The Palokhi Karen believe in “souls” (koela), and most of them agree that each person possesses thirty-seven souls, although this number varies from one informant to another, and also from place to place (Kunstadter, pers. comm.; see also Mischung [1980:57]). These souls are located in various parts of the body. The most important soul is the koela kho thi’, the principal soul which resides in the head and the loss of which would result in death. The Palokhi Karen also believe in “spirits” (tamyxa) and these spirits are said to inhabit such features of the natural environment as the forest, streams, tracks, and exceptionally large trees that are withered and standing. These spirits may be malevolent or, at best, indifferent in their relations with people, but all are manipulable through ritual intercession.

In addition to these spirits is an “entity” called xae which is often described as bgha’ and is sometimes also equated with si kho myxa (see also Marlowe [1979:179]; Mischung [1980:82–3]).^[2] Xae has been variously glossed by other researchers who have worked among Sgaw Karen as “ancestral spirits of kin groups called dopuweh” (Iijima [1979:107]), “ancestral or house spirit” which is the “controlling authority, the guardian of jural rules, and the dispenser of punishment for their infringement” (Marlowe [1979:178]), “ancestor spirit” (Madha [1980:199]), and “ancestor” (Mischung [1980:82]). In Palokhi, xae is conceived of as a very amorphous “entity” if, indeed, it may be said to be an “entity” at all. Most Palokhi Karen assert that it is neither soul (koela) nor spirit (tamyxa), while some say that it may be one or the other, but they are, nevertheless, uncertain on the point. The various meanings attributed to xae referred to above, however, do not in my opinion adequately describe how xae is conceived of among the Palokhi Karen. What xae “means” to the Karen of Palokhi can only be understood through an examination of the details of the ritual associated with it.

The ritual of ‘au’ ma xae (which, literally, means “to eat, doing the xae” and is often abbreviated to ‘au’ xae, that is, “to eat xae”) falls within the larger body of rituals that have as their purpose the restoration of health and well-being to individuals. A large part of these rituals involves the re-establishment of a concord between the domain of souls and spirits, on the one hand, and the domain of people and the physical environment, on the other. Any dissonance between these domains — whether through the witting, or unwitting, actions of people giving offence to the spirits or through the malevolent designs of the spirits themselves — results in illness. In these terms, illness is really symptomatic of three general conditions: “soul loss” (ba’ tamyxa), “spirit invasion” (ba’ soeta), or of something “amiss” with the xae (ba’ ma xae or ba’ xae).^[3] When illness presents itself, divination is, then, necessary to determine which of these three conditions exists so that the appropriate ritual may be conducted to restore a person’s well-being. As with most forms of divination, divination in Palokhi is based on binary principles and analogic reasoning to “sort out” or eliminate, the possible causes of illness. In the Palokhi Karen scheme of things ba’ xae usually (though not always) stands last in the order of possibilities.

If divination indicates that the ‘au’ ma xae ritual is necessary, then further divination is required to determine whether it should be conducted by the mother or father of the person who is ill. Both parents may, however, decide to conduct the ritual, without recourse to divination. In either case, the mother has to conduct the ritual first followed by the father on the next night. Whether or not divination is resorted to, the elements in the ritual are the same. The personnel involved in the ritual, however, varies depending on generational depth, the composition of the family and, in the case of the patient, where he or she is located in terms of generational levels. In very general terms, this describes the basic considerations which inform the organisation of the ‘au’ ma xae ritual. In practice, however, there are several permutations and complexities arising from these considerations and these are best explicated by referring to specific cases.

The first case which I consider here concerns a married couple, with children, and whose parents are all alive. This case would, therefore, involve either a nuclear family (that is, where the couple has moved out of the woman’s natal home because a younger sister has married) or a stem family, consisting of parents and a married daughter with her husband and children, as a result of the custom of uxorilocal residence at marriage. Assuming that a child of the couple is in need of the ‘au’ ma xae ritual, and divination reveals that the ritual must be conducted by both parents, further divination is then necessary to determine whether the ritual must be conducted using chickens for the ritual meal (a “small” ‘au’ ma xae) or chickens and pigs (a “big” ‘au’ ma xae). Pigs and chickens are specially reared for the express purpose of ‘au’ ma xae rituals, and these animals may not be put to any other use. ‘Au’ ma xae pigs are called thau’ kho thi’, and the chickens chau kho thi’. These pigs and chickens are acquired by a couple after they are married for use in ‘au’ ma xae. In Palokhi, the man usually buys a sow and hen and then “gives” them to his wife to rear for ‘au’ ma xae purposes.

