Carsten, J. (1991). Children in Between- Fostering and the Process of Kinship on Pulau Langkawi,...

CHILDREN IN BETWEEN: FOSTERING AND THE PROCESS OF KINSHIP ON

PULAU LANGKAWI, MALAYSIA

JANET CARSTEN

University of Manchester

This article analyses the way kinship is created in and through the conception and nurturance of children amongst Malays on the island of Langkawi. In Langkawi ideas about co-eating and sharing in the house are as fundamental to kinship as are ideas about procreation. These ideas are particularly associated with siblingship, representing the unity of children originating from within one house. Yet children spend much of their lives between houses, and are important mediators in exchanges between them. This mediating role is especially clear in widespread fostering arrangements. Analysis of these arrangements shows that they are centrally concerned with the conversion ofaffinal into consanguineal links, and with the maintenance ofequality between co-parents-in-law. Thus children represent points of transformation between two contrary images of community, one based on the unity of the house, the other on the plurality of exchange between houses. In their growth and their movement, children embody the process of kinship.

In this article I analyse the cultural construction of kinship among Malays on the island of Langkawi. Amongst these Malays ideas surrounding co-eating and sharing are as fundamental for kinship as are ideas about procreation. And both eating together and shared consumption more generally are intimately bound up with notions of siblingship. In fact, siblingship is in many ways more elaborated than filiation. 1

In order to analyse the specific meaning of kinship in Langkawi I shall focus on ideas about children: where and with whom they live and the ways in which they become kin. I shall argue that in Langkawi relatedness is not given at birth. Kinship can be viewed as a process of becoming, and this process involves a complex of ideas which centre on shared consumption in one house. The process of kinship continues during childhood and adulthood - as long as people consume together in houses.

Kinship in Langkawi involves the sharing of substance. This substance may be injected at conception in the form of seed, benih, from the father which is then nourished by blood, darah, from the mother during gestation in the uterus. But it may also be acquired through nurturance. In fact, the acquisition of substance through nurturance can, in certain circumstances, be more significant than that injected at conception. Kinship may begin with something you are born with, but it may not. That is, it is subject to change. As you grow, you acquire it. Eating together implies having blood in common. One becomes related to the people

Man (N.S.) 26, 425-443

This content downloaded from 137.132.123.69 on Sun, 13 Oct 2013 22:23:31 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/page/info/about/policies/terms.jsp

426 JANET CARSTEN

with whom one shares consumption. And these may or may not be one's genetic kin. In other words, although one is born with blood, one's blood also becomes. The substance of kinship is acquired in the course ofchildhood through developing social relations in the house.

I will focus here on notions about children and fostering arrangements to show how these articulate and reflect more general concepts of kinship and community in Langkawi. I will argue that there is a sense in which, from the moment they are born, all Malay children in Langkawi can be viewed as foster children.2 This occurs through the way they are attached at birth not only to their parents, but also to their grandparents, and through these to members of the community at large. The significance of children's simultaneous filiation to both their parents and their grandparents can only be grasped, however, through an understanding of the wider meanings of kinship in this society. In particular, it is necessary to examine notions about houses, consumption and siblingship and the way these are linked together.3

The island of Langkawi is situated off the west coast of Malaysia and adminis- tratively forms part of the state of Kedah. It should be emphasized that Langkawi's population, both today and in the past, has been rather mobile. Historically, the island seems to have been a kind of 'refuge area' for peasants evading warfare and corvee labour demands, or simply suffering the effects ofpoverty and land shortage on the mainland. Many inhabitants are connected by kin ties to the mainland; many have recent ascendants who have come as migrants to the island, or have themselves done so.

I would suggest that the remarkable flexibility of notions of kinship which I will describe in this article can be linked to this mobility of population. Com- munities were not necessarily stable but were capable of absorbing and connecting newcomers to a high degree, largely through marriage and fostering. Both population mobility and the extreme malleability of bilateral kinship are, however, rather typical of Southeast Asia in general. And it is in this context that fostering becomes significant.

The village in which I conducted fieldwork has a population of about three thousand, divided among several named hamlets. The chief source of cash income is fishing, and almost all men are involved in this commercialised activity. Sub- sistence rice cultivation provides an important agricultural base which is largely but by no means exclusively in the hands of women.

The house

The house has a fundamental structural significance for Malays, as it does in many Southeast Asian societies. Some implications of this emphasis on the house in Southeast Asia have recently been explored by, amongst others, Levi-Strauss (1984).4 To underline the fact that houses constitute a central feature of social organisation in these societies, Levi-Strauss has used the phrase societes a maison, house-based societies, to describe them.

Perhaps the most important principle embodied by the house in Langkawi is that of unity and resistance to division. Household unity is expressed in a number of ways. It is reflected in the spatial arrangements of the house which show a minimum of division. In particular, houses never have more than one hearth,



JANET CARSTEN 427

dapur. However many couples reside together in one house, they always cook and eat full meals, of which rice is a main constituent, together. This commensality is a prime focus of what it means to be of one household. Food, especially in the form of rice meals, becomes blood. To a great extent shared cooking and consumption create the shared substance that is at the heart ofkinship and co-residence.

Another reflection ofhousehold unity is the close bond that should exist between siblings. Disputes between adult siblings are always highly threatening and particularly disruptive when they occur. Such disputes are especially likely to arise over the division of household property through inheritance, and generally have far- reaching consequences for those involved. One indication ofthis is that the division of property tends to be deferred until long after the death of an original owner, resident descendants inhabiting and farming land in common until a disagreement makes the initiation of formal inheritance procedures unavoidable.5

The fact that siblings are strongly associated with the unity of the house results in a paradox. Married siblings never co-reside, and this is explicitly in order to avoid disputes between them. It is an attempt to avoid any conflict of interests that might arise between siblings through their obligations to their spouses and children. Although siblings, more than all other categories of kin, represent household unity (for it is the sibling group which is conceived as having its origins in one house), harmony within a group of married brothers and sisters can only be achieved through their residence in separate houses.

The importance of unity among siblings is reflected in a number of other ways. Siblings are expected to render each other aid and remain close throughout life, and this is especially clear in the warm, affectionate relations that obtain between adult sisters. One aspect of the stress on siblingship is its unique capacity, within the field of close consanguineal relations, to accommodate both equality and hierarchy. Equality tends to prevail when differences of age are slight while hierarchical modes of interaction are more apparent between siblings widely separated by age.

