. They are called angelitos (Sp. = v little angels); at burial little white paper wings are attached to their v' white 'habit' and they fly to heaven. There they are set to look after" God's irrigation system and send water to earth in the form of rain.^ Alternatively, people say that when we weep for angelitos God £ 63 Olivia Harris punishes them with his whip, and their tears of pain fall to earth as rain.27 During the feast of All Saints there is some feeling that all souls must reach heaven; in other contexts there are assumed to be children in Tacna; certainly children are thought to be especially vulnerable to abduction by the envious ghosts of the recently-dead; still for most purposes the fate that awaits children and adults at death is recog-nisably distinct. The much closer connection of the angelitos with" Christian cosmological themes perhaps reflects the Catholic doctrine that only the innocent are assured of salvation. On the other hand though they go to heaven, they too are harnessed to the needs of the ň agricultural economy, both directly in sending rain, and indirectly in | that their work in their new home is to maintain the irrigation ditches. One class of infants is excluded from becoming angelitos, namely i, those babies who die before being baptised. All babies are baptised by a. priest on his rare visits to Laymi hamlets, or when Laymi are able to make the long journey to a parish church in the mining centre of Uncia or the old colonial town of Chayanta. Full baptism is thus often delayed many months; however immediately after birth a rite is held in order to give the child a name and thus bring it into the chain of signification that is culture. The rite itself is called 'to pour water' (um waraňa), and a baby that dies before this rite has been performed is known as a little Moor (mum ivawa; cf. Sp. mow = Moor). The Laymi definition oimuru is 'without a name', and Bertonio's 1612 (1956) dictionary of Aymara translatees this term (sutiuisa) by the Spanish infiel (= infidel). The Italian Jesuit however also notes-at length that the Aymara concept of naming is not the same as Christian baptiism, which is rather centred around the idea of cleansing (Bertonio, 1612 (1956):330); it is clear from the way he writes that Aymara culture in the early seventeenth century ^ attached greater importance to the nname than to the accompanying^ rituals, and today too the focus of baptism is the name. A baby that dies unnamed has no place in heaven; it must be left outside inhabited space for the mountain spirits to eat. If this due is not paid the mountains punish the community by sending hail, which in the Andes can ruin an entire harvest in the space of a few minutes. The infants who are nameless belong not to society but to the spirits v who preside over the Laymi landscape - the mountains whose power is highly charged but ambiguous. To the mountains also belong those who have committed incest, known as condenados or condemned ones. They are extremely threatening to the living; anyone who is unfortunate enough to meet one will surely die or suffer terrible misfortune. They live out on the mountainside, or travel round the world, wearing a stone on one foot, a prickly cactus on the other. 64 The dead and devils among the Bolivian Laymi Devils, fertility, ambivalence x The connections between the dead and the regeneration of fertility in Laymi culture are not immediately obvious. Human fertility is clearly t-affected negatively by the dead, in that women of childbearing age must be kept separate from the corpse, and if they conceive as a result of contact with spirits of the dead the result will be a monstrous birth. As was noted above, the reproduction of Laymi society is represented through the land and its fertility, rather than human fertility. But even here the connections between death and fertility cannot be immediately made. In many respects the crucial place occupied by the dead in the agricultural cycle is concealed by metaphor. Symbolic usage concentrates in many ways on the involvement of the dead in the continued fecundity of plant and animal life but these are not in the main explicit. The power vested in the collective ancestors is also indirect. Nothing in Laymi attitudes to the dead today seems to conjure up the spirit of 'adoration' and 'worship' used by early colonial writers to describe the veneration of their ancestors by Andean peoples. An illustration of this apparent ambiguity can be found for example in the concept niňu. Deriving presumably from the Spanish word for child (and thus possibly