In Palokhi, there are slightly different interpretations as to what the “giving” of these animals means. Some say that these animals, therefore, “belong” to the woman, but others say that she rears them for the couple. To see the status of the kho thi’ animals in terms of “ownership” however, would be to miss an essential aspect of the way in which the Palokhi Karen maintain a conceptual separation between the domains of male and female activity. There is only a minimal sexual division of labour in Palokhi, but one area where male and female activities are marked is the domestic domain, in which the rearing of pigs and chickens and, to some extent, the preparation of rice are seen to be essentially female activities. A sow and hen are acquired initially for ‘au’ ma xae purposes because hereditary continuity of these animals is regarded, at least in theory, as being essential in conducting the ritual.^[4] The actual use of these animals in ‘au’ ma xae, however, depends on who holds the ritual and related considerations. If the wife holds the ritual, and both her parents are alive, then she uses a cock and hen corresponding to the sex of both her parents. If only one parent is alive, then she uses a fowl of the sex corresponding to that of the living parent. In the case of ‘au’ ma xae that require pigs as well, then only one pig is used — a sow corresponding to the sex of the woman herself.

There are eminently pragmatic reasons why strict symmetry in the sexual symbolism in the case of pigs, paralleling that for chickens, is not observed. The Palokhi Karen say that pigs are more expensive than chickens, they take longer to rear, and besides much of the pork would be wasted after the ritual because it would not keep as well. If the husband holds the ritual, the same strictures apply with regard to the use of chickens in respect of his parents. Similarly, in a “big” ‘au’ ma xae he would be obliged to use a boar. Nevertheless, even with all these stipulations, the animals cannot be used indiscriminately: due regard must be given to generational order and birth order in the lines of chau kho thi’ and thau’ kho thi’.

Furthermore, the pigs must be all-black and the chickens any colour but white. Given that the ritual may be held by the father and/or mother (husband and/or wife), the fact that there are two versions of ‘au’ ma xae, “big” and “small”, means that ‘au’ ma xae may be carried out within a single night or may stretch over four days. A full-blown “big” ‘au’ ma xae, for instance, would require the use of chau kho thi’ on the first and second nights, followed by the use of thau’ kho thi’ on the third and fourth nights.

Assuming that both parents are to hold the ritual, and that further divination indicates that a small ‘au’ ma xae should be performed, then the parents and their children must refrain from working on the days of the ritual. They must also wear “traditional” clothes. Towards the evening of the first night, the appropriate chickens are selected and killed. This may be done by either the father or mother.

It is worth noting here that the Palokhi Karen say that the blood of ‘au’ ma xae animals should not be spilled. The chickens are therefore killed in one of two ways. They either have their necks wrung or are beaten over the head with the flat of a bush knife until they are unconscious before they are cast onto the hearth where they are burnt to remove the feathers. Pigs, being larger animals, are despatched in a different manner. They are clubbed on the snout or head after which they are throttled to death by placing a piece of wood across their throats with someone standing on the two ends of the wood so as to apply the necessary pressure. This stricture on the shedding of blood of kho thi’ animals stands in striking contrast to the method of killing animals as offerings to the Lord of the Water, Lord of the Land and other spirits associated with rice cultivation and the expiatory rite for “crooked unions”. In these cases, blood must be let as part of the offertory process. The Palokhi Karen also say that care must be taken to ensure that the organs and entrails of the kho thi’ animals should not be damaged when the body cavity is slit open to remove them for cleaning. The chickens are cooked whole in a pot of water along with the the livers and hearts which are bound up with the intestines. The pigs are, of course, dismembered prior to cooking. Unlike the vegetable stews eaten on ordinary occasions, or the chicken and pork stews, which are cooked to feed helpers in cooperative agricultural tasks, the chicken or pork eaten during ‘au’ ma xae rituals is usually cooked without adding the myriad spices and herbs, as well as salt, so common to Karen cuisine.^[5]