The birth order of a group of siblings is thus always significant. This is ritually expressed in rules governing marriage which ideally should take place in the order of birth of siblings, and this is particularly important between sisters. The in-marrying husband who disrupts the natural order of the sibling group by marrying a woman whose older brother or sister is still unwed is said to 'stride over the threshold', langkah bendul. This phrase implies that the husband in such cases is viewed as violating the integrity of the house itself, and he incurs a ritual fine.

Marriages are loosely endogamous, this notion being based as much on locality as on kinship. New houses are not established at marriage. At this stage residence is rather mobile. Villagers lay great emphasis on its uncertainty and unpredictability, explicitly denying the existence of rules governing post-marital residence. This can be connected with a more general unwillingness to accord prior rights over a young couple to either the wife's or the husband's kin in a context where marriage is conceived as occurring between parties of equal status. The general pattern, however, is that initially a couple commutes between the two parental households, spending periods of varying length in each. This is followed by an uxorilocal phase lasting until a couple has one or two children. A first child should always be bom



428 JANET CARSTEN

in the house of the maternal grandmother. It is only after this that a new house will be established, more often in the husband's than the wife's natal compound. Often this occurs prior to the marriage of a younger sibling to avoid the necessity for co-residence.

Children and houses

The association between a set of siblings and the house in which they originate is made ritually at the time of birth. Each child belongs to a set of 'symbolic siblings' whose existence precedes birth. The child and the placenta, uri, are conceptualised as 'two siblings', dua beradik. When a child is bom the afterbirth, uri, conceived as the younger sibling, is washed and placed in a woven basket together with various ritual objects by the midwife, bidan. It is then buried by the father in the grounds of the house compound in a manner which recalls the burial of human corpses outside the village in the graveyard. Similar beliefs and practices have been recorded widely elsewhere in Southeast Asia6 and can be related to a complex cosmology which has recently been explored by Headley (1983; 1987a; 1987b). What I would stress here is the way that the sibling set, in this ritual, is physically anchored to the house.

Siblingship thus asserts itselfin the womb and it continues to influence a person's fortunes throughout life. The uterus itself may be considered as the siblings' first house.7 Houses that are occupied after children are born merely create a weaker form of siblingship than that created in the womb. The very notion of personhood can be said to involve the relation of siblingship since even an only child - highly undesired - has its symbolic sibling. Although an individual may lack or be separated from human siblings she or he is still part of a sibling set whose other members closely affect his or her well-being.

It is notable that when villagers are getting to know a stranger one of the first questions they ask is how many brothers and sisters he or she has. The form of this question in Malay ('how many siblings are you?', rather than 'how many siblings have you?') underlines the indivisibility ofthe sibling set. This same principle is expressed in naming systems for siblings which both emphasize their similarity and that they constitute a complete set.8 Individuality, then, is always highly qualified by siblingship. Furthermore, it is very clear that this sibling set is closely associated with the house which gives it life. Parallel to a stress on the house there is, then, a stress on siblingship and these two concepts are directly linked.

The close association between children and the house is not confined to the symbolic level. I have described how new domestic units are only founded after the birth of children. This is one aspect ofa very great stress on the birth of children in marriage. During their early years particularly, marriages tend to be unstable. Children of both sexes are highly desired, and infertility may well result in divorce. The birth of children stabilises marriage; it also marks the attainment of full adult status whereas marriage without children does not.

The birth of a first child which occurs in a woman's mother's house signals a very significant change in her status, and is surrounded with a complex of ritual procedures which I will not discuss in detail here. In the forty days following the birth of any child the mother and newbom infant are particularly closely associated not only with the house where they are confined, but also with the symbolic



JANET CARSTEN 429

centre of the house, the dapur, hearth or kitchen. The conclusion of this period is marked by a ritual cleansing of the space of seclusion - the dapur - performed by the midwife, at the end of which the mother and child descend from the house. In a ritual of attachment, the baby's feet are made to touch the ground at the base of the house ladder, and it is then carried around the outside of the building three times so that it should always remember its house of origin.

Before children leam to walk it is normal for them to spend most of their time inside the house. I have described how commensality is a central aspect ofhousehold unity: co-eating creates the shared substance of close consanguinity. This is most clearly expressed in the fact that marriage between two adults who have drunk milk of the same woman is prohibited.9 Maternal milk is a form of blood, and siblings are thought to share substance to a particularly high degree because they have drunk milk from the same mother. The mother's milk is conceived as the source of the powerful emotional bond between her and her children.10 Rice meals, too, are a source of blood. Throughout childhood stress is laid on the propriety of children eating full rice meals in the house where they reside. The behaviour of children who frequently eat such meals in the houses of others is remarked upon and reflects badly on the repUtations of adult members of their own households. Even more than adults, children may, however, accept snacks in other houses. These can be defined as food which does not include steamed rice as its main constituent, and is therefore neither a full meal nor proper food in Malay terms.

As children grow older and spend less time within the house, their association with houses becomes more focused on eating and sleeping. At nightfall, when malevolent spirits are believed to be most likely to attack, it is particularly important for children to return home. With certain exceptions which I will discuss below, children do not normally sleep in houses where they are not resident.

I have described the close association between children and the house, particularly between the sibling group and the house from which it originates. It might be said that if individuals are in mnany contexts conceptualised as inseparable from their sibling groups, houses are the embodiments of such sibling sets.

There is, however, a paradox here. No visitor to the village could fail to be struck by the extent to which children are present outside houses, and correspond- ingly absent during the daytime from the interior of houses. If the identity of houses is so intimately bound up with the children that constitute the reason for their establishment, how is it that these same children are so little present within them? In what follows, I shall address this paradox.

Children between houses

Once they have learnt to walk, children spend most of their daytime hours outside the house. Pre-school children play with their age-mates in the space between houses, usually within the compound. After they begin to attend school, children are necessarily absent in the mornings, and religious instruction in the Koran will occupy them for several aftemoons a week outside the home. During these years children still tend to play outside the house rather than inside, although in ado- lescence this is less true of girls than of boys.



430 JANET CARSTEN

Apart from their play and study, there is a further dimension to children's absence from the house which can be seen not merely in negative terms, but as giving them a positive presence between houses. Children are the most frequent mediators between houses. They are the vehicles of messages, gifts and items of exchange. It is children who are sent by women with the dishes of cooked food and raw produce which figure so prominently in the sharing among closely related houses of one compound and also in the reciprocal relations of the wider neighbourhood.

Children often collect the monetary contributions to rotating credit societies which women organise in the neighbourhood. They are sent to the shops for purchases, and may take cakes which women make at home to the local coffee shops for sale. Children also deliver words between houses. When relations between two houses are tense children may be the only means of communication between them. They are the last finally to sever connexions with other houses. Children also play an important role in more formal visiting. On such occasions children are greeted with evident delight, made much of, and almost always given money as well as snacks by their hosts. There are many senses, then, in which children can be seen as embodying exchange relations in the wider community. And in this connexion it is significant that they are so clearly associated with items of exchange, in particular with money.