from the Christ child), niňu when used in libation refers to the ancestors. When I inquired why the ancestors should be called children, I was told that they were the children of the1 sun and moon, supreme deities of the Laymi pantheon, and identified. with the Christian God and Virgin Mary. Niňu however has also a multiplicity of other meanings that undermine this clear identification with the moral force of celestial deities. It can, for example, be used to refer variously to severe fevers, the carrion condor, the mountain spirits, bulls, and wife-taking affines. The rite performed to ward off the ravages of epidemics is known as 'taking out the niňu' (niňu apsuňa); bulls are the incarnation of telluric energy, but their very power lies in their uncontrollability; wife-takers, identified in myth and ritual with the condor, are necessary for reproduction but also the source of discord and fragmentation. Many of the meanings of niňu are clearly sources of power which are not fully controlled. Another referent of niňu suggests a close connection with the representation of fertility; this is the small crosses kept in the Calvary chapel (Calvario) and used in the feast of Carnival to symbolise fertility and life. The complex meanings embodied in this term are startling, particularly if it indeed derives from the unambiguous and positive power residing, according to Catholic doctrine, in the figure of the Christ 65 w Olivia Harris child. In Laymi usage niňu also refers in some contexts to the mountain spirits (kumprira), and this meaning perhaps gives the clearest indica-tion of its semantic parameters. The mountain peaks are the guardians of their lives; they are the source of bad weather, of hail and thunder and rain. Sacred and powerful places, they are also the source of life; they are simultaneously protectors and malevolent beings, bringers of fertility, but also disaster and illness (Martinez, 1976). The mountain spirits are classified as part of the 'evil sphere' (saxra parti); this is opposed to 'God's sphere' to which belong the sun, moon, and Catholic saints. The 'evil sphere' is that of the devils (in Aymara the word saxra means both bad, evil, and evil spirit/devil), or at least of indigenous deities who became 'evil' as Christian doctrine was imposed and Andean structures of power dismantled.28 It will be recalled that Carnival, marking First Fruits and the new year, is called the feast of the devils; disguised with goatskins and abundant vegetation these embodiments of the ancestors celebrate fertility at the same time as they pose a serious threat to the revellers. The Laymiw ancestors today do not represent a fully cultural force, embodying social morality and authority; the only unambiguous source of" morality is the sun (inti), whose cult was central to Inka statecraft, and ! who is today identified with God. The sun however is remote (Tschopik, 1951); it is the spirits of the 'evil sphere' who affect more directly the daily life of the Laymi. The devils work by night, and inhabit wild places, such as gullies, waterfalls, rocks and mountain tops; graveyards are classified together with these places, feared and avoided in everyday life. The placing of the dead in the 'evil sphere' is, finally, evoked by the fact that one of the common names of the mountain spirits is precisely the term for ancestors, or grandparents (achachawila). In some sense, then, the dead belong in a category of beings identified with the wild, a source both of fertility and of misfortune. An opposition between the wild and the social should however be applied with caution: the boundaries are shifting and relative. For example, while in one sense all the dead are 'devils' and categorised with other manifestations of the wild, in another sense the clear identification of unbaptised babies and of those who have committed incest with the mountains implies by contrast that other ghosts are socialised. Within this group beliefs about the journey taken, respectively, by the married and the single after death again argues that the former are closer to human society than the latter. It would however be a vain exercise to v attempt to fix the significance of the dead; they are a transitive force,'/ who move between the existences of dangerous, envious, ever-v 66 The dead and devils among the Bolivian Laymi present devils, and the worlclof Tacna where they toil in the fields and1 live a life that closely parallels that on earth. While they are devils, they V are more benign than some. They are a type of devil for which Masses are offered in