When the chickens have been dressed and cooked, they are placed in a bowl together with the organs and intestines, and the bowl is set in the centre of a large circular eating tray. Rice is then spooned out around the bowl in the tray and the parents and their children will then sit around the tray or squat about the tray. As this first ‘au’ ma xae is held by the mother of the child, the mother’s parents are required to attend the ritual regardless of whether or not they are living in the same house. The parents of the woman sit a little away from the tray, and the ritual meal commences with the child’s father picking a small piece of chicken from a thigh of each fowl and a handful of rice, which he then eats. The child’s mother then does the same, followed by the children according to birth order.

When all the members of the nuclear family have eaten in this fashion, including the child who is ill, each of them takes a sip of water in the same order. The mother then calls to her parents, who have been sitting to one side throughout, to “come and eat, come and drink”. Her parents approach the tray, and her father takes a piece of flesh from the cock which he wraps, with some rice, in banana leaf. Her mother does the same, but takes the flesh from the hen. The pieces must all come from the same thigh from which the child’s father has initially picked. As they do this, they pray for the well-being of their daughter, son-in-law, and grandchildren specifying particularly the symptoms of the grandchild who is ba’ xae. These prayers are addressed to the souls of their respective parents whether living or dead. When they finish the prayers, they shove the wrappings of rice and chicken through the slatted bamboo floor of the house.

The following is an example of an ‘au’ ma xae prayer. It comes from an ‘au’ ma xae rite performed by Rae’, the ritual specialist in Palokhi, for one of his grandchildren. Rae”s wife had died several years before, while his son-in-law, Thi Pghe, had died in 1978. This particular ‘au’ ma xae, thus, involves a three generation stem family in principle (see below) but the prayer may be regarded as being representative of all ‘au’ ma xae prayers since there is little variation in these prayers. As with most prayers and ritual texts in Palokhi, this prayer is long and repetitive; I therefore present below only a portion of the prayer sufficient to show its principal features.

The prayer continues in this vein with Rae’ referring to the various conditions “afflicting” his two other grandchildren, and ends with the line “Remain with me for ten years, for one hundred months”.

Although the rite was specifically held because his youngest grandchild was sick, it is worth noting that the conditions of all the members participating in the rite are mentioned. It is also significant that the relationships between members are described in terms of generational dyads, as indicated by the possessive pronouns, and that they are referred to according to generational and then birth order. Rae’ begins by referring to himself, and then his daughter by the expression “my daughter”, followed by his grandson whom he refers to as “her son” (the possessive ‘a’ being attributive of coe’ phomy), and then only as his grandson. This is a good indication of how the Palokhi Karen conceive of generational relationships, that is, in terms of successive steps of affiliation. Note also that the participants are not referred to by any collective term; they are referred to individually, which is indicative of the degree of importance accorded to individual social identities notwithstanding the fact that the ritual expresses their commonality through commensalism. The specific references to subsistence activities is, again, of considerable significance; they point to a perception that the health, well-being, membership and subsistence activities of domestic groups are inter-related. The ritual is a “totalising” one: it contains a great many conceptual associations which cover various facets of the lives of individuals and domestic groups.

Once these prayers are completed, everyone proceeds to eat his fill without regard for the order of precedence that dictated the ritualised eating of the chickens or, as the Palokhi Karen express it, the “eating of the kho thi’”. During the eating of the kho thi’, however, all the participants must be extremely careful not to disrupt the proceedings by any sort of disorderly behaviour which might result in dropping the eating utensils or upsetting the cooking and water vessels. If any of these should happen, the very strict adherents of ‘au’ ma xae say that the ritual must be conducted again because it has become ineffective. Similarly, if a child should complain that it has not had enough to eat, then the ritual should be performed again. Most Palokhi Karen, however, eschew observing these conditions on ‘au’ ma xae on the grounds that if they were to do so, the ritual would be too burdensome as they would end up consuming their kho thi’ animals faster than they reproduce.