There is, however, one form of visiting from which they are almost totally excluded. These are visits to the sick and visits to the dead. Villagers are strongly enjoined to visit the house of any member of the community who is ill, or who has died and is awaiting burial. Such visits take place with extreme formality and contrast strongly with other forms of visiting in that they involve a minimisation of all forms of exchange. There is little conversation and the visit is conducted in hushed tones; snacks are not necessarily served; the atmosphere is quiet and subdued. In marked contrast to other occasions, children are not welcomed at these events.

By contrast, at the ketiduri, or communal feast, an occasion replete with many different forms of exchange, children figure proninently. At the kenduri rules governing domestic cooking and eating are inverted. Men cook outside houses, and men, women and children eat separately. The most frequent and elaborate kenduri are held to celebrate marriages, and it is appropriate that these are occasions at which the ties within domestic units, whether consanguineal or affinal, are suspended and exchanges between them, particularly of different kinds of food with different symbolic associations, are enmphasized.

Many of the principal participants at these events are men and women who are grandparents, and they often bring a grandchild along. These children together form an undifferentiated group and eat separately from their grandparents, as if to create a community of grandchildren by themselves. Such an image of community as one of shared grandchildren is highly significant and I shall return to it in my conclusion.

I lhave described how children embody both the close consanguineal unity of the house and the exchange relations between them, relations which often seem to be disruptive of, and opposed to, the unity of the household. Both marital relations and the conmmercial ties of fishing are highly unstable and are thought liable to cause disputes between close kin. The central ambivalence surrounding children is nowhere more clearly captured than in institutions of fostering.



JANET CARSTEN 431

Fostering

The majority of children live with their parents from birth until they have completed their education. Many continue to reside in the parental home until after they have married. At the time of my fieldwork, however, about one quarter of all minors in the village were not living with both their conception parents.11

The fostering of children covers a number of different arrangements which may or may not be terminologically distinguished from each other. The simultaneous possibility of drawing or erasing distinctions between conception children and different kinds of foster children parallels the way affinal relatives can be both conflated with, and distinguished from, consanguineal ones (see below). Such distinctions are always conceived in a highly ambiguous manner.

A foster child, anak angkat,12 may have lived from birth to the age of seventeen with her foster parents, or have spent a few years with them; the term may cover a university student who stays with a family for two weeks or an anthropologist who arrives for two years. Nor do different terms exist for a foster child who has no rights of inheritance from his or her foster parents and one who has been formally adopted.13 Ideally, there is a sense in which different kinds of foster children should not be terminologically or behaviourally distinguished from each other; likewise, ideally, there is a strong idea that an anak angkat should not be distinguished from a conception child although, in practice, such distinctions can be significant.

For the purposes of the analysis presented here it is possible to divide children not living with both conception parents into two categories: first, those living with a divorced or widowed parent who, in about one in two cases, had sub- sequently remarried; secondly, children being brought up by a couple in which neither partner was the conception parent.14 In the latter category villagers themselves often make a further distinction between fostering a child and bringing up the child of a close relative. Not infrequently, children reside more or less per- manently in the house of a close relative. Such residence varies in duration: it may endure for most of a childhood, for a few months or for some years; it may begin when the child is a baby or when he or she is already an adolescent. It may or may not involve visits of varying duration and frequency to the parental household. When a couple bring up their grandchild or a younger sibling they usually do not refer to the child as an anak angkat, foster child, but rather say they are 'caring for', or 'bringing up' the child, and specify the kin relation involved. In these cases the child uses appropriate kin terms for these relatives, whereas an anak angkat always uses parental terms of address and reference for her or his foster parents.

In fact, an anak angkat is not necessarily unrelated to her foster parents although the connection in these cases tends to be rather distant. In the great majority of cases, if there is a distant kin connexion it is through the foster mother.

The considerable, and often ambiguous, blurring between these categories is mirrored on the affective plain. Indeed, an attenuated or non-existent genetic link may even be 'compensated' by a strong emotional one. It is considered normal for a transferred child to love those who bring it up more than its conception parents. Such children are also considered the obvious favourites of their adoptive parents. But it is not only affection that is acquired through living together. Children transferred from a young age are presumed to take on the physical attributes of



432 JANET CARSTEN

their adoptive parents, coming to look like them, in the same way that conception children do.

The causes which precipitate fostering by distant kin or non-relatives are almost always death of one or both parents or their divorce. From the point of view of the foster parents, bringing up another person's child confers merit and prestige. Fostering is a means of avoiding divorce in cases of infertile marriages; it can also correct an imbalance in the sex ratio of children in accordance with the ideal of having roughly equal numbers of boys and girls.

In the case of divorce or death of one parent a child may be brought up by the other parent, but the father never has sole care of the child. He can only continue to look after his children if he lives with at least one female relative or if he remarries. By contrast, widowed or divorced women may bring up their children even if they live independently. There is a general reluctance to put children under the care of a step-mother who is regarded as potentially malevolent.

Categories of close kin that are particularly favoured for looking after children, in order of preference, are grandparents, older sisters, and sisters of either the mother or father of the child. Roughly equal numbers of female and male children are brought up by their close kin.15 The reasons for a child being cared for by close kin are often practical, as when a couple have a large number of children or when they are so close in age that their mother is overburdened with work. This is particularly likely if these kin have no young children of their own. In such cases, when a baby is bom it is generally an older sibling who goes to the home of relatives. Although children are often cared for by grandparents, older sisters or aunts and uncles, villagers always emphatically deny that these kin have absolute rights to a child, even if they are childless. On the contrary, people say they are malu (embarrassed, ashamed) to give a child away, that they should have an excuse for doing so. If a child is given to close relatives the shame involved is obviated. The grandparents on either side and a mother's sister are regarded as having particularly strong claims to a child if they themselves have no young children.

In general, however, villagers stress the voluntary aspects of fostering over the obligatory ones. The rights of parents to their own children have priority over any other claim. It is only once a couple have three or four children that the desire of one partner's parents for a grandchild, or of a married sister who is childless, would be taken seriously into account. Informants do not accord prior rights to either the husband's or the wife's kin, but ties through women are regarded as being emotionally stronger. Thus after grandparents, the most frequently cited preferred relative for this role is the child's mother's sister.16 This, as McKinley (1975; 1981) has forcefully argued, can be seen as reflecting the emphasis on closeness between siblings - in Langkawi this is especially true between sisters. In general, I would suggest that the distinction villagers make between looking after the children of close relatives and fostering non-relatives or distant kin relates to the ideological importance and specific values associated with the categories of close kin most frequently involved, especially siblings and grandparents.

