Marriages, in Palokhi, are not arranged in the sense that parents choose spouses for their children. Young men and women make the choices themselves after a period of courting that is initiated by men. Courting, which is idiomatically called “to go and visit young women” (lae ha’ mykoe’nau), usually takes place during the planting and harvesting seasons when there is a great deal of co-operative labour exchange between households in Palokhi swiddens and wet-rice fields.^[9] As there are no prohibitions on village endogamy or exogamy, this also applies to other villages where the Palokhi Karen go to assist in the work on a labour exchange basis.^[10] Often, however, during the slack season in the agricultural calendar, visiting between households in Palokhi itself and between villages occurs, and young men also take such opportunities to woo young women they are interested in.

When a young man and woman are certain that they do want to get married, they then inform their respective parents who, by this time, usually have an inkling of their intentions. Parents in Palokhi say that they generally do not object to their children’s choice of spouses unless these choices contravene the rules of marriage in Palokhi. Some Palokhi parents do say that, apart from this major consideration, they would object under two circumstances: opium addiction in the case of a prospective son-in-law and being an unmarried mother in the case of a prospective daughter-in-law. A potential son-in-law who is likely to be, or already is, an opium addict represents a liability as the Palokhi Karen know only too well from their experiences in Palokhi and other villages in the area. Not only is he likely to be incapable of contributing his labour to agricultural production, but he may well attempt to dispose of family possessions or livestock to obtain the wherewithal with which to buy opium.

A woman with an illegitimate child, however, represents a liability of quite another kind. Apart from the moral reprobation which her condition provokes, an unwed mother is an anomaly which forcibly confronts the ideas and beliefs which the Palokhi Karen hold about marriage and the status of offspring from a conjugal union.^[11] Yet, at the same time, many Palokhi parents say that they feel a great reluctance to object to their children’s choice of spouses (whatever the reasons and this includes matches which would contravene the rules which govern marriage in Palokhi) because they fear that the couple, especially if they are not susceptible to dissuasion, will commit suicide (kha lau sa’, literally, “to kill down [one’s own] heart”).

Some parents and elders, elaborating on this fear, express their concern in terms of the limited repertoire of Buddhist concepts which they have acquired through their on-going contacts with Northern Thai. Specifically, they say that they would be “affected by demerit (ba’ ba’).^[12] On the other hand, some also say that those who commit suicide become transformed into malevolent spirits (tamyxa) which would be a danger to the villagers. Thus, while parents and elders may feel an obligation to prevent, or at least object to, a match which they regard as undesirable, nevertheless, children who are bent on their choice of spouses can, even in such circumstances, have their own way, namely, by seeking recourse in the threat of suicide. This does not, by any means, imply that such threats are devoid of real intent. There have been sufficient precedents in Palokhi (and elsewhere) to justify the apprehensions of parents in this regard. Ultimately, of course, this means that parents and elders have no real means of enforcing their objections, which is of some consequence in the context of infringements of marriage rules in Palokhi.

If, however, parents feel that a match is acceptable, then the onus of initiating formal discussions lies with the parents of the man. This is done through a go-between (a man) appointed by the man’s parents. He approaches the parents of the woman to ascertain that they consent to the marriage, after which various details of the marriage are worked out such as the time when the marriage ceremony should be held (which should not fall in an odd month), the size of the pigs to be slaughtered for the ceremony, the amount of rice liquor to be brewed and the time that this would take, and who should constitute the party of elders essential to the ceremony.

As there is no system of payments, apart from certain prestations made to the woman by the man and which are fixed by tradition, these discussions therefore do not entail bride-price or bride-wealth negotiations. Discussions can, however, stretch over a period of two or three months, mainly because of financial considerations, and whether or not the parties concerned are able to muster enough rice (for making the liquor and for feeding guests at the marriage ceremony) as well as money for the purchase of pigs if they do not have pigs available. One other important consideration in contracting marriages is whether the elder siblings of the man and women are married or not. The Palokhi Karen say that children should get married according to their birth order (see also Madha [1980:58]). If a person marries before his, or her, elder siblings, then he or she has to present them with a pig each before the marriage can take place. This, therefore, is a consideration which can also prolong the period of discussions initiated by the go-between before a marriage can take place.