On the next day, the child’s father must conduct the ritual. Given the custom of uxorilocal residence at marriage, the father’s parents must be called to attend the ritual from wherever they happen to live. The ritual is conducted in the same way as on the first night, with the father’s parents addressing their prayers to the souls of their respective parents. If for some reason the father’s parents are unable to attend the ritual, then the father, as he eats the kho thi’, will address a similar prayer to his parents’ souls, while at the same time taking pieces of the appropriate fowl and some rice, and wrapping them separately in banana leaves. These bundles are kept aside for his parents in the eaves of the house, and after the ritual has been completed, they are taken to his parents at a later date. The rice and chicken which are kept aside are simply called “dried rice” (me xe). When the dried rice is sent to the parents, they eat it and say prayers for the well-being of their son and his family, just as they would had they been able to attend the ritual.

The foregoing description of ‘au’ ma xae with chickens applies equally well to ‘au’ ma xae with pigs. There is, however, one difference: ‘au’ ma xae with pigs usually includes divination which is called “divining at the gall-bladder of the pig” (ka thau’ ‘a’ soedi). If the gall-bladder is small, this means that the ritual will not be successful, although the ritual is nonetheless performed. If it is full, but is held by the mesentery across the junction of the two major lobes of the liver, then this is also taken as a sign that the ritual will be unsuccessful. If, however, the gall-bladder is full and is aligned with the junction of the two lobes, then it augurs well. If divination augurs ill, the Palokhi Karen say that the ritual should be conducted again, but they tend to ignore this for the same reasons that they tend not to observe the conditions on the ritual mentioned before.

‘Au’ ma xae (whether “big” or “small”) where the parents of both of husband and wife are all alive represents an ideal type situation which is rare in Palokhi. More common are situations in which at least one of the parents is dead. In such circumstances, the ritual is carried out in much the same way and the structure of the ritual remains the same: the surviving parent of either spouse attends the ritual and says the same prayer while making the offerings of rice and chicken, or pork, to the souls of his, or her, parents.

If, however, both the parents of the husband and/or wife are dead, then the ‘au’ ma xae ritual is conducted in a slightly different manner, although the kho thi’ is eaten in the same way described above. If the child’s mother’s parents are deceased, then there will be only one chicken, a hen (corresponding to the sex of the mother), used on the first night of the ritual. At the end of the eating of kho thi’, the mother will then take a piece of the chicken and some rice, and wrap them in banana leaf saying a prayer to the souls of her deceased parents for the well-being of the family. The offering is then shoved through the floor of the house, after which the family proceeds to eat its fill. Similarly, if the father’s parents are dead, then a cock is killed for the second night of the ritual, and the father will say a prayer and make the offering to the souls of his parents after the eating of the kho thi’.

The next case I consider is when both parents of the spouses are dead and when one of the spouses is also deceased. Under these circumstances, only the surviving parent performs the ‘au’ ma xae ritual. The order of commensalism remains the same, and at the end of the eating of the kho thi’, the father or the mother as the case may be, says the prayer and makes the offering to his or her parents’ souls.

If, however, both the husband and wife are dead, then the orphaned children can no longer have ‘au’ ma xae performed for them. In Palokhi Karen reckoning, it is not possible for orphans to be ba’ xae. If grandparents of the orphans are alive, they cannot perform the ritual for the orphans, nor can married siblings of the orphaned children. The Palokhi Karen offer no explanation for this beyond saying that it is parents who must perform ‘au’ ma xae for their children.

Thus far I have considered cases in which the personnel involved in the ritual span two or three generations, and the person who is ba’ xae is in the lower (or lowest) of generation. In the various permutations of these cases, the ritual is always held by the parents of the person who is ba’ xae. Another significant feature of the ritual is that these parents call upon their parents to attend the ritual. The latter then call upon the souls of their deceased parents to heal and restore the well-being of the descendants in need of the ritual. If the grandparents of the person who is ba’ xae are dead, then their souls are called upon by his or her parents to heal and restore the well-being of the family. In other words, the ritual entails a “connecting up” of generations in such a way that the parents in the last surviving generation (whichever the level) call upon the souls of their deceased parents. At each level, however, there is in the structure of the ritual an evident recognition of the bilaterality of parenthood. Or, in more general terms, the ritual complex, as a whole, appears to place an emphasis on generational continuity through filial relationships, traced bilaterally.