The affective strength of ties through women, which is physically expressed in uxorilocal residence at the time of the birth of a woman's first child, thus receives further expression in fostering. A child's emotional tie to his or her foster mother is realised in co-residence. If, as I have suggested, the co-substantiality of kinship



JANET CARSTEN 433

is largely created through shared consumption, then, through cooking and feeding, women play a decisive role in this creative process. The acquisition of substance after conception begins with feeding in the uterus as the child acquires blood from its mother. The process continues with the provision of breast milk and cooked food, but these may or may not be acquired from the conception mother. In constituting themselves as foster kin who, ideally, are undifferentiated from conception kin, women actually create kinship. They provide the locus for this shared consumption to take place as well as fumishing its material and emotional content.17

The residence of children is often talked about by villagers in terms reminiscent of post-marital residence. In several cases where children were living with their grandparents I was told that this was 'not fixed' or certain, tak tentu, and that the child was alternating between the parental home and that of the grandparents. This alternation is described using the same term as that used for the alternation between the two parental homes in the first phase of married life, berulang (literally, repeated action). Great emphasis is always placed on the child's own wishes. I was frequently told that there was no reason for a child to be living with grandparents, she or he simply preferred to do so, or was 'following his/her own wishes', ikut suka dia. Following the individual wishes of a child is, of course, another way of stressing the uncertainty of residence, since such desires are seen as intrinsically unpredictable. In this way the voluntary and contingent aspects ofa child's residence are emphasized. However, the amount of attention paid to the wishes of a child may be greater in theory than in practice. The reasons for this relate to the crucial role played by children as mediators between affines.

Children between affines At this point it becomes clear that fostering cannot be understood in isolation from other exchanges involving the transfer of persons, in particular, that of marriage.28 I have described how marriage in Langkawi is conceived as taking place between those of equal status and involves great uncertainty over the residence of a newly married couple, who in fact are rather mobile in the first phase of married life.

In many ways the equality on which marriage is premissed may be said to focus on the relation between co-parents-in-law, bisan. Bisan constantly visit each other's houses and are engaged in a complex web of exchanges involving food, labour, services and loans. These exchanges do not, in fact, necessarily take place between the actual bisan but they are usually described as if they did, and this is indicative of the special significance with which this relation is invested. Bisan are frequently called upon during rice harvesting, for work in the vegetable gardens, housebuild- ing, for loans of cash or equipment. They send each other gifts, especially of cooked food and of raw produce. Apart from immediate neighbours, they are often the most frequent visitors in each other's houses. A high value is placed on reciprocity in these exchanges. Neglect of obligations to give assistance is a frequent source of bitterness, and such grudges may be brought up much later in a dispute.

The mutual assistance supplied by bisan is broadly of the same kind as that given to close kin or immediate neighbours, but there is a difference. This has to do in part with the emphasis on reciprocity already mentioned. In the case of kin and neighbours there is a greater tolerance of imbalance in the short term. Following from this, the exchanges between bisan not only occur with great frequency, but



434 JANET CARSTEN

are also particularly marked. They receive constant attention in behaviour and in speech, while those between close kin and neighbours tend to be taken for granted. Nor is there any other affinal relation that receives this kind of stress. On one level, then, it is as if an attempt were being made to assimilate a bond of affinity into the realm of consanguineal kinship. On another, this attempt is belied by the very emphasis it receives, by the intensity of the exchanges and the importance of reciprocity.

It should be emphasized here that notions about affinity simultaneously involve both an elision with, and a distinction from, consanguinity. This can be associated with the very loose form of endogamy to which I have already referred. Marriage should take place between those who are 'close' (dekat), a notion which conflates kinship, locality, appearance and status. It may, but does not necessarily, take place between kin. Although affines are terminologically distinguishable from consanguines by the use of an appropriate suffix, it is highly impolite to make such distinctions, whose use is principally associated with disputes. In general, affines use appropriate consanguineal kinship terms when addressing or referring to each other. Thus spouses and affinal siblings should in theory be addressed using sibling terms, and parents-in-law are addressed using parental terms.

Bisan, however, constitute something of an exception to this rule in that there is one way in which they may actually address each other using an affinal term. This is by calling each other tok bisan, 'grandparent bisan'. It is significant that this usage combines a term of consanguinity with one of affinity for, as will become clear, 'bisanship' can be seen as a point where, in local conceptions, consanguinity and affinity merge. Furthermore, by using this form of address they draw attention to one of the most important constituents of their relation: their common grandchildren.

Responsibilities towards grandchildren are an important aspect of grandparents' lives. These responsibilities begin with a first pregnancy which almost always occurs in the house of the young woman's mother. In the seventh month of pregnancy the services of the village midwife, bidan, are secured by the husband's mother who is responsible for payment of the midwife. The midwife has jurisdiction over all births in the community. At childbirth women's isolation from their mothers and close kin is notable. It is significant that the midwife takes on responsibility for the birth on behalf of the grandmothers rather than the mother of the child.

As the pregnancy advances, the husband's mother begins to buy articles of clothing and other necessaries for her grandchild. At the time of birth she is expected to go to the house of her bisan again and to stay there for some days. The crucial role of the grandmothers, already marked before birth, is at this point confirmed. Wilkinson observes that 'The honour of the first introduction [the midwife] gives to the child's grandmothers' (1957: 41). The child's attachment to its grandmothers is thus stressed at the birth itself.

The obligations performed during pregnancy and at the birth of a child are perceived in terms of the relation between bisan. Women say that these duties are a 'responsibility between bisan', tanggungan antara bisan. They continue after the birth of a child. The husband's mother pays for the feast and services of the midwife at-the time of the first shaving of the child, cukor anak, seven days after birth. The relation between bisan continues to focus on the welfare of their common



JANET CARSTEN 435

grandchildren throughout their childhood. Especially during their early years, it is important for the female bisan to make protracted visits to the home of her counterpart in which the young couple resides whenever a grandchild is sick or for any social occasion conceming them.

Not only are grandparents expected to visit their grandchildren, but the latter must also visit the former at frequent intervals and often stay the night there. Once again, where the children and their parents are normally resident with one set of grandparents it is particularly important that they make visits to the other set.

I have described how children are brought up by grandparents more often than by any other category of kin apart from parents. Very often one set of grandparents bring up a grandchild in this way when their counterparts reside with or in the same compound as the young couple and the rest of their children. Once again this can be seen as a way of balancing the affiliation of a couple and their children to one set of parents.

