Village Livelihoods and Symbolic Spaces
The Context of Eastern Indian Rainfed Farming Project

Martin Rew

Paper presented at the conference Livelihoods and Poverty Reduction: Lessons From Eastern India, 25-27 September 2001, by Martin Rew, Department of Anthropology, University of Cambridge, UK

To download, print, or bookmark, click:
To cite, quote this address and the download date. Not for commercial use.
2001 Martin Rew. Distributed by
Do not remove this notice from digital or paper copies of this text. 



Agricultural Production
Labour and Exchange Relations
Spatial Dangers and Dangerous Change: The Village, the market, and a festival
Land, Kinship and Household


This paper comes out of research carried out over the period of January, 1999 to August 2000. The research, undertaken by both the University of Cambridge and CDS, Swansea, was primarily concerned with trying to examine the nature of seasonal livelihood dynamics within the District of Keonjhar, Northern Orissa. The national partner institution which has been collaborating in the research since its inception has been the Institute of Socio-Economic Development (ISED), based in Bhubaneshwar.

Keonjhar District is one of the key areas in which the Eastern Indian Rainfed Farming Project (EIRFP), funded by DFID, has been working since 1995; the implementation of the project has been the sole responsibility of KRIBHCO since this date. The main orientation of the project has been to improve the livelihoods of ‘poor farmers’ through both ‘low-capital’ agricultural technologies, and the encouragement of  ‘self-help’ groups, primarily geared towards credit and savings.

In order to examine the nature of livelihood dynamics within the project area the research team carried out intensive household interviews in two project, and two non-project villages. The rationale behind this was to examine the labour opportunities and linkages both between these villages, as well as to assess the differential access to labour opportunities of the various villages within the immediate district area and further afield, such as the urban centers of Cuttack and Bhubaneshwar. 

With all this in mind this paper largely concentrates on the symbolic and spatial context in which livelihoods are understood at the local village level within Keonjhar. A particular emphasis is placed on the manner in which local ideologies of gender orientate the appropriate social strategies in which the organisation of rural labour takes place. This has wide ranging implications for both project policy, and wider development policies, which are concerned with designing suitable interventions for improving women’s access to, and participation in, the rapidly changing labour market. 


It is clear that the political and social sensitivities of the context in which ‘the Project’ took place greatly affected local perceptions of what the aims and goals of ‘the Project’ were. It is also clear that the concept of ‘context’ itself, and by extension ‘locality’, were also highly contested domains within the institutional realm of ‘the Project’. There was constant debate in ‘the Project’ as to whether the District of Keonjhar was ‘predominantly ‘tribal’’ or ‘traditionally Hindu’, ‘a non-cash based society’ or ‘becoming slowly monetarised’, ‘mono’ or ‘multi-cropping’, whether it was ‘fundamentalist’ or ‘anti-Christian’, and so on and so on.

The Projects’ ideology and modes of operation certainly struggled with attempts to historicise and socially contextualise the realm of its operation. Here however, I will attempt to offer my own interpretation of ‘context’ in relation to some of ‘the Projects’ priorities. Of key concern will be to examine the nature of rural labour relations particularly in terms of agricultural labour, and to comment on some of the symbolic and spatial realms in which ‘the Project’ takes place.      

The nature of rural labour relations has obvious implications for a project which is concerned with improving agricultural productivity within the district. Part of ‘the Projects’ remit was to institutionalise self-help groups in order to undertake certain agricultural activities. There were implicit assumptions within ‘the Project’ and within DFID that encouraging a ‘participatory approach’ through these groups and activities would do much to ameliorate long existing inequitable caste relations, debt-bondage to money lenders, and forms of discrimination against women.  

What will be at issue here is that the spaces and exchanges - using exchange in its widest sense - in which activities such as agricultural production take place are highly symbolically laden and do not translate easily into the metaphor of ‘participation’, which at its heart is an idealised abstraction of egalitarian exchange relations. In particular the contexts of the market and explicit political activity within Keonjhar present social relational forms which are highly ambiguous and potentially unbounded, and which often present a potential challenge to existing norms of authority, status and practice.  

It becomes quite clear in the Keonjhar context that when looking at the position of women, for example, social restrictions on their movement into certain social spaces affect the forms in which women could be said to potentially become more incorporated into, and participate in, the increasingly monetarised economy of the area.  Rather, within Keonjhar, the prevailing norm is that for women to participate in such spaces could present a potentially serious threat to embedded ideals which often articulated a communal sense of prestige with relation to the propriety of its women.   

Agricultural Production

when I was young we were tied to the higher castes and the Maharaja. We had to work for them when they said and we were often beaten when we did not work efficiently. I can remember times when we had nothing to eat but the fodder that was given to the animals. Things were terrible then, but they are better now. At least now we can choose whom we work for and get paid for it. We are less exploited (atayachaar) now…….“things are very different today; now the Sahu[1] themselves can even become the Mulya.[2]

During an animated and heated discussion with Dhurba Patro during the monsoon of 2000 Dhurba explained to me in the above terms his perception of historical change since Indian independence. His statement neatly summarises the gradual untying and breaking of former ties of absolute labour bondage which had existed within the rural areas of the district both prior to and immediately after Independence. Dhurba’s description of the not too distant past points to a period where the landless of the area, which consisted mainly of those from the lower ‘tribal’ and ‘untouchable’ castes, were still in a position of bethi (free-labour) to the landowning households of their respective village. The system forcibly tied their labour to the, largely, Brahmano landowners of the village and could be called upon at any instant the landowners required it. The present, however, in Dhurba’s view, is now very different.