church; with all the ambiguities of the Laymi interpretation of Christianity and the subversion of its proclaimed meanings in mortuary practice, the identification of the dead with the church renders them susceptible to the controlling power of the priest. On the other hand the close association of the dead with church buildings has the effect that these are dangerous at night and are avoided. Night is the time of devils and particularly of the spirits of the dead. While the Laymi ancestors do not today embody social morality and authority, the feasts for the dead are the most important in the Aymara ritual calendar. This being the case, it could be argued that the dead are the major source of cosmic power from Laymi society. Why then do they pose such a threat to the living, particularly since the Laymi perceive little threat from biological processes in themselves? In part the answer to this question must lie in the fact that the processes of fertility and growth are themselves not fully controlled by the living. But there is a further dimension that concerns the exercise of political authority overall. As other papers in this volume suggest, the cult of i the dead may be a way in which social and political authority is t rendered eternal. v The fragmentary evidence available suggests that in Andean societies before 1532 the disposal of the dead and the cult paid to them was closely bound up with hierarchy and sources of authority. Today, too, there are some traces of a close connection between political office and the cult of the dead. The local headman (jilanqu) is responsible for his community's tax payments, formerly for the Spanish crown, and more recently for the Departmental coffers of the Bolivian Republic. At the same time the headman incarnates the prosperity of those he represents. It is he who ceremoniously initiates the ploughing which brings fallow land into cultivation; known as the shepherd, he must pour many libations throughout the year for collective fertility. As the person who assembles the tax payments of the community he is the critical agent mediating between State power and subsistence cultivators; he must also hear and settle disputes. His ritual importance is signalled by the fact that unlike other office-holders, he must be accompanied in all formal rites by his wife, or at least by a female partner, thus signifying the complementary duality which in Andean thought is the essence of continued reproduction (Harris, 1978a). The new office-holder is installed at the beginning of Carnival; indeed it is his instalment on the Thursday before Carnival that 67 Olivia Harris initiates this great feast. This, in the Laymi calendar, is the time not only of First Fruits and New Year, but also the point at which the new agricultural cycle begins, with the first ploughing of the fields which will be brought into cultivation the following spring (September). In other areas of Aymara culture the connections of this office with the dead are more explicit: in Kaata for example the various secretaries of the national peasant union preside over every funeral (Bastien, 1978:178-87), and in many parts of the altiplano the office of headman is actually initiated at the feast of All Saints.29 Today the office of headman rotates annually. In one sense it could be argued that the rotating nature of the office eternalises it, since it is thus separated from an identification with any particular individual or individuals; however it also exemplifies the extreme limitations on power within the structures of Laymi society. All landholding adult * males are expected to take their turn in office, regardless of their ^ suitability for the post, so that in years when it is held by a weak or! ineffectual person, little authority is wielded. The same holds true for the moiety authorities of the ethnic group as a whole. Today the sources of power lie outside the reach of Laymi culture, in the army, the Church, among those who control wealth and the State apparatus. The ethnic groups of northern Potosi have not been able to forge a 'pact of reciprocity' with the Bolivian State (Piatt & Molina, in press) by which the exercise of power might be clearly demarcated. The way that external power is represented suggests that for the Laymi it is ambiguous. The Catholic priest, for example, is the representative of the sun-God, but he is also a secret evil-doer who steals life-giving fat from the bodies of Indians to use for his own nefarious purposes. The townspeople who lose no opportunity to remind the Indian population of their inferior, savage