This became particularly clear to me during a dispute involving the members of three households. An aged couple who co-resided with a married daughter and her husband faced a conflict between their allegiance to their co-parents-in-law, bisan, who lived in a different neighbourhood, and that to the household of a sibling and next door neighbour.19 If they continued to maintain their close relations with their neighbours they risked offending their affines. If they offended their bisan they put at risk the continued residence of their daughter and her husband in their own house, and might well jeopardize this daughter's marriage. It was highly significant that their daughter's husband was the sole income earner of the family. Eventually the couple had no option but to break off their relations with their neighbours. Members of these two households ceased to interact in any way.

At the same time, and as if to reinforce this statement of allegiance, a child of the resident daughter and her husband was transferred to the husband's parents. At the time, members of both households told me that this was done because the child herself wanted to stay at her paternal grandparents' house. The child's behaviour, however, was not quite in accordance with their assertion.

Forced to ally themselves with either their close neighbours and kin or the parents of their son-in-law, the senior couple had no choice but to break their relations with the former in order to secure both the livelihood of the whole family and their daughter's marriage. The transfer of a child at a time when all relations had reached a crisis point was an extreme expression of this allegiance. I would suggest that the 'gift of a child' can in this society be seen as the 'supreme gift', to use a term Levi-Strauss has adopted in a different context (1969: 65).20

I have already referred to the problems raised by the affiliation ofa newly married couple to one or other parental household in a context where the equal status of the two families is strongly stressed. Only the gift ofa child can correct the imbalance created by affiliation to the household of one set of parents, compensating the loss that one house or the other incurs after marriage. Often, but by no means always, this involves the 'replacement' of a married daughter by a female grandchild. Villagers vigorously deny, however, that either set of grandparents has rights to a grandchild, just as they deny rules of post-marital residence. Not to do so would imply a superior claim on the part of one side or the other.



436 JANET CARSTEN

The 'sharing' of common grandchildren is, then, a crucial element in the relation between bisan.21 People in Langkawi often stress the particular closeness of bisan and that it is having grandchildren in common that makes them so. It is thus not surprising that one of the terms of address they use, tok bisan, emphasizes their role as co-grandparents. It is through shared grandchildren that the relation between bisan - one of affinity - is actually transformed into one of consanguinity, and this is explicitly emphasized by villagers.

This also explains why the birth of a first child receives such ritual attention. For not only does the first child establish the consanguineal principle in a new nuclear family (and couples never establish a new household until they have at least one child), it also creates the consanguineal link between co-grandparents. Moreover, while the first child is herselfrarely fostered, she establishes the possibility for the fostering of subsequent children.22

Exchanges between houses are conceived in many respects as exchanges between bisan, and this is clear from the way these exchanges are conducted and described by the participants. In other words, relations between households are based on an idea of symmetrical exchange between equals. If these relations are conceptualised as relations between bisan, grandchildren are often the means by which these links are activated, their frequent focus, and a path of mediation between these houses.23

Two images of community

From the ethnography I have given it is possible to draw out two models of community in Langkawi, and these appear to be opposed to one another. In sketching out these models I would stress that they constitute polar images, neither of which can ever be wholly satisfactory because each implies a rigidity and fixity that is in fact absent. As I shall argue, it is always possible to move between the two and to transform one into the other, and this possibility is an essential part of the creative dynamic of this system.

The first model is based on the house and on siblingship. Persons united through these ties should ideally hold property in common; they are conceived as an undivided group who share consumption. During the course of the domestic developmental cycle new houses are established as the siblings, having grown up and married, have children. Frequently, these houses are on the same compound land as either the husband's or the wife's parents. Eventually, neighbouring houses of one compound come to be occupied by adult siblings, or by the descendants of a sibling group. To a considerable degree compound members share resources, particularly land. As compounds themselves expand they are subdivided, and new ones are formed in the vicinity. Neighbourhoods come into existence in which houses are connected to each other through ties that can be traced back to siblingship. The village community is formed through the process of house expansion.

This continuity between the house and the community is expressed in the term kampong which can refer to a compound, a small hamlet, or a large administrative and religious unit consisting of several small hamlets. The compound itself may have one house or ten houses built on it. In other words, the house and the wider community are directly linked terminologically: the latter is conceived as the former writ large.



JANET CARSTEN 437

Similarly, the community is conceived as being united through ties ofsiblingship. Sibling terms or derivatives of these constitute some of the most common forms of address and reference for co-villagers. When two people are asked to trace out a kin connexion, however distant, they always begin or end at a point where two ascendants can be named as siblings. When villagers emphasize their connectedness to each other in a general way they say kita semua adik-beradik disini, orang lain t'ada, 'we are all kin here, there are no strangers'. The term for kin, adik-beradik, is itself derived from that for younger sibling, adik.

The conception of the community as an expanded house is expressed at communal feasts at which the most powerful domestic symbolism is used to evoke a sense of community. Such feasts involve members of the community consuming together one extravagant meal. Yet at the same time, as I have already shown, such feasts constitute in another way a direct negation of the house. Domestic ties are denied, domestic rules of cooking and eating are inverted.

This ambivalence can be related to the second image of the community: that which is based on exchange. Houses engage in a series of reciprocal and equal exchanges of visits, gifts and labour with each other. The commercial transactions of the fishing economy and the tense equality of marital exchanges are explicitly seen as potentially threatening to domestic harmony and unity - particularly that among siblings.

Inasmuch as houses represent sibling groups which share resources and consume together, with different houses eventually coming to contain different members of sibling sets, they also represent bisan exchanging on an equal basis. If siblings share consumption to a degree that renders marriage between them incestuous, bisan are those who, by definition, exchange sons and daughters in marriage. Both bisan and sibling sets are intimately linked to the conceptualisation of the house. How can this tension be resolved? In the concluding section I will argue that it is children who provide a solution to this problem.

Children as agents of transformation I have argued that in Langkawi notions of relatedness are founded on shared consumption in one house. Siblings share consumption more fully than all other categories of kin. They are normally brought up in the same house and share food from birth on. It is feeding that creates blood, and human milk is seen as a form of blood. The prime category of incest is between those who have drunk milk of the same mother, in other words siblings. Siblings are conceived to share consumption to a degree which would negate the exchange that is marriage, and to a greater degree than all other categories of kin. It is thus siblingship and not filiation which constitutes the closest possible tie as well as the paradigm for moral relations.