Certainly the types of rural labour relations and the process of agricultural production that we witnessed during the teams fieldwork were a far cry from the bethi labour system that Dhurba describes. Throughout the locality the annual agricultural cycle forms the basis around which all income generation is geared. Whether ‘tribal’ or upper caste Hindu, most activities are ultimately orientated towards the peak cultivation period from mid-June, when the monsoon starts, through to the end of November when the harvest is completed.  Both periods are sanctified in terms of the Oriya, Hindu calendar with the celebration of Raja taking place at the onset of the monsoon, usually on the 14th of June, and the celebration of Makar Sankranti, on the 14th of January, which solemnises the reaped harvest and marks the beginning of the agricultural and Hindu New Year.

Of key importance within the annual cycle are the three discrete yearly seasons of the rainy season (borso-dino), winter (sitto-dino) and the summer season (khora-dino). The dynamics of the cycle within the locality usually produce an intense period of activity throughout the rainy, and the latter half of the winter seasons, with agricultural activities coming to a virtual standstill throughout the intensely hot summer period from March through to the beginning of June. Only two of the 206 households, across the two villages, were involved in any form of second, winter cropping (rabi), with both households having access to small-scale pump irrigation; the majority of households are, therefore, highly dependent on a yield which is only harvested once annually during mid-November.

During the peak cultivating period running through June to late November both men and women of the locality are engaged from early morning onwards either in cultivating their own fields or working on others fields for a wage. Men begin ploughing the fields as soon as the rains appear in mid-June. Ploughing is very much the domain of men, and of a certain caste, with strict sanctions and restrictions placed against both Brahmano’s and women from engaging in it. ‘Traditional’ Brahmanical norms are very much in evidence within the area with it generally seen as inappropriate for the highest and priestly caste within the Hindu caste hierarchy to be engaged in the commercially orientated and material nature of cultivation.  Here conceptions of Brahmanical male purity (pavitrata) are in contrast to the impure (apavitra) conceptions associated with women, particularly those who are menstruating, and are instrumental in legitimising the strict prohibitions placed on women from even touching the plough let alone using it.         

After the fields are ploughed work groups of both men and women are occupied in the intensive task of either transplanting or broadcasting paddy seed in the tilled fields.  The choice to either transplant or to broadcast is primarily dependent on the decisions of the landowner or owners as to which system they think is both more appropriate and effective. Generally transplanting is viewed throughout the area as the more ‘modern’ of the techniques but is not always chosen on the grounds that it is more labour intensive, and is, therefore, highly dependent on your ability to mobilise labour from within your own kin or caste group. If your labourers are not kin, or of the same caste, this will almost certainly have to be paid for, in either cash or paddy. Seed germination as a result of transplanting also takes a longer duration and is consequently highly dependent on a consistent water supply; realistically it can only be seriously entertained as a practical possibility on low lying land where water retention is at highest. As a general rule the most affluent households are those owning prime ‘lowland’ and are therefore the more likely to be adopting transplanting methods, particularly if the household also has access to small scale irrigation such as a water pump.

Paddy seed is then allowed to germinate with the pattern of the rains watched by all the members of the household. The timing and level of rain is a constant source of conversation as to whether its pattern was likely to affect the ultimate yield that year[3]. Throughout this germinating period the fields are weeded on a daily basis. There was no clear demarcation of labour here with both genders often involved in the activity. On the whole, however, it was mainly the women of the household who weeded while the men maintain a sustained surveillance over the fields and watched against possible theft.

In the last weeks of October and the first weeks of November all activity is concentrated on reaping the harvest with both men and women, including often the children of the household, involved in collecting the germinated paddy seedlings and bundling it into piles (bida) in preparation for its threshing. The seed is then thrashed out on the floor immediately outside the house, or on ‘collective’ threshing floors positioned centrally within the village, and the seeds then collected for de-husking. The tasks of both threshing and the de-husking is almost entirely the responsibility of women who crouch at the dhenki (paddy de-husker) and whisk away the de-husked rice at great speed from underneath the hammer head of the long wooden pole. The paddy is then laid out to dry and by the next day it is ready for consumption, with a certain proportion of it stored away in the household granary for next years planting. 

Labour and Exchange Relations

During these key productive months a high demand is placed on the labour available within the area. Strong norms of reciprocal exchange exist within the separate hamlets (sahi) of the village at both the kin and intra caste levels. As hamlets within the villages I conducted fieldwork usually consisted of both a single caste and identical patrilineal kin group this practically amounted to the same thing.  The men of the sahi were expected to collectively engage in binimaya (literally meaning ‘exchange’) with their fellow agnatic kin. This not only denoted the physical reciprocal exchange of labour in terms of cultivation or the repairing and building of houses but also the need for men to engage in discourse with each other, literally, to exchange views. Discourse across sexes, however, is highly regulated with discursive exchange between unmarried men and women actively discouraged; unmarried women, in particular, are encouraged to simply maintain communication with those within the household and not beyond the immediate caste or even kin group.