While the foregoing are general considerations which attend marriages in Palokhi, the most important consideration is, of course, the rules which govern marriage in Palokhi.^[13] These rules may all be subsumed under a general prohibition on “crooked marriages” (thau pgha ke’ko), as the Palokhi Karen call it.^[14] The Palokhi Karen do not offer any categorical definitions of what would constitute “crooked marriages”, though from the examples cited by them (see below) it is apparent that such unions are those which result in the superimposition of affinal kin links over consanguineal kin links. Some Palokhi Karen, however, do say that “crooked marriages” are those where, metaphorically, “elder brothers become younger brothers, elder sisters become younger sisters” (waecau’ kae’ py, waenau kae’ py), or where the parties to such unions “love (each other), mutually prying apart that which is closely bound together” (‘ae’ khoekhae’ lau sa’). The first formulation, or rationalisation, refers to the contradictory applications of kin terms which would result or, what Freeman has called, “dysnomia” (1960:74, 161 n. 17). The second, interestingly, refers to what are thought to be the divisive effects of such unions where not only is there a confusion in the application of kin terms but also in the order of commensalism in the ‘au’ ma xae ritual which I discuss later. Such a blanket prohibition would, therefore, rule against virtually all inter- and intragenerational unions within the range of kin, recognised by the terminological system, who are genealogically traceable or “operative” kin to use Firth’s term (1963).^[15] While this may seem improbable — and it is certainly true that in some societies dysnomia may be no bar to the marriage itself — nevertheless, the fact of the matter is that the Palokhi Karen express the prohibition on “crooked” unions in just this manner. It is also believed that the families of the couple in a “crooked marriage” will be “affected by their xae” (ba’ xae; see Appendix A on ‘au’ ma xae).

At the same time, however, the Palokhi Karen also point out that “crooked marriages” are “destructive” (tahaghau) because they invite sanctions from the Lord of the Water, Lord of the Land. In general, these sanctions manifest themselves in ways where the natural order of things is upset or reversed, such as “tigers entering the village” (bauso’ ny’ zi pu).^[16] It is significant, however, that the Palokhi Karen specifically say that “the village becomes hot, the land becomes hot” (zi ko, hau ko or ko ba’ zi, ko ba’ hau, “the village is affected by heat, the land is affected by heat”) resulting in the “destruction of the rice crop” (by haghau). There is, clearly, a conceptual association between marriage and the cultivation of crops.

In practice, it would appear that three kinds of unions constitute “crooked marriages”. These are: cousin marriages, marriages between traceable kin across generations and multiple marriages between two hitherto unrelated families where the relative birth orders of the siblings concerned cross-cut each other. The first two cases are shown in Figure 3.2, extracted from an actual genealogy in which two such “crooked” unions did occur in Palokhi.

When Ty’ married Chi’ Ghe, the situation was clear enough. ‘Ae Nae’ used the term py in respect of Ty’ and the term for younger brother’s wife, demy, in respect of Chi’ Ghe. Nu’ was pyde to Ty’, and ‘Ae Nae’ was waenau to Ty’, Chi’ Ghe and Nu’. Lo’, on the other hand, referred to and addressed Ty’ and ‘Nu’ as phati, and Chi’ Ghe was called mygha, while she herself was phodo’ to all of them.