I have described how children, in so far as they are part of a sibling set, embody houses, and are anchored to them. This association is so strong as to overcome the fact that children may not genetically belong to such a set, just as it may overcome the actual division of a sibling group into different houses. Residence and co-consumption in a house can confer membership in the sibling group associated with that house, even for a child who is fostered there. I have also described how children mediate between houses. They seem almost to embody



438 JANET CARSTEN

exchange as they take words, gifts and money from house to house. And this role is most fully captured when they themselves take up residence in another house through the various forms of fostering that occur in Langkawi.

The different aspects of fostering reflect children's capacity to participate in the two images of community which I have outlined. In so far as they are brought up by their older siblings, or by parents' siblings, they may be seen as demonstrating the unity of the sibling group and the extensive sharing that occurs within it, as McKinley (1975; 1981) has argued. They are living out the notion of the village community as an expanded house. But when they move to the houses of distantly related or unrelated co-villagers they are involving themselves intimately in the exchanges that occur between different houscs of the community.

This idea is also present when they reside with their grandparents, as they frequently do. Residence with grandparents, as we have seen, reflects, and may compensate for, the imbalance caused by post-marital residence in the context of an ideological stress on marriage between equals. In this light it must be considered as an aspect of marital exchange itself. It is also clear, however, that the birth of grandchildren signals the formation of new sets of siblings and the establishment of new houses. Sets of grandchildren represent both undivided domestic consanguinity and affinal links between houses. In fact their birth transforms the latter into the former: once bisan have common grandchildren they are, in local conceptions, united by a consanguineal tie.

This, then, is at the heart of children's 'inbetween-ness'. It is because they themselves have a dual aspect - representing the point of transformation of two opposed images - that they can perform in their daily lives a constant mediation between representations of these images. As products of their parents, children divide houses from each other, and emphasize their exclusivity. Through the mobility of their parents, the possibility of fostering and as products of their grandparents, they unite different houses.

If grandchildren can take up residence in the house of either set of grandparents then it follows that these two houses can actually come to contain siblings. The circle is completed as two houses are linked by both an affinal tie and one of siblingship. The fostering of children can utilise pre-existing ties of siblingship and reinforce them; it can also actually create such ties.

The movement of children can always be interpreted in a number of different ways, from sharing to exchange. In fact, the movement of children blurs the distinction between sharing and exchange in that it may be interpreted either as exchange between discrete units or as sharing within an expanded unit. It is sharing in that, from one point of view, it breaks down the boundaries between houses and is constitutive of the image of community as an enlarged house; it is exchange in that, from another point of view, it is part of a series of reciprocities between discrete and bounded household units. And in the last analysis it is this ambiguity which lends it force.

The ambiguity surrounding children's movement can now be linked to the considerable terminological and behavioural ambiguity between different fostering arrangements which I described above. Distinctions may be made at one level only to be elided at another, just as affines may be distinguished from consanguines only to merge with them at another level. It is this ambiguity which allows the



JANET CARSTEN 439

rapid and easy absorption of newcomers to the village through their transformation into affines and foster kin, and the transformation of these categories into consanguines.

Furthermore, I would suggest that children have this capacity both to unite and to divide houses to a particularly heightened degree because they are not yet fully formed. It is their 'unformedness' or 'incompleteness' which lends them their ability to move, seemingly unaffected, between houses. This unformedness might be considered to some extent an inherent quality of childhood. Physically, children, as they grow, are acquiring substance more obviously than adults. In Langkawi young children, in particular, are not considered to be fully in control of their desires. If two children quarrel it is always the youngest who is favoured when elders settle the dispute. The younger child is thought to be less capable of understanding particular rights and wrongs through rational consideration.

I would also suggest that the incompleteness of children relates to the fact that they have not yet married or had sex.24 They have not directly participated in, or incorporated, the division which affinity introduces into the unity of the sibling group. In this sense we can understand why it is that they have the capacity to involve themselves in exchange without being affected by it, why the actual separation of young siblings does not appear to raise major problems. Because their association with their own sibling group has been untainted by the divisive dangers of affinity, it is strong enough to overcome partition. In this respect we may recall children's 'non-individuality', which I discussed above: conceptually, they are not fully separated from other members of their sibling group; practically, they are ideally placed to be thus separated and to involve themselves in both sharing and exchange between houses.

It is having children that confers adult status; conversely, childlessness is the mark of childhood. Incomplete themselves, children, who constitute one half of a house, have the capacity to complete the other half- the married couple without children. Likewise, children must eventually be themselves completed by marrying and having children. The child becomes a completed adult through having children and then grandchildren. Here we may recall Schneider's (1984) discussion of 'Western' notions of kinship. Kinship for us, he argues, is rooted in 'biology', that is, a completed, fixed genetic component which is implicitly or explicitly distinguished from and opposed to the social. Such an opposition does not apply to Malay notions of relatedness. Children's incompleteness can be seen as a demon- stration of this. Their biological substance is not given ready-made at birth. Rather, both their substance and their sociality can complete, but must also be completed by, that of others.25

I would argue that in Langkawi the stress on the production of children in marriage can equally be seen as a stress on the production of shared grandchildren of bisan, and this process is itself intimately bound up with the reproduction and expansion of the community at large. It is through the production of children and grandchildren that consanguineal ties are created from affinal ones, and new sibling groups come into being which will in the future intermarry. In this way links between households are reproduced through grandchildren. The community is envisaged as one of shared grandchildren. This image, which is projected at communal feasts, is realised in the widespread sharing and exchange of children



440 JANET CARSTEN

that occurs in fostering. Grandchildren resolve through their 'inbetween-ness' the opposition between siblingship and marriage, between sharing and exchange, and between unity and division. They constitute the practical reality and the full complexity of links between houses. Their growth and their movement embody the process that is kinship.

NOTES

Fieldwork was conducted on Pulau Langkawi between October 1980 and April 1982. It was financed by a grant from the Social Science Research Foundation (now ESRC). Additional funds were provided by the Central Research Fund of the University of London. A further visit in 1988-9 was financed by grants from the Wenner Gren Foundation, the British Academy and the Evans Fund of the University of Cambridge. During the initial writing of this paper I was sup- ported by a fellowship from the Evans Fund. Various earlier versions of the article were presented at seminars at Goldsmiths' College and the School of Oriental and African Studies, London, the College de France and the University of Manchester. I am grateful to participants in those seminars and to Maurice Bloch, Esther Goody and Maria Phylactou for their comments. In addition, I owe a greater than usual debt to the editor and referees of Man for their helpful suggestions.

l This contrasts with our own notions of genealogical closeness in which filiation takes prece- dence over siblingship (see Schneider 1984: 173). The most complete discussion of Malay notions of siblingship is that of McKinley (1981). See also Errington (1987) for a discussion of the importance of siblingship more generally in Southeast Asia.