In contrast to the seemingly egalitarian nature of intra-caste reciprocal labour exchange the more affluent, and usually ‘higher’ castes draw heavily on the pool of labour from the ‘lower’ castes both within the village and from nearby. Varying forms of labour relations currently exist within the village and denote varying degrees of hierarchy and patronage. For example, throughout the locality completely landless households can be forced into a contractual relationship with the large landowning households in terms of either a kothia or halia labour contract. A kothia relationship usually requires the labourer from the household to move to his employer’s home whereupon his labour can be called upon at any instant. In return he receives a previously agreed, annual fixed sum, in either paddy or cash, usually paddy, which is paid to him at the time of the annual winter harvest. A halia contract operates along similar lines although the labourer is not tied residentially to the employer but comes on a daily basis to carry out cultivation activities for, again, an annual, fixed sum payment, paid either in cash or paddy. Raju Nayak, who had no land at all, explained to me how long term illness had forced him to send his only son, Dhumbi, off at the beginning of the previous year to work as a kothia;

what could I do? We were on the point of hunger……… day a Sahu came to the village looking for labour. He wanted a young man to work on his fields so I agreed to allow my son to work on his land for a fixed sum of 5,500Rs. He had to work for the whole year, every day, and would get half the money at the beginning of the contract, and the other half at the end, plus some fresh new clothes twice in the year. Dhumbi went off to live in the village and work and has only just come back from finishing the contract. I do not want to have to do that again. I need him nearby…………I just hope he will get daily work within the village for some time. Now he is doing daily wage work for the Brahmano’s within the village and getting 3 pai paddy per day……..             

Kothia and halia contracts are, however, becoming increasingly rare[4] in the locality, and are usually chosen as a last resort by potentially destitute households in order to guarantee some form of income over the year. Increasingly the choice preferred by most households is to generate additional income through daily-wage-labour, primarily paid in terms of time.

Across the two villages only 5 (13%) of the sixty households that I had prolonged interaction with could be described as producing a paddy surplus in terms of the yield cultivated from their own land, and daily-wage-labour, therefore, contributed a substantial amount to a households income. All these ‘surplus’ households were from high or ‘middle’ ranking caste groups within Bahatipira, with three of them Brahmano, one of them the Panchayat Secretary from the Mali caste, and one of them a Gopal whose son was highly entrepreneurial and had used money from his flourishing hotel business in order to recently purchase 2.5 acres of reasonably fertile low to medium lying land. The relative inequity of land distribution was clearly evident in the land holdings of two of the 

Brahmano households who each owned approximately 25 acres of land, with their yield far in excess of the average within the village. The vast majority of households, however, are forced to supplement annual yields through a number of different means, primarily through wage-labour but also through other small-scale activities such as the collection of forest materials for sale in the local market, such as sal-leaves made into leaf plates, or in the production of rice beer (handia), again to be sold in the market or to neighboring villages.     

All rural labourers are referred to generically in terms of Mulya, pointed to by Dhurba, at the beginning of this chapter, but it is, however, the daily wage-labourers who are most commonly described in these terms. It is here that the gradual commodification[5] of rural labour relations since Independence becomes explicitly apparent where Mul - derived from Hindi, and denoting ‘price’ - directly refers to the monetarised value of an individuals labour.  

The term is used not only to describe those working as rural labourers but also those who seek paid work outside the immediate locality and in the urban centers of Bhubaneshwar and Cuttack.  A significant proportion of the households I had discussions with were often involved in juggling the needs of cultivating their own land during the rainy season and were then supplementing this income through paid wage-labour on an alternate basis; this usually involved spending so many days of the month cultivating their own land and then seeking out paid agricultural wage-labour on the other days. During the lean agricultural periods from late November through to the next monsoon, those with land only in the region of a half to one acre would also seek wage-labour in the urban areas of the State, often in manual and low skilled jobs such as stone crushing plants or brick-kilns.

Both men and women were involved in this migratory process but with very definite distinctions. Payments for women’s labour both in terms of agricultural labour and in construction work outside the district was, often, significantly lower.[6]  What also became clearly evident was that the further up one went through the caste hierarchy the less likely it was for women to be seen in the pursuit of agricultural wage-labour. At an idealised and normative level women’s participation in agricultural labour is generally seen as a loss of social prestige (samman), despite the fact that it is also often perceived as an economic necessity. Ideally it is not even thought appropriate for a women to work on the land owned by any of her affinal kin, but sanctions were often relaxed in order to meet the high labour requirements for cultivating land, particularly during transplanting and harvesting.

What did compromise the prestige of the household most severely was women entering into a contractual relation of wage-labour outside the households patrilineage. Indeed, a criterion often used in categorising a household as being non-Oriya, particularly tribals such as the Munda, was their complete relaxation of any sanctions governing women’s participation in wage-labour relationships. ‘Their women regularly go for Udi[7] as one leading Brahmano landowner of Bahatipira told me rather disparagingly. In this regard women signified social continuity in any discussion concerning the status of a particular household within the village.