The features of the ritual described above are also present when the person who is ba’ xae is a parent. There is, however, an important difference in the organisation of the ritual in terms of who is required to attend and, here, a certain asymmetry emerges in the form of a matrilateral bias in the organisation of the ritual. For example, if the person or any of his or her siblings is ba’ xae then either of the parents or both (depending on the outcome of divination) will have to perform the ritual. Whatever it is, the ritual is conducted in the manner described earlier, that is, the kho thi’ is eaten first, followed by a prayer and offerings to the souls of the person’s grandparents. The order of eating, however, is a little different because the participants span three generations. The person’s father eats first, followed by the mother after which the person and his or her siblings eat according to birth order. If the siblings are married and have children, the children of sisters attend the ritual but the children of brothers do not. The spouses of siblings do not attend the ritual either. Accordingly, if the person who is ba’ xae is a woman with children, then her children are required to attend the ritual but not her husband. On the other hand, if the person who is ba’ xae is a married man with children, his wife and children do not attend the ritual. Participation in this kind of ‘au’ ma xae is, thus, based on links traced through the female line. The children of female participants eat immediately after their mothers. For instance, if the person who is ba’ xae has an elder and younger sister both of whom have children, then the order of eating the kho thi’, as far as they are concerned would be: eZ, eZ’s children (according to birth order), ego, ego’s children (also according to birth order, assuming of course that ego is a woman), yZ, and so on. Should a sister be dead, her children would nonetheless be required to participate in the ritual. It is only in ‘au’ ma xae of this sort (where the participants span three generations) that orphans participate, that is, by virtue of links through the female line.

The corollary to this kind of ‘au’ ma xae is, of course, to be found in ‘au’ ma xae conducted for a person’s father or his siblings, by that person’s paternal grandparents. He or she would, necessarily, be excluded from these ‘au’ ma xae by the principle that dictates attendance at the ritual in this larger, all-encompassing form, namely, the tracing of links of affiliation through the female line.

This principle in the organisation of the form of ‘au’ ma xae described above operates only in the context of the ritual. It does not constitute the basis for organising kin groups for any other purpose whatsoever in Palokhi. The parallels between this basis for organising ‘au’ ma xae in its more embracing form and the way in which the lines of pigs and chickens are maintained for the purposes of ‘au’ ma xae are, however, striking.

As I noted before, sows and hens are acquired, initially, by married couples for use in their rituals. Specifically, sows and hens are acquired because the females of the species bear offspring which the males do not. Furthermore, as some informants explained the matter, the only means of ascertaining continuity of lines was through the females of the species — because any number of males could copulate with them and, consequently, any attempt to trace continuity of these lines of kho thi’ animals through boars and cocks would be futile. In other words, the obvious facts of “maternity” and parturition make the female of the species a “natural” criterion by which the continuity of lines may be unambiguously traced because the reproductive process is physiologically “marked” in females, whereas is it not in the case of males.

Although the Palokhi Karen do not express it in any overt or even implicit manner, I suggest that the parallel ways in which lines are traced in ‘au’ ma xae animals, and people in the organisation of ‘au’ ma xae rituals (of the kind described above), are not coincidental. They are based on analogous principles — sexual reproduction through females, and the lack of ambiguity in tracing lines of “descent” (or affiliation) on the basis of the facts of “maternity” and parturition. Yet, although the principles are analogous, there is a major difference between the two in the ritual complex: ‘au’ ma xae rituals also involve — at each generational level of the participants — a recognition of the bilaterality of parenthood, that is, of maternity and paternity. ‘Au’ ma xae, as a ritual complex, therefore contains a parallelism based on the analogous principles of tracing “descent” on the basis of “maternity” and parturition, as well as an opposition based on “non-paternity” in animal reproduction and “paternity” in human reproduction.