2 The analysis of fostering has tended to focus on its numerical and demographic significance in the face of a prevailing assumption that it is a phenomenon affecting a niinority of children. See Goody (1982) for West Africa. I have not included numerical data in this article. They are recorded in Carsten (1987a: 118-26).

3 See McKinley (1975) for a detailed study of the relation between child transfers and siblingship among urban Malays.

4 See also Levi-Strauss (1979; 1983). Barraud (1979), Errington (1987), Macdonald (1987) and Waterson (1987; 1990) have discussed house-based societies in Southeast Asia. For a discussion of the ways in which this model is relevant to Pulau Langkawi, see Carsten (1987b: n.d.).

5 See Carsten (1990) for a discussion of inheritance procedures. 6 See Laderman (1983) and Massard (1985) for a description of these rites elsewhere in Malaysia;

H. Geertz (1961: 89) for Java; Snouck Hurgronje (1906: 375) for Acheh; Hooykaas (1974: 93-128) for Bali.

7 See Headley (1987a). This idea is particularly powerful in the case of twins. It is notable that the more hierarchical Southeast Asian societies seem to have a particular fascination for cross sex twins in terms of incest and marnage, see Boon (1977: 138-40, 201-2) and Errington (1987). McKinley (1975: 226) suggests that once a child begins to be able to socialize in the house it no longer needs to interact with its placenta sibling from whom it becomes progressively detached.

8 For a full discussion of this see McKinley (1981) and Carsten (1987a: chap. 3). 9 This is an Islamic prohibition. 10 It is significant that if a baby has to be removed from the natal home before drinking its

mother's milk it should, instead, be given water cooked in the house hearth. 11 I use the term 'conception parents' for the parents responsible for the conception of a child,

and the term 'conception child' for the child produced by the parents' act of conception, in order to distinguish these relations from those of fostering.

12 The term literally means 'raised' or 'lifted' child, in the sense of having been taken up and put down elsewhere, see McKinley (1975: 61). He translates this as 'transferred child' to convey the broad scope of the Malay term.

13 Formal adoption, prohibited in Islam, is in fact very rare among Malays. Djamour (1959: 31, 34-5) notes the same lack of distinction being made among Malays in Singapore. See also McKin- ley (1975: 62-3), who links the ambiguity of terms to the conditionality and tentative quality of all forms of parent-child ties in Malay culture - in contrast to the categorical nature of siblingship.

14 See Goody's distinction between 'crisis' and 'voluntary' fostering (1982: 23). Rosemary Firth (1966: 105) has given the following statistics for Kelantan, which are surprisingly similar to those



JANET CARSTEN 441

for Pulau Langkawi (Carsten 1987a: 119): 25 per cent. not living with both parents; 11 per cent. living with adults other than parents; 14 per cent. living with a divorced or widowed parent.

15 This fact accords with villagers' oft-repeated assertion that they value male and female children equally. Nevertheless, there is a slight imbalance in favour of girls, which is more pronounced in the cases of grandchildren being raised by their grandparents. The imbalance is even more marked in Massard's material. She stresses (1983: 104, 111) how fostering involves the transfer of female children between women.

16 H. Geertz (1961: 40-1) cites a preference for giving a child to a female relative and especially the parents' siblings, particularly the mother's sister. Djamour (1959: 93) observes that grandparents frequently bring up children. Jay (1969: 72) states that in Java siblings have rights to a person's children, while Swift reports that in Negeri Sembilan women have rights to their sister's children (1965: 12). McKinley (1975: 103-6) shows how transfers of children between siblings are the most frequent, followed by the transferral of children to their grandparents. Massard (1983: 103) states that children are most frequently fostered by their mothers' sisters. Throughout this area, then, the rights of grandparents and of the mother's sister to children seem to be particularly strong.

17 The importance of feeding in creating kinship may be quite explicit. McKinley (1975: 62, 180) reports how transferred children say that their adoptive parents are their real parents because these were the ones to feed them. It might be suggested that in view of women's crucial role in the creation of kinship, Langkawi kinship could be labelled 'matrifocal' as, for example, Hildred Geertz (1961) has proposed for Java. While fully acknowledging the importance of women's creative role in constituting relations, I hesitate to use this term which seems misleading in view of the enormous significance of siblingship.

18 This point has been stressed by Barraud (1979) and Massard (1983) amongst others. Barraud (1979: 187-203) discusses how in the Moluccan society of Tanebar-Evav, marriage involves the circulation of women, and adoption the circulation of men. Adoption involves the relation between houses, it assures their continuity when marriage fails to do so. Massard (1983) has also described adoption as an alternative to marriage. In this community in Pahang girls are exchanged between women at adoption, and men circulate at marriage. See also Cunningham (1964), who describes how the loaning of children to parents among the Atoni reaffirms a fragile tie. I would disagree with Maeda (1975), who sees the frequent bringing up of grandchildren by grandparents among Malays in terms of a dyadic tie between them.

19 The details of this dispute are not relevant here. I have described them in Carsten (1987a: 395-408).

20 See also Massard (1983), who views adoption in a similar light. It is significant that on my return to the village in 1988, I discovered that this child had returned to her parents' home when the families involved in this dispute had made their peace a few years after these events.

21 Goodenough (1970: 405) has used the notion of 'sharing' rights to children to emphasise that what is involved is not a surrender of parental rights. He stresses the importance of distinguishing different kinds of parenthood transactions.

22 A similar stress on the production of children is found in other Southeast Asian systems. Hildred and Clifford Geertz (1964) have discussed how, in contrast with African systems which appear always to 'look upwards' towards past generations and an apical point, Balinese kinship can be seen as 'downward-looking'. They have shown how teknonymy plays an important role in this stress on future generations. Errington (1987: 420-22, 436) has discussed the importance of children and grandchildren in converting affinity into consanguinity in the Centrist Archipelago of Southeast Asia. See also Waterson (1987: 106-7), who links a close relation between baisen among the Toraja to teknonymy and a downward-looking attitude to descent.

23 See also Cunningham (1964) on the importance of children as mediators among the Atoni. Children do not have a fixed position: they combine different elements and are not settled in a descent group. They can mediate both between descent groups and between the living and the dead.

24 Aries (1979: 103-4) has discussed how children were in some sense immune to sexuality in sixteenth century Europe. It is clear that the association he makes between children and irrational- ism (1979: 116), and which he sees as belonging to twentieth century Western history, may occur in other contexts.