It became apparent over the course the research that the restrictions on women’s social mobility was based primarily on normative conceptions of space which were evaluated in relation to the varying sexualised, commodified and politicised spatial domains of the village, the market and the household. Wage-labour outside the kin based domain of the land owned by the patrilineage is, as we have already seen, one sphere which compromises the integrity of female identity and the prestige of the patrilineage. These conceptions have had significant implications for ‘the Project’ in terms of its ability to ‘operationalise’ its policies of ‘Gender Empowerment’, and to fulfill ‘the Projects’ remit of alleviating forms of social exclusion.          

Spatial Dangers and Dangerous Change: The Village, the market, and a festival

The weekly Saturday haat in Dhenkikote is integral to the area, populated each week by vegetable growers of the area, outsider traders (Bepari’s), from as far afield as the urban centres of the State, such as Cuttack and Bhubaneshwar, who come to purchase the vegetable produce, along with cloth, spice and kitchen utensil sellers.  Linked as they are through the medium of the local Saturday haat both the villages of Ramchandrapur and Bahatipira are situated and engaged in similar flows and ebbs of the local economy, and are both subject to the processes of agricultural market change and local government administrative management.

It is precisely this constant potential for change, however, which denotes the haat area as an arena of potential relational disruption and ambiguity. It is indeed a regular occurrence, in discussion with someone from any caste, gender or social group, to hear the haat described as a place which ‘belongs to no one’ (kaharo nahi). The metaphorical implications of such a statement within the locality of Keonjhar, however, is not simply a case of acknowledging the insubstantiality of property relations within the haat but is also to recognise that it has other associated meanings; signifying in addition further connotations of anonymity, in opposition to kin based relations, and danger, as opposed to the safety and protection of the household.

Gell (1982) explains the spatial and relational dynamics of a weekly Indian tribal market, in the Bastaar District of Madhya Pradesh, to show how they throw into relief the clear demarcation and differentiation of social groups within the area ‘in ways not possible in the context of the daily routine of village life’ (Gell 108: 1999). Markets as both ‘maps’ and ritual, one might wish to say.  Gell points to how hierarchical relations are sharply articulated through the complete lack of negotiation over the price of commodities between the Hindu seller and the ‘tribal’ purchaser. Through the commodities the Hindu sells and the ‘tribals’ purchase, such as ornate forms of jewelry perceived by tribals to be ‘Hindu’ in symbol and design but which are largely seen as anachronistic amongst the local Hindu populace, certain myths of Hindu and ‘tribal’ identity are exchanged and consolidated in the commodified exchange.

Gell’s depiction is, in my opinion, rather open to criticisms of arguing for a somewhat totalising structuralist perspective on Hindu ‘tribal’ relations but in terms of the impersonal nature of transactions within the markets of Keonjhar it certainly bears a strong resemblance to commodity exchange within the Dhenkikote haat. The transactions involving the purchase and sale of vegetables that I observed in the Dhenkikote weekly haat were certainly not negotiable in any real sense. The Bepari would stand authoritatively with his weighing scales on the roadside and pick off any person with a few vegetables to sell and quote them a price after weighing. A price would be given and if it was not acceptable to the seller they would move on with the Bepari remaining firm in his assessment. No prolonged discussion ever really took place.

It was also quite often the case to hear the impersonality of transactions voiced to me in a classic alienated sense in terms of the dominance of commodity exchange in haat relations. ‘The real producer is no longer there’ was a comment I often heard in my discussions with local farmers at the market. This sense of alienation became all the more apparent when it became clear that the main strategic purpose of ‘outside’ Bepari’s was to purchase small amounts of goods, and convert them into bulk wholesale which were then sold in the lucrative urban markets of Bhubaneshwar and Cuttack at a much more inflated price.         

However, rather then witnessing  a case of Hindu ‘tribal’ relations being reified in market interactions (it seemed to me that Hindu’s as well as ‘tribals’ felt potentially alienated from the products of their labour) what seemed most evident to me in the Dhenkikote haat was the lack of women involved in any form of transaction the further one went up the hierarchical scale of caste. The women who were involved in transactions were always sellers of extremely small-scale produce from either ‘tribal’ or ‘low’ caste groups. Women purchasing goods are few and were usually always protected from the potential dangers of the frantic and highly populated market by either their brothers, fathers or husbands.   

In no sense do I want to pretend that transactions are not deeply embedded in hierarchical relations, of which the Hindu populace in the market of Dhenkikote are certainly dominant, but to emphasise the fact that the impersonality of relations within the Dhenkikote haat have further associational meanings when considering local ideals delineating the appropriate spaces in which women can enter. It is important here to stress the fact that these associational meanings are never fully articulated at the discursive level but are inferred through descriptions such as the haat being depicted as ‘belonging to no one’. In this sense the anonymity of the market for local people infers a sexualised realm of unordered relations which women should be kept away from. This becomes all the more clear when we examine other local depictions of space, such as the village and the household.