Despite the importance of this ritual complex in the lives of the Palokhi Karen, they are nonetheless unable to provide reasons to explain the practice of ‘au’ ma xae. Nor are they able to recount a myth (in the Malinowskian sense of a “charter”) to account for the performance of the ritual, and which might provide a further basis for understanding the ritual. Mischung, however, has reported such a myth from Chom Thong. The myth is suggestive indeed and lends support to the interpretation of ‘au’ ma xae that I have advanced above. Part of the myth documented by Mischung, relevant to my argument, reads as follows:

As Mischung remarks, in his commentary following the myth (1980:87), the central piece of information being communicated in the myth is that without the rite, the Karen would be like animals instead of human beings. This is, indeed, the theme of the myth. But, it is significant that the theme should be expressed in terms of such an evident concern with the distinction between animality and humanity based on the sexual and reproductive behaviour of animals and humans, and on the issue of non-monogamous unions of which the Karen disapprove. The myth, thus, expresses precisely those concerns which, as I suggested, are embedded in the ‘au’ ma xae rituals of Palokhi.

These concerns are also to be seen in the notion of xae. I refrained, earlier, from trying to define xae because, as it should be obvious by now, it is not any easy thing to do — in the absence of unequivocal indigenous explanations — without taking into consideration the various permutations and complexities of the ‘au’ ma xae ritual. An examination of these permutations and complexities of the ritual does, however, bring us closer to an understanding of what xae “means”. Rather than calling the xae “ancestral spirit”, and so forth, I think it would be more accurate to regard it — despite the risk of some reification of the Karen concept — as a kind of shared “consubstantial essence” (or, perhaps, even a shared “state”) which comes into existence with the formation of a marital union and the birth of children.

To begin with, the linguistic evidence argues against interpretations such as “ancestor spirit”, because xae (although possessing the synonym bgha’) is not “soul” (koela) or “spirit” (tamyxa). Furthermore, xae is not invoked in the prayers said during the ritual. It is the souls of deceased parents in the immediate ascending generation that are invoked. Xae, therefore, stands on its own apart from these entities and, consequently, the indications as to what it “means” must be sought elsewhere in the ethnographic data.

The most pertinent of these data are to be found in certain aspects of the ritual. First, ‘au’ ma xae can only be perfomed by parents for their children, and not even grandparents or married siblings of the children may do so. Second, unmarried children cannot themselves perform the ritual and it is only when they marry and have children of their own that they may, then, perform the ritual even when their parents are alive. Third, and perhaps most conclusively, orphans cannot have ‘au’ ma xae performed for them. They only participate in ‘au’ ma xae under two conditions: when the ritual is performed by their maternal grandparents for siblings of their mother, and when they themselves marry and have children of their own. Finally, when persons are ba’ xae, the ritual is not conducted solely for them; the ritual also involves primary kin in the case of nuclear families holding the ritual, and lineal and collateral kin organised on the basis of links through the female line in the case of larger kin groups when the ritual is, nevertheless, performed by the parents of those who are ba’ xae. The commonality, or better still, the communality of those involved in ‘au’ ma xae is indicated by two important considerations: first, the transitive consequences of infringements of marriage rules (resulting in “crooked unions” in which kin of parties to such unions fall ill; second, by the enumeration in ‘au’ ma xae prayers, by name, of those (other than the person who is ba’ xae) involved in the ritual and their respective “symptoms” of “dis-ease”.

The weight of the evidence is, I think, sufficient to warrant the wider interpretation of xae that I have proposed. Nevertheless, we are left with one aspect of the ritual unresolved, namely, the status of xae in kin groups larger than the nuclear family. I think it is safe to say that xae is the same, and that its genesis and transmission lie in nuclear family formation, even when ‘au’ ma xae may include larger kin groups. The groups themselves are organised on “natural” principles, that is, the tracing of group affiliation on the basis of the facts of “maternity” and parturition. However, what makes ‘au’ ma xae a preeminently “cultural” thing is, as I argued earlier, the recognition of paternity in the bilateral structure of the ritual. It is, in other words, important to recognise that there are two rather different principles at work in the ‘au’ ma xae ritual.

If the ideology of kinship and ‘au’ ma xae could be said to be “sociological”, it would be so in a way which stresses the marital relationship and the procreation of children, rather than the larger kin groups which participate in the ritual. The sociological aspects of the kinship system, that is, marriage and residential patterns, household formation and fission, and the structure of ‘au’ ma xae make this clear. The ideology of kinship and ‘au’ ma xae also stresses the continuity of domestic groups associated through shared xae in a three generation cycle, which ends with the death of parents, whereupon the families of married siblings become fully separate units conducting their own ‘au’ ma xae when the cycle repeats itself.