25 I am indebted to Marilyn Strathern for suggesting the ideas expressed in this paragraph.



442 JANET CARSTEN

REFERENCES

Aries, P. 1962. Centuries of childhood. Harmondsworth: Penguin. Barraud, C. 1979. Tanebar-Evav: une societe de maisons tourne vers le large. Paris: Cambridge Univ. Press. Boon, J.A. 1977. The anthropological romance of Bali 1597-1972. Cambridge: Univ. Press. Carsten, J. 1987a. Women, kinship and community in a Malay fishing village on Pulau Langkawi,

Kedah, Malaysia. Thesis, University of London. 1987b. Analogues or opposites: household and community in Pulau Langkawi, Malaysia. In

De la hutte au palais: societis 'a maison' en Asie du Sud-Est insulaire (ed.) C. Macdonald. Paris: CNRS.

1990. Women, men, and the long and the short term of inheritance in Pulau Langkawi, Malaysia. Bydr. Taal-, Land- Volkenk. 146.

n.d. Houses in Langkawi: stable structures or mobile homes? Unpublished m.s. Cunningham, C. 1964. Borrowing of Atoni children: an aspect of mediation. In Symposium on new

approaches to the study of religion (ed.) J. Helm. Seattle: Univ. of Washington Press. Djamour,J. 1959. Malay kinship and marriage in Singapore (Lond. Sch. Econ. Monogr. social Anthrop.

21). London: Athlone Press. Errington, S. 1987. Incestuous twins and the house societies of Southeast Asia. Cult. Anthrop. 2,

403-44. Firth, R. 1966. Housekeeping among Malay peasants (2nd edn.) (Lond. Sch. Econ. Monogr. social

Anthrop. 7.) London: Athlone Press. Geertz, H. 1961. TheJavanesefamily: a study of kinship and socialization. Glencoe, Ill: Free Press.

& C. Geertz 1964. Teknonymy in Bali: parenthood, age grading and genealogical amnesia. J. R. anthrop. Inst. 94, 94-108.

Goodenough, W.H. 1970. Epilogue: transactions in parenthood. In Adoption in eastern Oceania (ed.) V. Carroll. Honolulu: Univ. of Hawaii Press.

Goody. E. 1982. Parenthood and social reproduction:fostering and occupational roles in West Africa. Cambridge: Univ. Press.

Headley, S. 1983. Houses injava: the missing kin. Unpublished paper presented at seminar on Cognatic forms of social organisation in Southeast Asia. University of Amsterdam.

1987a. The body as a house in Javanese society. In De la hutte an palais. societis 'a maison' en Asie du Sud-Est insulaire (ed.) C. Macdonald. Paris: CNRS.

1987b. The idiom of siblingship: one definition of 'house' societies in Southeast Asia. In De la hutte au palais: societes 'a maison' en Asia du Sud-Est insulaire (ed.) C. Macdonald. Paris: CNRS.

Hooykaas, C. 1974. Cosmogony and creation in Balinese tradition. The Hague: Martinus Nijhoff. Jay, R.R. 1969.Javanese villagers: social relations in Modjokuto. Cambridge, MA: MIT Press. Laderman, C. 1983. Wives and midwives: childbirth and nutrition in rural Malaysia. Berkeley: Univ. of

California Press. Levi-Strauss, C. 1969. The elementary structures of kinship (rev. edn). Boston: Beacon Press.

1979. La voie des masques. Paris: Plon. 1983. Histoire et ethnologie. Annales 38, 1217-31. 1984. Paroles donnees. Paris: Plon.

Macdonald, C. (ed.) 1987. De la hutte au palais: societes 'X maison' en Asie du Sud-Est insulaire. Paris: CNRS.

Maeda, N. 1975. Family circle, community and nation in Malaysia. Curr. Anthrop. 16, 163-6. Massard, J. 1983. Le don d'enfants dans la societe malaise. Homme 23, 101-14.

1985. The new-born Malay child: a multiple identity being.J. Mal. Br. R. Asiat. Soc. 58: 2, 71-84.

McKinley, R. 1975. A knife cutting water: child transfers and siblingship among urban Malays. Thesis, University of Michigan.

1981. Cain and Abel on the Malay Peninsula. In Siblingship in Oceania: studies in the meaning of kin relations (ed.) M. Marshall (ASAO Mongr. 8). Lanham, MD: Univ. Press of America.

Schneider, D.M. 1984. A critique of the study of kinship. Ann Arbor: Univ. of Michigan Press. Snouck Hurgronje, C. 1906. The Achehnese (2 vols.) Leiden: Brill. Swift, M.G. 1965. Malay peasant society in Jelebu (Lond. Sch. Econ. Monogr. social Anthrop. 29).

London: Athlone Press.



JANET CARSTEN 443

Waterson, H.R. 1987. The ideology and terminology of kinship among the Sa'dan Toraja. Bijdr. Taal-, Land- Volkenk. 143, 87-112.

1990. The living house: an anthropology of architecture in SoutheastAsia. Singapore: Oxford Univ. Press.

WiLkinson, R.J. 1957. Papers on Malay customs and beliefs.J. Mal. Br. R. Asiat. Soc. 30: 4, 1-79.

Des enfants a part: I'adoption et le processus de la parente au Pulau Langkawi, Malaisie

Resume Cet article analyse la facon dont la parente est creee dans et par la conception et l'education des enfants parmi les malaisiens de l'lle de Langkawi. Au Langkawi, les ides a propos de la prise des repas en commun et le partage dans la maison sont aussi fondamentales pour la parente que les idees concernant la procreation. Ces ides sont particulirement associees a la notion de meme parente, representant l'unite des enfants originant de l'interieur d'une maison. Et pourtant les enfants passent la plupart de leur vie entre des maisons et sont des mediateurs importants dans les echanges entre elles. Ce role d'intermediaires est particulierement clair dans les arrangements d'adoption tres repandus. Une analyse de ces arrangements montre qu'ils sont concernes centrale- ment par la conversion de lieus d'affinite en liens consanguins, et par le maintien de l'egalite entre les parents conjoints selon la loi. Ainsi des enfants representent des points de transformation entre les deux images contraires de la communaute, une bas&e sur l'unite de la maison, l'autre sur la pluralite de l'6change entre les maisons. Dans leur croissance et leur mouvement, les enfants incarnent le processus de la parente.

Department of Social Anthropology, University of Manchester, Roscoe Building, Brunswick Street, Manchester M13 9PL, England



Carsten, J. (1991). Children in Between- Fostering and the Process of Kinship on Pulau Langkawi,...

Documents

Transcript of Carsten, J. (1991). Children in Between- Fostering and the Process of Kinship on Pulau Langkawi,...