Although in different generations, the age difference between Lo’ and Nu’ was not great, and when they wanted to get married Lo”s parents and the village elders objected on the grounds that it would be a “crooked marriage”. The reason adduced in this case was that the marriage would result in a variety of conflicting kin terms in use: Nu’ being either pyde to Ty’, or phodo’ by virtue of being Lo”s husband, ‘Ae Nae’ would be mipgha and would have to be addressed as mo whereas she was waenau prior to the marriage, and so forth. When the intended marriage was opposed by the elders in Palokhi, Nu’ and Lo’ attempted suicide by drinking insecticide. Although the attempt was unsuccessful, the village elders remained adamant, despite considerable misgivings lest Nu’ and Lo’ should repeat the attempt. Nu’ and Lo’, however, presented the elders with a fait accompli by leaving the village for a nearby unit of the Royal Forestry Department where Nu’ managed to find employment as a wage labourer. Their parents and the village elders were forced to acknowledge the de facto union, but the headman and elders demanded that an expiatory rite be performed to appease the Lord of the Water, Lord of the Land. This was done, and a simple wrist-tying ceremony followed in lieu of the marriage ceremony which could not be performed under the circumstances.

The second union in Figure 3.2 (Pi No’-Pau’), is clearly a union between first cousins. The general consensus of opinion among the Palokhi Karen seems to be that cousin marriages are not permissible. First cousin marriages are definitely prohibited but, nevertheless, there are some who say that cousin marriages, beyond the first cousin range, are permissible. Those who say that marriage between cousins of whatever degree is prohibited allude to the old saw:

which, roughly translated, means “first cousins fall (?) close together in the (same) navel, second cousins taken back (must be) caught back”. The first line of this couplet quite unequivocally points to the closeness of first cousins who, therefore, are unmarriageable. The second line, however, gives rise to ambivalent interpretations in Palokhi. Those who regard all cousin marriages as prohibited unions interpret the second line to mean that the “catching” of second cousins is too difficult to achieve and, therefore, indicates that second cousin marriage is not permissible. Others who say that cousin marriage (beyond the first cousin range) is permissible take the second line to mean that “catching” a second cousin, although difficult, is possible and that, therefore, cousin marriage is permissible.

Even for those who say that cousin marriage is possible, there is however disagreement over whether this is possible at the second or third cousin range. It is interesting to note, in this regard, that some Palokhi Karen say this would depend on residential arrangements regardless of the range: they say that if physical proximity such as living in the same village or nearby lead to on-going intimate social relations, then cousin marriage under these circumstances should not be permitted. There is, nevertheless, a generally consistent agreement that cousin marriage (beyond the first cousin range) is possible only if the man belongs to an “elder” or “senior” dau’takhwa, and the woman to a “younger” or “junior” dau’takhwa, where “elder/senior” and “younger/junior” are defined in terms of the birth order of the parents or grandparents in the relevant connecting ascending generation. The rationale invoked to support the alleged permissibility of cousin marriage in this form is that when the couple perform the ‘au’ ma xae rituals (as they will have to eventually), the man has to eat before the woman and this would at least be consistent with the rule that food in the ritual should be partaken of according to sibling order. According to this formulation of permissible cousin marriage, therefore, same sex siblings would only be able to marry cousins in dau’takhwa forbidden to their different sex siblings (see Figure 3.3). It must be stressed, however, that this is by no means a prescribed or even preferred form of marriage. All it implies is that if cousin marriage is at all allowed, then it should take this form.

The first cousin marriage shown in Figure 3.2, therefore, goes against all the views on cousin marriage in Palokhi because it is, first, a union between first cousins and, second, because Pau’ belongs to a “senior” dau’takhwa since her father, Wa’ is elder brother to ‘Ae Nae’, the mother of Pi No’. The circumstances leading to this “crooked” union were no less interesting than those surrounding the Nu’-Lo’ union. Pau’, though unmarried, became pregnant when she was living at a unit of the Royal Forestry Department several kilometres away where her father was working as a wage labourer. Pau’ claimed that Pi No’ was responsible for her condition which he apparently denied vehemently. Wa’, however, pressed the matter with the parents of Pi No’ (reasoning that a prohibited union for his daughter would be less shameful than her being an unmarried mother). The Palokhi elders decided that an expiatory rite should be performed and the two could then live as husband and wife after a wrist-tying ceremony similar to that for Nu’ and Lo’. For reasons best known to himself, Pi No’ consented to these arrangements while still protesting his innocence. He added, however, that as proof of his innocence, the child that Pau’ was carrying would die on being born. As it turned out, the child did die but Pi No’ did not carry the matter further. This union was, needless to say, regarded as a ke’ko union.