The villagers of Bahatipira and Ramchandrapur also revealed a concern with how political relations represented potential disorder and how they were at odds with metaphysical principles which underpinned the fundamental and organising principles of religion and caste. The village of Bahatipira has been historically and politically key to the locality for over a century and a half. Its potential as a spatial sphere of disorder is encapsulated in local perceptions, both from within and from outside, which refer to the village as a associational sphere of raj-nithi, denoting politics and rule (raj) in terms of inherent relational problems of ‘dislike’ and ‘difficulty’ (nithi). This stems largely from the fact that the village is a Panchayat village and is consequently intimately involved with local networks of power, with money circulating in complex forms, and is also highly skilled in its ability to mobilise State resources through the local government system.  Indeed, in comparison to the rest of the village hinterland around Dhenkikote the village is noticeably well resourced, having electricity, a bank, the Panchayat office, a school and an extensive irrigation canal currently under construction through funds obtained from the Block Development Office.

Depictions of the village as a sphere of raj-nithi implied possibilities of change with strategic and established alliances always being perceived as potentially under threat. In contrast, however, there is a strong emphasis within the village placed on maintaining the continuity of ‘traditional’ distinctions, and there is a obvious tension between the two. Continuous distinctions between caste groups is accentuated by higher caste groups in terms of maintaining divisions of bhed-bhoa, ‘discrimination’ or ‘difference’ (bhed), based on value (bhoa). This metaphor of discrimination and distinction is used interchangeably to distinguish either caste position within the local hierarchy or economic status within the village. Demarcations of difference, are also in tension with a sense of increasing egalitarian relations voiced at a discursive level and which also contradicted overall perceptions of the village as a place of political intrigue and uncertainty. Mausa, a retired school clerk, who became a key informant over the course of my fieldwork, described the increasing sense of horizontal exchange between opposing and hierarchically ranked castes;

we attend each others marriage ceremonies…..the Munda’s have now stopped rearing pigs for food………Pani-chuan[8] are now allowed to consume chi (tea) or food from the local village tea shops (my emphasis)

A state of complete egalitarianism was, of course, a long way off; it was noticeable that someone who was Pani-chuan had to wash their own utensils at the tea shop to avoid bodily impurities being exchanged between ranked caste groups. However, it was obvious that Mausa felt pleased that these changes had occurred and that things were, for him, moving in the right direction. 

The multifarious interpretations of local village dynamics are also played out every year at the site of the villages renowned Shiva temple, which has acted for many years as a religious and cultural focal point for the immediate locality and well beyond. It dominates the skyline of the village standing as it does on a mound of enormous rocks reaching a height of approximately 70 meters; its metaphysical possibilities, you might say, embodied in the scale of its reach. The history and patronage of the temple is central to an understanding of the growth of the importance of the village. According to legend the Keonjhar Maharaja at the time granted what is now the bulk of the village land to key allied Brahmano households during the early part of the 19th Century. As the result of a dream, in which Lord Shiva is said to have appeared to the Maharaja, he appointed these Brahmano households as religious custodians of the area and instructed a temple to be built in honour of Shiva.

The religious importance of the site, and of the village, has been consolidated ever since. It is a focal point for the surrounding locality and beyond, particularly on January 14th every year when the coming of the agricultural New Year is celebrated in the festival of Makara Sankranti. The festival draws people from all social and caste groups within the area and consequently generates a substantial amount of revenue for the village which is then fed into its resource base.

The festival marks the beginning of the agricultural year in terms of the entry of the sun into the month of Magha and the worship of Shiva, ‘Sankar’ designating another term for the deity. The period is firmly associated with a healing of the earth and is inaugurated on the early morning of the 14th by a member of the Sabar caste who places honey (madhu) as a healing and cleansing agent on Shiva’s linga within the temple. The process ultimately acts to ameliorate the destruction perpetrated upon the soil through constant ploughing and cultivation and it is here that the process of rejuvenation is played out in terms of both identity and its temporal continuity because it is the Sabar who are perpetuated in local myths as the original inhabitants of the immediate area. This symbolic healing of the desecrated earth is also part of a process in which the continued reproductive potential of women and their identity is signified and where women are made idiomatically similar to the earth.         

It is the enormity, however, of the transactions taking place during the festival that I want to emphasise at this moment. During the festival period the vast number of these transactions present a serious metaphysical dilemma for the inhabitants, particularly the Brahmano’s. Jagubhandu Bhere, one of the Brahmano priests of the village, pointed out to me how the festival had grown substantially over the past thirty years, stating that it had originally taken place over a period of three days which had now expanded into a period of ten. Traders now came from all over the District, setting up a huge variety of food, clothing, jewelry, and trinket stalls with various food stall owners at the festival stating that income generated during the period could constitute as much as 10% of their annual income accumulated over this brief time. This dilemma was by no means exclusive to the Brahmano’s but it did present a more explicit threat to the sanctity of their religious authority within the village and their association with non-materialism.

What Brahmano caste members were keen to point out to me was that despite the enormous numbers of commercial transactions that were taking place during the festival the metaphysical principles of the ritual had remained unchanged and uncorrupted. Offerings (bhoga), they often stipulated, made to Shiva throughout the ten days by pilgrims to the temple and the festival were still conducted in the same manner, emphasising the fact that bhoga were made largely in the non-monetary form of coconuts, fruits and flowers.

