Despite the general rule on uxorilocal residence upon marriage, in Palokhi, this rule does not lead to the formation of extended families consisting of parents, married daughters and their husbands and children. The reason for this is that the Palokhi Karen stipulate that a house should not contain more than two married couples on a permanent basis. Thus, if in a two generation family with one married daughter, another daughter marries, then the daughter who married first and her husband (and children if any) must make way for the second daughter by moving out of the natal family and building a separate house for themselves. On the other hand, if it is a three generation family to begin with, then even the first daughter who marries and her husband cannot expect to remain indefinitely in the house of her natal family. She and her husband would usually move out within a year of being married. Married daughters who move out of their parents’ houses usually build houses near their parental homes, although virilocal and neolocal residence have been known to occur. The reasons have been mainly economic, having to do with wage labour opportunities available at the various units of the Royal Forestry Department in the area, and personal preferences as well.

The Palokhi Karen offer no satisfactory reasons for the limitation on the composition of stem family households beyond saying that “it must be done so” (ba’ ma di’ ne), although some informants say that the reason for the limitation is that otherwise there would be “too many people” for the ‘au’ ma xae ritual.^[19] In fact, however, the reason for the limitation on household composition becomes readily apparent when we examine the sleeping arrangements of households in Palokhi. Houses have one main room, and this room is roughly divided in half by a hearth, as Figure 3.6 shows. In a nuclear family household, the parents and children would sleep in the rear half of the room. As their children grow older, pubertal daughters sleep in the front half of the room, while pubertal sons sleep either on the inner verandah or the rice barn, or in other rice barns along with their age-mates.

In traditional times, “communal houses” (blau) served the purpose of accommodating young men who were expected to sleep on their own, as well as other ritual functions (Marshall [1922:139]); but, in Palokhi, there is no such communal house, hence the practice of young men sleeping in rice barns or inner verandahs. When a daughter in a nuclear family marries, the front half of the room is given to her and her husband to sleep in, thus maintaining a spatial, and therefore symbolic, separation of the two married couples in the stem family that is formed. With the marriage of a second daughter, such a separation of all the couples would be impossible to maintain if the second daughter and her husband were to reside in the house as well, given the structure of Palokhi houses. In the same way, in a three generation (stem) family, such a separation would not be possible to maintain when a woman married, because her grandparents would occupy the rear half of the room, and her parents would occupy the front half of the room.^[20]

This is not to say that house form in Palokhi determines the composition of households by any means, but, rather, that sleeping arrangements which are conditioned by house form are taken into account in an ideology of kinship which emphasises the discreteness of conjugal pairs, even within stem families.

Palokhi residence customs, however, are a little more complicated than I have indicated in the foregoing. If a mother dies, the house is taken apart and burnt, and the father then builds a new house in which he and his children live. The Palokhi Karen say that the reason for this is that “the house belongs to the wife” (phau’mypgha ‘a’ doe’), at least, initially. The house which a widower builds, however, “belongs” to him, so that when he dies the house is likewise destroyed. If, on the other hand, a father dies before the mother, the house is not destroyed, and the mother and her children may continue to live in the house.^[21] In either case, the rule that there may not be more than two married couples living in the same house still applies, with the surviving parent and deceased parent regarded as constituting a “union” nonetheless.

This condition on residential arrangements, the stipulation that there should not be more than two couples in a house, may in practice be circumvented in either of two ways. The first, a long term one and one which was resorted to on only one occasion in Palokhi, entails building a separate hearth (which, in effect, means extending the house) for the couple not ordinarily supposed to live in the house. This also requires the construction of separate step-ladders so that, for all practical purposes, the separate hearths and steps represent separate houses.^[22] The second device which is a short term one, and customary practice in Palokhi, involves one or the other couple sleeping on the inner verandah of the house until the new house for the first married daughter and her husband has been built and is ready for occupation.

In the event that both father and mother die, the house is destroyed and alternative arrangements are made. If there are any children who are married, then the unmarried children may live with either the married sister or married brother, although the former is more usual given the fact that the brother moves away to live with his wife. If there are no married children, then the children are fostered out to patrilateral, or matrilateral, kin depending on the circumstances: that is, whether they can be located easily or not, whether they can afford the upkeep of the children, and so forth.^[23] Whatever the resulting arrangements, married siblings cannot live in the same house when their parents are dead, and the reason for this lies in the way ‘au’ ma xae rituals are structured, as the Palokhi Karen conceive it, and which I discuss later. In sociological terms, however, this is understandable. First, the rule on uxorilocal residence at marriage which leads to brothers marrying out would act to prevent the formation of households consisting of co-resident married brothers, or married brothers and sisters. Second, the succession of married sisters resident in their natal homes, according to the condition on uxorilocal residence described above, would itself act to prevent the formation of households composed of co-resident married sisters.

Residential arrangements are, of course, part of the total process of household formation and household fission within the overall developmental cycle of domestic groups. This cycle, in ideal type terms, spans three generations. During the cycle, brothers marry out, and as sisters marry successively, nuclear family households are formed, while the natal family at any point in time retains a fixed structure, consisting of two conjugal pairs, regardless of whether it is a two or three generation stem family. The cycle ends with the death of grandparents, followed by parents, and what is left are siblings with independent nuclear family households which then repeat the process. This cycle in the family sociology of Palokhi, and its residential concomitants, are fully reflected in the structure and organisation of the ‘au’ ma xae ritual.

A full account of the ‘au’ ma xae ritual may be found in Appendix A. It is a quintessentially domestic ritual and it is an integral part of the religious life of the Palokhi Karen. As I show in my discussion of the ritual, it is relevant to a further understanding of the institution and rituals of marriage, the family sociology of the Palokhi Karen, as well as the issues which I consider next.