The third prohibited form of marriage is shown in Figure 3.4. If Ri’ were younger than By Zo, or if La’ were younger than Mau’ Ghe, the use of kin terms before and after the two marriages would be clearly unambiguous. Ri’ would address La’ as wae by virtue of his being the husband of La”s younger sister Mau’ Ghe, and this would also be the term he would use in respect of La’ as his elder sister’s husband. By the same token, the other members of the two marriages would have no difficulty in using the appropriate kin terms since these would all be consistent with one another if the two sibling orders were not crossed in marriage. As it is, Ri’ is older than By Zo, and La’ older than Mau’ Ghe. Ri’ married Mau’ Ghe first and, some time later, La’ and By Zo decided that they wanted to get married. This was met with objections by their respective parents and the village elders for the reasons I have already discussed, including the dysnomia which would result. For instance, La’ could be either wae to Ri’ as Mau’ Ghe’s elder brother, or ca’li as the husband of Ri”s younger sister, and so on. Moreover, the dysnomia would not be confined to these people (and their siblings) but would also extend into the next generation. Their children, Ki Da and Phan, could be either waenau/waecau’ and py to each other, or neither. When it became clear that La’ and By Zo were determined to get married, By Zo’s father attempted to apply the maxim “Those who love each other prying apart that which is closely bound together, cannot drink water on the same steps” (ae’ khoekhae’ lau sa’, ‘au thi toe tauxau toe’ se) literally by giving them a bag of rice and ordering them to leave the village. After three days, the hapless couple returned and begged to be taken in. Duty done, and after consultation with the village elders and headman, By Zo’s father accepted them into his house when the appropriate rituals had been conducted for the Lord of the Water, Lord of the Land.^[17]

It is obvious, from the cases which I have described above, that the rules which govern marriage as they are conceived by the Palokhi Karen are violated. Indeed, these cases were referred to by my informants in Palokhi to point out where the rules should not be contravened. This, however, raises two important issues in the operation of the kinship system in Palokhi: first, what is it about the structure of Palokhi kinship which would allow for the possibility of the contravention of these rules; second, what is there in the system of authority and social control which fails to prevent infringements of marriage rules, given the fact that “crooked marriages”, as they are defined in Palokhi, are said to be prohibited.

To answer the first question: we have seen that the Palokhi Karen rationalise the prohibition on “crooked marriages” by referring to, amongst other things, the confusion arising from the contradictory application of kin terms — or what Tambiah has called, in dealing with a similar type of problem in Northeastern Thai ethnography, “an unacceptable linguistic asymmetry between husband and wife” (1973:131). There is, however, an important difference between the Thai and Karen cases. In the Thai case, husbands address their wives by the term naung (“younger sister”) while wives use the reciprocal term phii (“elder brother”). The “unacceptable linguistic asymmetry” that Tambiah discusses arises, therefore, when a man is younger than his wife and thus pertains to relative age differences between husbands and wives.

In the Karen case at Palokhi, the kinds of terms employed by husbands and wives have nothing to do with sibling terms. Instead, they constitute a set of terms in their own right within the corpus of kin terms. The linguistic asymmetry, which is rather more of a contradiction as the Palokhi Karen see it, therefore does not concern spouse terms and relative age but, rather, the kinds of terms employed with respect to, and between, traceable kin connected with “crooked marriages”. In other words, it is the result of the superimposition of affinal links on kin links prior to the unions which constitute the domain of difficulty. “Crooked marriages”, moreover, are implicitly defined in terms of previous marital unions to which the parties in “crooked marriages” are linked either consanguineally or affinally. Furthermore, where there is a broad rule such as the Palokhi Karen have against the contradictory application of kin terms, then such a rule must logically rest on the kinds of kin recognised in the society through its kinship terminology. The conclusion which may be drawn from this, therefore, is that marriage with genealogically traceable kin is undesirable.