status, and contrast it with their own mastery of civilisation, are called in Aymara the 'undressed' (q'ara). For the Laymi there is thus no longer a clear source of power from which the social order is mythically derived. It is true that in some contexts Laymi talk of the sun (inti) as such a source, but the sun is a general source of morality and social life, and does not often intervene directly in human affairs. The weight of Laymi representation of thel/ sources of good and bad fortune, of punishment and reward, is in the v 'evil sphere' where the mountain spirits guard jealously over their » prerogatives, and where the devils and the dead are the source of illness and death as well as fecundity and life. As power was removed from the ethnic group the veneration of their embalmed lords was forbidden; today among the Laymi, social office has ceased altogether 68 The dead and devils among the Bolivian Laymi to be incarnated in particular individuals. There are no monuments and few permanent records of the dead; those who are named receive full burial and feasting and in turn with their presence they ensure fertility. A source of ambiguous power, like the mountain tops, like the bulls who plough, they are harnessed to the service of society in an alliance which is always precarious. As long as their name lives on in someone's memory they will be offered a libation. Thereafter they become part of the crowd of 'nameless ghosts' for whom a Mass is offered at the feast of St Andrew - the end of the Inka month of the dead. Many anthropologists have written of the victory of society over thev deaths of individual members. This victory is a victory of symbolic. integration, of the harnessing of the potential anarchy of death to the moral organisation of society. For the Laymi today such integration is only partial. For them it would perhaps be more accurate to talk of an uneasy truce. NOTES I wish to thank Javier Albó and Tristan Piatt for their detailed comments on an earlier draft, and also the contributors to this volume for valuable discussion. The errors of interpretation that remain are mine alone. 1 Arriaga (1621 (1968)) and Duviols (1971) give details of the strategies employed. 2 Duviols (1973; 1978). Archaeological evidence for Qullasuyu also reveals traces of what have been assumed to be mummies and embalming cloths (Ibarra Grasso, 1965). 3 Duviols (1973:164-5) argues that mummies, known in Quechua as mallqui, were the equivalent of huari (venerated stones) which were the 'husbands' of the maize (zarasmamas). 4 As Cieza de Leon (1553 (1947)) wrote: 'There are great differences [in the mode of burying their dead], for in some parts they make holes, in others they place their dead on heights, in others on level ground, and each nation seeks some new way of making tombs.' 5 Harris (1982). Field research was conducted from 1972-4, and again in 1981. In mortuary practices I have been able to detect little systematic differentiation between the two zones. I participated in funerals and feasts of the dead in both, and accordingly my account will be a composite of observations made in both likina and suni, unless explicitly stated to the contrary. 6 Piatt (1982) gives a detailed discussion of land rights in northern Potosi today, and their relation to the nineteenth-century cadastres. 7 Carter (1968) and Tschopik (1951) also stress the overwhelming authority of old men in the Titicaca region. 8 Harris (1978a; 1978b); Godoy's (n.d.) study of the neighbouring Jukumani suggests that the relative absence of authority in Laymi society may not be typical for the region as a whole. 69 Olivia Harris 9 Carter (1968) gives a detailed account of similar games in Irpa Chico. J. Albó (personal communication) notes that the game played with sheep knuckle bones is, like the sheep, of Spanish origin. 10 While I am not clear as to the reasons for this contrast in location of graveyards between the two ecological zones, Bastien (1978:174) suggests that in Kaata the dead return underground to the mountain top which is both point of origin and return (uma pacha in Quechua). Hyslop (1976:153-7) notes that in the lakeside Lupaqa kingdom, hilltop burial sites remained in use long after habitation had been moved down to the plains bordering the Lake in the Inka period. 11 While this pattern is typical of most Laymi rituals, there is no doubt that women fear the dead and prefer to stay away (Harris, 1980). This normal avoidance is occasionally breached by a powerful symbolic inversion; when illness has ravaged a community, the women, or even children if circumstances are bad enough, replace men as pall-bearers and run to the graveyard with the corpse. 