Perceptions of economic change increasing at a speed at odds with the principles of meaning which do not equate sanctified objects with monetary exchange were also often voiced when discussing modes of exchange outside the sphere of structured village ritual. Mausa, again (who had become the owner and manager of one of the village tea shops after his retirement), who is from one of the ‘higher’ ranking caste groups of the Mahakuds, explained to me how he was dissatisfied at the manner in which the previously sanctified and hierarchical conception of all farm animals, known collectively as goru, were now perceived in overall terms of equivalence in relation to monetary exchange and changes in price. What he was referring to was the manner in which price rates for certain animals had over the years come to be equated with the previous price rates for other more sacred animals. In other words the cost of a chicken was now what the price of a goat was some years ago, and a goat was now the price of a bullock. At one point in the discussion he exclaimed[9];

How can this be? Those who buy these animals should have to follow a ritual penalty (prash-chit) after they buy it……….they must eat some bullock shit mixed with water as soon as they come home from the market.             

Land, Kinship and Household

Links between the two villages of Ramchandrapur and Bahatipira are strong; the residents of Ramchandrapur are frequent visitors to the village of Bahatipira during the festival of Makar Sankranti and the Munda of Ramchandrapur have already established social links to the village, primarily through marriage. Many of the cultural dynamics occurring within both villages could easily be described as similar with both of them having sizeable Munda and Nayak communities. However, perhaps because there were only these two communities in Ramchandrapur, the divisions or separations between the two communities in relation to an overall ‘village’ unit, might be described as more marked than Bahatipira. In addition, articulations of village identity are not as assertively objectified as they are in Bahatipira with its lack of highly visible and objectifiable assets, such as banks, shops or temples, often cited by villagers as the embodiment of the lack of a ‘communal identity’ inherent within the village. ‘Why do you come to us, we have nothing here’ was a question often directed at me during my initial discussions with villagers.

Superficially social proximity and similarity between the two communities seems initially all the more evident when considering the fact that both groups currently have formal Orissa State government recognition as Scheduled Tribes. However, despite their official government status as a Scheduled Tribe the Nayak claim a Hindu, ‘high’ caste Kshetriya identity and to all intents and purposes are Hindu, both culturally and religiously. The Munda are perceived by the Nayak as the more ‘recent’ inhabitants of the village, and are increasingly beginning to emulate Hindu customs and religious practice. The Munda, however, are more willing to acknowledge their adivasi status, although adherence to this sense of identity is gradually becoming more tense and ambiguous given the pressures on people within the locality to adopt ‘more Hindu’ customs and norms. Difference in terms of insider/outside distinctions is also strong demarcated by the Munda in their referral to non-Munda’s as diku[10]. At the moment, however, the Munda still do retain their distinctive language, and some of the practices which are marked in the locality as ‘tribal’ in orientation, although spoken Mundari amongst the younger generation of men is declining in favour of Oriya[11].

Established links of patronage to authority within the immediate locality are largely concentrated through the Nayak. Members of the Nayak community have monopolised lines of power and authority within the village through a style of ‘big man’ politics and a member of their community has always been nominated, and then elected, to the village Panchayat since the systems inception in Orissa in 1975. Such ‘big man’ forms of command have a strong kin based dimension with new lines of authority usually established through claiming close kin links to previous office holders[12].

In part the lack of command over village institutions by the Munda is explained by one of the most noticeable differences between the two communities. The Munda in contrast to the Nayak migrate outside the village for wage-labour, often for several months, and to places far afield from the village. At one level the lack of sustained and continuous periods of time in residence does a great deal to compromise the Munda’s practical ability to participate politically and socially within the village.  However, constant cyclical migration has important social and symbolic dimensions, particularly in reference both to gender, kinship and to denotions of belonging in reference to place, with place being largely embodied in the connection between kin and land.  

As with attitudes concerning women’s participation in paid agricultural labour within the village or nearby villages, or their movement within the realm of the market, migration to a distance ‘far’ outside the immediate locality highly compromised the status of the Munda within the village and does much, one might argue, to consolidate their internal perception of marginality. In the eyes of the Nayak the Munda are not able to sustain a very localised sense of kin in relation to the land because of the frequency of their movement outside the village. From mid-November onwards, after the peak cultivating period was over, it was not unusual for whole households to migrate to either Cuttack or Bhubaneshwar to seek wage-labour. I myself, returned from a trip to Bhubaneshwar in the December of 1999 to find that there were only a handful of Munda women left in the village. Everyone else had left in groups to find paid work. By contrast only a handful of Nayak men had sought wage-labour outside the immediate locality.    

In addition, continued absence from the village further consolidated Nayak perceptions that the Munda were not the original inhabitants of the village. The Munda are often referred to by the Nayak as not having mula, which denotes the sense of ‘original root’ one has in relation to the land, particularly in terms of kinship. These perceptions are very explicit despite the fact that communal origin was in fact extremely difficult to verify. In discussions centering around genealogical origin in terms of the village it was apparent that genealogical memory was relatively shallow, with neither of the two communities being able to define the origin of their respective households beyond three generations; no household could ever remember from what region or place their great grandfather had come from, although the usual account was that it was from ‘elsewhere’.