Given, then, a general preference for marriage with non-kin (however, obliquely this may be expressed in Palokhi), this nevertheless leaves the problem of cousin marriage unresolved. If we consider the logic of the restricted form of cousin marriage described earlier, in terms of dysnomia, it is apparent that even in such cousin marriages dysnomia would result. For instance, if a man marries a woman from a “junior” dau’takhwa, there would still be a situation where the use of affinal kin terms would clash with the system of address for cousins which, in theory, is based on sibling terms applied per stirpes from the connecting ascending generation. Apart from considerations which have to do with ‘au’ ma xae, the explanation for this inconsistency lies in another aspect of Palokhi kinship.

In cognatic type systems, the application of kin terms may be extended indefinitely so long as genealogical connections may be traced. The way in which the Palokhi Karen recognise cousinship could, hypothetically, be applied ad infinitum but, as I have suggested, the notional limit to the reckoning of cousinship seems to lie at the third cousin range. There is a good reason for this: the Palokhi Karen do not have very deep genealogical memories. The use of kin terms for kin, and non-kin, in ascending generations, the strong social sanctions which actively discourage children from referring (much less addressing) kin and non-kin in ascending generations by name (if they happen to know them), and the practice of teknonymy all serve to produce “genealogical amnesia” (Geertz and Geertz [1964:101]). Most Palokhi Karen do not know the names of their parents or grandparents, so that for them to be able to trace third cousins to a common forebear by name is very unusual indeed. Of course, in the Palokhi system, to be able to classify someone as a second or third cousin, it does not require a person to know who the common ancestor was. All he or she would need to know is how his or her parents addressed, or referred to, the parents of the person concerned. Given, however, that relative age sometimes overrides this system of address for cousins, and teknonymy, cousinship beyond the first cousin range under these circumstances tends to be a vague relationship. This holds true especially when the appearance of new generations, the movement of households and the absence of prolonged social contacts lead to the gradual sloughing off of otherwise traceable genealogical connections. This, of course, would apply equally to kin other than second or third cousins.

Looked at from this perspective, the “problem” of cousin marriage takes on a quite different complexion. The ambiguities over cousin marriage arise because of difficulties in defining where cousins become sufficiently removed to be non-kin. And the restricted form of cousin marriage is just what it implies: it is a means, at least in a functional sense, of restricting marriages between cousins and reducing the superimposition of affinal links over consanguineal links to a minimum. In terms of the logic of the rationale of dysnomia, this is wholly consistent. But, note that this consistency is achieved by invoking a different principle; this principle, which underlies the restricted form of cousin marriage, is sibling order in the ascending connecting generation.

The rules of marriage and definitions of who constitute eligible and ineligible mates in Palokhi do not, therefore, form a neat or tidy set of proscriptions and prescriptions.

The foregoing also points to the significance of the role of parents and elders in determining whether prospective unions should be allowed or not. The reason for this is that they possess more genealogical information than those in descending generations. This is due to the general properties of the kinship system and teknonymy which lead to gradual losses of such information through successive generations.^[18] This is not to say, however, that there is an absolute attrition of genealogical information over time. Each generation acquires a certain amount of information from existing operative kin links, but this is slowly lost with the appearance of successive generations where new information comes into existence; it is this which reproduces the continuities and discontinuities — of which the ambiguities over cousin marriage are a part — of the system.

We can now turn to the question of why the possibility of infringements of marriage rules occurring cannot be prevented with absolute certainty. Headmanship and the position of elders in Palokhi do not carry with them any capacity for social controls and punitive action. There are, consequently, no real means available for enforcing the rules of marriage in Palokhi, nor are there any effective sanctions which could act as deterrents to the infringement of these rules. In the final analysis, however, it is the fact that the Lord of the Water, Lord of the Land may be placated through propitiatory rites, as a last resort, which allows for the possible occurrence of “crooked marriages”.