12 Some of the most common icons bear the implication of mediation, e.g. ladders, birds, snakes. Bread babies in the form of a woman carrying a child are also common, and Nash (1975:150) suggests that they make a positive connection between death and human fertility. 13 Piatt (personal communication) argues that water, like mirrors, is an 'enemy of the soul'. 14 Other feasts are regularly the scene of fighting between individuals as well as groups (Harris, 1978ö). 15 A further message implicit in the helmets festooned with ribbons is the negation of fighting. 16 In the hamlet of Muruq'umarka this place was called the dogs' graveyard (anupampiuna); see note 23 (below) for the significance of dogs to the dead. 17 I owe this information to Wilier Flores, regional delegate for Northern Potosi for the National Institute of Archaeology. 18 In Quechua-speaking areas of northern Potosi, according to Wilier Flores, the rite at the end of All Saints is known as misq"a Carnaval, that is 'Carnival in advance', thus providing an explicit pointer to the feast with which All Saints is paired. 19 It may be significant that the quena, another form of Andean flute, was in other areas of the altiplano fashioned from human femurs taken from ancient burial grounds. The Laymi play this flute occasionally in the dry season, but the music of the quena is in the class of zuaymi, that is, rainy season music associated with the dead. 20 The rite is known as uruni, 'dedicated day'. In a hamlet of the neighbouring Macha that I visited, I was told that the actual sponsors of Carnival are those obliged to perform this rite for their dead kin - another overt association between the dead and the feast of Carnival. 21 The distinction between the Spanish-derived alma and the Aymara word amaya depends on context as far as I can judge. For example alma is used in the feast of All Saints, and in general when libations are poured for the souls of the dead. It is almas again that haunt the church at night. Amaya on the other hand is appropriate for contexts not connected directly with Christian theology, for example the road travelled by the dead is known as amay t"ak"i. The degree to which these two terms are associated with a 70 The dead and devils among the Bolivian Laymi non-corporeal essence is unclear to me. Laymi use the term alma to refer specifically to the corpse. Not surprisingly, Bertonio's dictionary (1612 (1956)) gives the meaning of dead body (cuerpo muerto) rather to the Aymara amaya. Today Laymi sometimes say, echoing the catechism, that after death we abandon the saxra kurpu (evil body) and go to heaven. There is thus no positive association for the bodies of the dead that might suggest some submerged memory of historical practices of embalming. 22 Today the terms alma and amaya refer exclusively to the dead. Shade (ch'iwu) and the Spanish-derived animu are employed to refer to the spirits of the living that survive bodily death. Duviols (1978:133-4) points out the close identity between the idea of a shadow and the animu as a non-corporeal element that leaves the body during certain illnesses and at death; he suggests connections with the sixteenth-century Quechua terms camaquen and upani. In Chucuito, Peru, the Aymara term axayu seems to correspond to the Laymi use of animu (Tschopik, 1951:210). 23 Dogs are associated with incest, and their bite can turn people mad. In the suni, graveyards were carefully sealed off not only to prevent the ghosts escaping but also to stop dogs getting in and savaging the corpses. Carter (1968:245) argues a close connection between dogs and the spirits of the dead. According to J. Albó, in the altiplano, black dogs are associated with humans (jaqi anu) while white dogs belong to the devils (saxra anu), In their journey to the land of the dead, then, ghosts are assisted by dogs that recall their human rather than 'devil' status. 24 In my innocence I frequently told people that I came from the other side of the world, from a place where it is day when night falls in the Andes, and where the seasons are reversed. They evinced no surprise, but some anxiety at my description. It was a long time before a woman confided to me in friendship that everyone had been forced to assume from what I said that I had come from the world of the dead to haunt them. In other contexts manq"a is better translated as below or lower. 