No reciprocation? Wife-givers and wife-takers and the bartan of the samskara among the Khatris of Mandi, Himachal Pradesh Klaus Hesse The aim of this paper is to question the paradigm—which is also central to Parry’s theory of the ’Indian gift’-of the unilaterality of the gift structure among wife-givers and wife-takers in north India on the occasion of marriage and other samskara. On the basis of detailed ethnography, it will be shown that the gift giving system (bartan) of the urban Khatri in Mandi (Himachal Pradesh) does not substantiate this theory of unilaterality. In Mandi there is definitely an asymmetric, but not a clear-cut unilateral, structure. Even with regard to kanyadan, reciprocation seems to be more complex than the soteriological argument of Trautmann and Parry suggests. It will be further argued that, with regard to Mandi, proportional, asymmetric reciprocation is central to the system of bartan, which implies a gift exchange in Mauss’ categories. Bartan is conditioned by the content and structure of the rites de passage, by the general hierarchical social order, by the kin structure, and by asymmetry of giving, receiving and returning of gifts within the life cycle and over the generations. This paper is concerned with the structure of gift giving (bartan) among the Khatris of Mandi on the occasion of the life cycle rites (samskaras). Here, some basic structures of the system of prestations will be considered and, in outline, the gift relations of wife-givers and wife-takers and their kindred at marriage (vivaha samskara).’ The central problem, as indicated by the title of the paper, is the question of ’non-reciprocation’ within these relations.’ The thesis will be put forward that the generally accepted picture of Klaus Hesse is at the Institut und Sammlung für Völkerkunde, Universität Göttingen, Theater platz 15, D-37073 Göttingen, Germany. 1 The paper is based on 18 months’ fieldwork in Mandi. The research was supported by the Centre for Historical Studies, Jawaharlal Nehru University, and financed by the Volkswagen Foundation, Germany. 2 In this paper, the term ’reciprocity’ is mostly avoided, since it has been used too ambivalently. For an example of the ambivalent use of the term, see Sahlins 1972. For a critique of the term reciprocity, see MacCormack 1976; also Cheals 1988, Weiner 1992, and others. To avoid ambiguities, reciprocity is here generally equated with reciprocation, and the relation of returning a gift with retribution. 110 unilateral gift relations between wife-givers and wife-takers in north India has to be modified in respect to Mandi. Reciprocation within asymmetric relations is, as will be shown, an underlying principle of the bartan among the Khatris. Thus, this contribution is directly related to the general problem of reciprocation within the theory of the gift, and to some of the more recent debates on this question with regard to north India. I Introduction . The Khatris of Mandi are a traditional urban caste of traders, state servants and landowners. Mandi, situated in the lower hills of Himachal Pradesh, was the former capital and the religious as well as trading centre of Mandi State. Since 1948 the town has been the district headquarters. With about 22,000 inhabitants, Mandi is a small town; however, it has a distinctively urban character. The Khatris conceive themselves as an endogamous caste or biradari. Occasionally urban Khatris marry girls from the Khatris of the countryside, but they do not give daughters to them.’ Traditionally, the Khatri biradari was divided into twenty endogamous patronymic sub- groups, each usually associated with a particular gotra or gotra name. Some of these sub-groups have had a notion of common descent, others not. The Khatris have a flat social structure. Remembered genealogies of more than five generations are rare exceptions. The Khatris of the town biradari consider themselves a ’close-knit community’, and this applies to their close kinship ties within the community and town, their system of prestations or bartan among kin, close friends and neighbours, as well as to their solidarity against other groups and castes. With regard to marriage rules, in principle, the four-gotra exogamous rule is given, though reduced to a specific number of generations, in regard to FM, MF and particularly MMF gotra. Furthermore, wife-giver families should not take women from their wife-takers.4 The direct or delayed exchange of daughters (batta satta) has been a social practice among the Khatris, though to a much lesser extent than reported, for example, by Madan (1965, 1975) in the case of Kashmiri Brahmans, and it is considered 3 As stated, the Khatri could marry Khatriani from the countryside. This, however, was considered to be improper for a first (regular) marriage, and they would hardly marry a daughter to Khatris outside the town biradari. (These rules have been relaxed during the last twenty to thirty years, and intercaste marriages with other high castes also occur). Membership in the biradari refers to patrilateral, or preferably, bilateral filiation from town Khatris having property within the town. 4 Despite the strict prohibitive norms of marriage rules given, there has always been flexibility to a certain degree, conditioned by the social circumstances and the amount of social criticism within the biradari. 