The interplay of origin and the maintenance of time and place acts in differentiating the two communities. This becomes all the more evident when we consider the relationship between conceptions of kin, land and the household.  In terms of the basics of production the activities of both the Nayaks and the Mundas within Ramchandrapur are primarily geared to the agricultural season and a predominantly rainfed, mono-cropping system of paddy cultivation, as described for Bahatipira at the beginning of this chapter. Participation in agricultural cultivation is, however, far from symmetrical and is symptomatic of the relative imbalance of land distribution, and of the capacity for land investment, within the village. Across the Munda households of Ramchandrapur land ownership never goes above 1.5 acres, with approximately one third of the households landless. Between these two polarities are a ‘semi-landless’ group of households who have may have no more than roughly half an acre within their homestead, which yields, only after a successful monsoon, a few vegetables and pulses (tomato, brinjal and chilies) used for purely household consumption.       

The Nayaks have a significantly higher distribution of land which reaches a ceiling of approximately 4.3 acres, with the majority of households, however, holding between 2 to 2.5 acres. The critical element, and a leveling factor for all households within the village, is the fact that few of the households hold what could be described as sustainable fertile land, with most of the land, classified in purely agronomical terms, highly acidic, low in phosphate, and undulating, with poor water retention, as a result. Only three of the households in the village, all of them Nayak, could be described as having a significant portion of low lying land (talajami) under their cultivation, the highest of which only came to approximately 1 acre.

Annual yields are, therefore, low in comparison to the ‘plains’ land in the southern half of the district and the area can certainly be described as highly risk prone, with a poor monsoon in a particular year significantly altering the delicate balancing act of subsistence for any of the households of the village. The same generic description could also be used to describe the relationship between productivity and land in Bahatipira with, of course, the one obvious exception of a very high proportion of the best low lying land being held solely by the Brahmano’s.       

Nonetheless, whether from a ‘tribal’ caste or from the Hindu/Oriya castes the problems of cultivating sufficient yield are allied to perceptions and forms of kinship cohesion needed at the time of cultivation. As pointed to earlier both villages consist of several, separate sahi’s (hamlets) which, without exception, are dominated by an exogamous patrilineage denoted as one’s kutumba, which broadly includes all agnatic kin living within the hamlet. Marriageable links always exist outside the particular sahi with marriages frequently taking place with a households bandhu; a sphere of affinal alliances, usually in neighbouring villages.[13]

Indeed, the paucity of land holdings was often described to me in terms of the demands of conforming to appropriate kin based rules. While careful about discussing publicly rights to property within his household Ravindar Nayak, just recently married, once explained to me the sparseness of his land holding in terms of the fact that his fathers land had only been divided, as he stated, ‘when all the brothers (bhai) of my family were married………….but we are so many bhai (a household of 5 brothers, in fact) that I only received half an acre’.[14] Ravindar was not voicing this in terms of a complaint but in terms of the fact that such rules had to be adhered to in order to maintain equitable harmony amongst all his brothers.

The normative ‘rules’ of kinship related to property and articulated to me in this sense, however, belie the fact that the cultivation of land, even after its distribution amongst all the brothers, is, more often than not, organised on a collective basis with brothers assisting each other in land preparation, ploughing, transplanting, harvesting and in the final distribution of a ‘collective yield’. Here binimayo, as formalised exchange, was the main organising principle. Indeed, it is usually the norm for men not to receive their share of patrilineal land until after their father has died, by which time they could be well into their married lives. Further delays might also play a factor in an individual brother receiving his own land because the elder brother would not usually sanction the division of property until all the brothers were married. For all the brothers to be married could take a considerable amount of time given the period it might take to organise suitable marriages for all the siblings within the household. This is largely due to the fact that a strict order, and sense of priority, is placed on the sequence of marriages amongst siblings and it is almost always seen as the responsibility of the male household head and the eldest brother (boro-bhai) to organise marriages for all of the daughters of the household, and only then for the male members to begin marrying according to age.      

As an overarching concept ‘home’ or ‘family’ within Keonjhar  is usually articulated in terms of the collective sense of gharo. The metaphorical scope of the term is, though, far reaching in that gharo refers to both the specificity of place and origin of an individual person but also to the network of relationships which involve kin, primarily those within the Kutumba. Here gharo and mula are very much mutually dependable in that it is not only the lack of mula that the Nayak perceive in the Munda but also their inability to sustain a gharo and, consequently, kinship links within the village. For the Nayak long term spatial severance from the land and the kutumba implies a breaking of kin. 

The harmonious implications of the gharo does not, of course, disguise the ambiguities and divisions in household belonging which often occur, and the sentimentality of the bonds it implies certainly must not be overemphasised. Tensions regularly occur, for example, over the appropriate time at which a married couple decides to create a new kitchen, although this does not imply leaving the home. At a more formalised level this decision can be described as predominantly a male domain and is done mainly in consultation with co-brothers and the male household head. However, the women of the household often described to me how decisions to separate were made in terms of the practical difficulties of running a single kitchen for all of the brothers and their wives. Indeed, when I discussed the issue together with both genders at the same time the consensus was that ‘separateness’ was usually articulated in terms of whether a new kitchen had in fact been established, although all kin members may be living under one roof. The reality of ‘separateness’ is, of course, never described in absolute terms and is often simply articulated in terms of either ‘we live together in one place’ (ethaki-rahiba) or ‘we live separately’ (alga-rahiba). Ultimately, however, practical difficulties cited by either gender in terms of household organisation are never at the expense of undermining the inclusiveness of the collective gharo.      