25 This mythical use of a real place may reflect a previous epoch when the Laymi had direct economic links with that part of the Pacific coast (J- Murra, T. Piatt, personal communication). Tacna is known to have been an important zone of chilli production. Thus the term manq"a could in this context also refer to the lower lands of the ethnic group. 26 Archivo Nacionál de Bolivia: Tierras e Indios 149 (1592). Today Laymi use a wild pepper growing in the likina, and otherwise buy chilli from the market. 27 In this" belief we can perhaps see the assertion that children, who are sinless, belong to God in a special way. God in this context is the Spanish dios (Aymara: tyusa), and heaven the Spanish cielo (Aymara: si'/h). 28 Piatt and Molina suggest that in some contexts the best translation of saxra is 'secret' (in press, Ch. 3). The three 'spheres' of Laymi cosmology are discussed in Harris (1980). Taylor in a recent discussion (1980) argues that the word used commonly in the Andes to denote devils - supay - may well have referred in pre-Christian times to the ancestors. 29 Carter (1968), Albó (1972); the discussion of rotating offices today is only applicable for the highland suni: in the likina all indigenous offices of authority were abolished after the 1953 Agrarian Reform and replaced by a local branch of the national peasants union. 71 Olivia Harris REFERENCES Albó, J. 1972. 'Dinámica de la estructura intercomunitaria de Jesus de Machaca', America lndigena, 32:773-816. Arriaga, P.J. de, 1621 (1968). Extirpation de la idolatria del Perú, (trans. L. Clark Keating) Kentucky: University Press. Bastien, J. 1978. Mountain of the Condor. Metaphor and ritual in an Andean ayllu. St Paul: West Publishing Co. Bertonio, L. 1612 (1956). Vocabulario de la lengua aymara. Reprinted in facsimile, La Paz. Buechler, H. 1980. The Masked Media. The Hague: Mouton. Carter, W. 1968. 'Secular reinforcement in Aymara death ritual', American Anthropologist, 70:238-263. Cieza de León, P. de. 1553 (1947). Primera parte de la crónica del Perú. (trans, for the Hakluyt Society by C. Markham (1864)) Madrid: Biblioteca de Autores Espaňoles, 26, 1947. Duviols, P. 1971. 'La luttecontre les religions authochthones."Trai>. Inst.Fr. des Etudes Andines, 13, Lima. 1973. 'Huari y Llacuaz. Agricultores y pastores. Un dualismo de oposición y complementaridaď Revista del Museo National, 39, Lima. 1978. 'Camaquen, Upani: un concept animiste des anciens Péruviens', in Amerikanistische Studien. Festschrift f Ur Hermann Trimborn. edsR. Hartmann & U. Oberem. St Augustín: Collectanea Instituti Anthropos 20. Godoy, R. (n.d.) From Indian to Miner and back again: Small scale mining in the Jukumani ayllu of northern Potosi, Bolivia. Harris, O. 1978«. 'Complementarity and conflict. An Andean view of women and men', in Sex and age as principles of social differentiation, ed. J. La Fontaine. ASA monograph 17. London: Academic Press. 1978b. 'De ľasymétrie au triangle. Transformations symboliques au nord de Potosi", Annates E.S.C., 33 (5-6). 1980. 'The power of signs; gender culture and the wild in the Bolivian Andes', in Nature, culture and gender, eds. C. MacCormack & M. Strathern. Cambridge: Cambridge University Press. 1982. 'Labour and produce in an ethnic economy', in Ecology and exchange in the Andes, ed. D. Lehmann. Cambridge: Cambridge University Press. Hertz, R. 1960. Death and the right hand, (trans. R. & C. Needham) Oxford: Oxford University Press. Hyslop, J. 1976. An archaeological investigation of the Lupaqa Kingdom and its origins. Unpubl. PhD thesis, Columbia University. Ibarra Grasso, D.E. 1965. Prehistória de Bolivia. La Paz: Amigos del Libro. La Barre, W. 1948. The Aymara Indians of the Lake Titicaca plateau. American Anthropologist, Memorials no. 68, Menasha. Martinez, G. 1976. El sistema de los uyiviris en Isluga. Isluga, Chile: Centro Isluga de Investigaciones Andinas, publ. 1. Nash, J. 1975. We eat the mines and the mines eat us. Columbia: University Press. Piatt, T. 1978. 'Symetries en miroir: le concept de yanantin chez les macha de Bolivie', Annales E.S.C., 33 (5-6). 1982. El ayllu andino y el estado Boliviano. Lima: Instituto de Estudios Peruanos. (n.d.) The Macha Ayllu. 72 The dead and devils among the Bolivian laymi Piatt, T. & Molina, R. (in press) Qhuya Runa Lima: Instituto de Estudios Peruanos. Taylor, G. 1980. 'Supay', Amerindia, 5: 47-63. Tschopik H 1951. The Aymara ofChucuito, Peru. Anthropological Papers of the American Museum of Natural History, 44 (2), New York. Valcarcel, L. 1948. 'The Inca calendar', Handbook of South American Indians, 2, Warr!ľnSpímaOIde Ayala, F., 1613 (1936). Nueva coronica y buen gobierno. Facsimile edn, Paris. 73