111 as ’not truly honourable’.1 The traditionally preferred type of marriage is kanyadan, the unreciprocated ’gift of a virgin’. Marriage in north India and in Mandi is not prescriptive, and thus an ’open game’ beyond the limits of the exogamous rule among the town-endogamous Mandi Khatri biradari. However, tendencies towards repetitive marriages are by no means absent. Matrilateral and patrilateral cross cousin marriage is absolutely barred, but third degree matrilateral cross cousin marriage constituted a borderline case, even in the past. Thus, delayed repetitive marriages between wife- givers and wife-takers are not only possible, but have been quite frequent. Traditionally there was also no bar to marrying sisters-at the same time or consecutively. Up to the 1950s, marriage negotiations started from the bridegroom’s side, and there has been a clear tendency to search, outside the prohibited degree of kinship, first among the relatives of the wife- givers of the family, and their wife-givers and wife-takers.6 Nonetheless, it should be emphasised that the marriage strategies are complex and an ’open game’. Neither a disguised structural pattern of continuity within a set of wife-givers and wife-takers and their kindred or a kind of delayed prescribed repetition of marriage, nor a ’closing of the chain’ within a generalised exchange system can be directly inferred. However, the poss- ibility is given within the kinship and social structure of the Khatris in Mandi, and in social practice it happens, more so in the present, with the narrowing down of the prohibited degrees of kinship in marriage.’7 II The basic structures of bartan Bartan is an ambiguous term. Literally it means ’vessel’, but in Mandi, Kullu and elsewhere it can also imply shares in communal lands or forests. 5 For the poor members of the biradari, ’daughter exchange’ (batta satta), as well as temple marriages, are an important means to reduce the general costs of marriage and the amounts of prestations and dowry. However, even in these marriages a minimum standard amount of gifts, dinners and garments for the bride and bridegroom have to be given. Among the élite of the biradari it is still a means of acquiring suitable marriage partners within the same status group. However, they practice mostly delayed exchange, and large amounts of gifts and dowry are given. Traditionally they sometimes arranged quite complicated chains of exchange. The particular weak point in these arrangements was the breach of promised giving or retributing. Though batta satta, in one form or the other, has been a social practice among the Khatris, it was always tainted with non-accordance with the ritual preferences and commit- ments of the biradari. 6 Today the marriage negotiations start from the bride’s side, with similar results. This change hardly had an influence on the social position or standing of women, which is stronger and better among the Khatris than among other castes in Mandi. 7 This particularly happens among the top families of the town, who want to marry within the same socio-economic strata. Orthodox families still observe the traditional norms of exogamy. 112 In Mandi and elsewhere bartan basically refers to the giving of prestations or gifts among kin at the life cycle rites or samskara and on other socio- religious occasions, actions of great social importance. Bartan among the Khatris of Mandi implies exchange of gifts, and the relations of giving, receiving and retribution implicit in the meaning of gift (Mauss 1968) are locally expressed in the duplication, artan bartan. This connotation has already been referred to by Eglar (1960) in her pioneering study of vartan bhanji in west Punjab. Her study, as well as the ethnographic material presented by Lewis (1958), Mayer (1973) and Tambiah (1973), have shown that asymmetric reciprocation is part of the system of marriage; this paper relates to their findings. Bartan in Mandi is closely connected with the samskara of the ashrama (life stages). The samskaras relate to the life cycle and its rituals, and also to the creation and perfection of the body, mind and soul, and to purification, thus implying an ontological category. Marriage and other samskaras have their particular structures which are linked to the general pattern of ’rites de passage’, as outlined by Van Gennep (1986) and, in a more sophisticated form, for instance, by Leach (1978).