To reiterate, in describing the practicalities of needing to create separate kitchens or in the division of property, a member of any caste group is careful in articulating this to anyone outside the immediate household in terms of the conscious concern with maintaining a public show of household cohesion. The Indian anthropological literature has made much of the tension between conforming to an idealised sense of the family (parivar) with extensive kinship ties and the transition to a more ‘nucleated’ family structure. Cautionary notes (see Parry: 1979) have been voiced to the conceptual opposition itself, stemming as it does from a Western conceptual framework which is  often forced to describe non-nucleated structures in opposition to itself.  I would concur with these cautions by simply stating that in Keonjhar there is no conceptual term to imply a ‘joining’ or a ‘separation’ in any clearly articulated sense other than that households are constantly concerned with showing and producing appropriate wider kinship ties amongst kutumba and bandhu, and in relation to an idealised intra-caste and Oriya sense of belonging. In local terms the sanctity of the gharo and its sustained cohesion is the embodiment of these ideals.       

All this has obvious implications for expressions of gender identity within Keonjhar and here it is best to look at the totality of the gharo in relation to what are best described as ‘local ‘Brahmano ideologies, which are often used generically across other ‘higher’ castes in Keonjhar, and also at a very formalised level in terms of ritual ceremonies such as marriage; the Nayak are certainly one of these groups. Within Keonjhar ‘higher’ castes speak of grihasth which literally refers to the ‘householder’ and expresses a significant stage in the male life cycle. It is the ‘complete’ context in which married life, religion, social/kinship obligations and cultivation activities take place. Household identity is also spoken of in less metaphysical terms such as ghar-grihasthi, and is often used to refer to the more practical requirements (money, food) needed to secure one’s ‘house’ and livelihood.  These terms are then further differentiated by the notion of grihini, which refers to the householders wife. It was clear that within Keonjhar the term grihasth associated agriculture as primarily a sphere in which men worked and belonged, and a distinct stage in the male life-cycle, while the term grihini denotes the person who looks after the house, and is seen as primarily the domain in which women worked and belonged. 

For the Nayak the sanctity and purity of the gharo was highly dependent on the wife of the household maintaining spatial continuity within the household and for her not to interact with non-Nayak caste members. In any discussion I had with a Nayak household concerning the need to find supplementary wage-labour for the household everyone in the household was always keen to point out that none of the women in the household, either wives or daughters, had been outside the household for wage-labour.  The inability of the Munda wife to uphold such high caste ideals because of her constant movement outside the household consolidates the Nayak consensus that the Munda are not fully Hindu and that their sense of village belonging is tenuous.  

Within Keonjhar debates such as the appropriateness of either a ‘joint’ or a ‘nucleated’ family system did not seemed particularly inappropriate when it came to discussing the collective organisation of agricultural production, or in terms of conceptualising kinship and the household. Rather, the continuity and social prestige of the gharo for the ‘higher’ castes seems to rest on the tension between maintaining the prestige of the gharo in relation to social and spatial sanctions placed on women’s participation in exchange relations of any kind, including wage-labour relationships. Castes ‘lower’ down the hierarchy, such as the Munda, were also often the most impoverished and necessity usually dictated that most able adults within these household had to obtain additional wage-labour in order to supplement the households income.       


The District of Keonjhar has indeed undergone rapid social and economic change since the post-Independent period, as the statement from Dhurba Patro at the beginning of this chapter exemplifies. The potential reversal of roles that Dhurba’s statement points to is a perceived structural shift, albeit more imagined than real, whereby one of the historically key landowning and entrepreneurial castes of the area, the Sahu, are also perceived as themselves vulnerable to the vagaries of the market, and might indeed be forced to gain employment and income by becoming a Mulya.

Furthermore, the perception of Dhurba points to the manner in which social relations, firmly embedded up to now through the hierarchies of caste, are emerging into positions of possible dislocation. Indeed, discontinuity has been the main subject matter here. Within the context of Keonjhar, however, discontinuity manifested particularly in relation to two old anthropological staples; namely local concerns with the commodification of social relations, and the underlying principles of exchange that underpinned either the reproduction or dissolution of caste boundaries.

What I have been at pains to stress here is the manner in which communal solidarity and continuity are, in part, dependent on the marker of gender. For all castes within the locality the inherent  problem is one which sees the continuities of caste as relying upon a spatial and embodied conception of the household, and of both female labour and purity, which is always under threat from the increasingly commodified and inter-caste realm of exchange relations which are outside immediate kin relations and the household.


[1] No members of the Sahu caste actually live in either of the two villages but they are synonymous throughout the area as the key moneylenders and business caste of the locality. To a large degree it has become a generic title so that anybody who lends money or who offers employment can be referred to as a Sahu.

[2] Excerpt from an interview with Dhurba Patro, an ‘elderly’ man (age unknown) from one of the largest Scheduled Castes of the area, on July14th, 2000.

[3] I started intensive fieldwork just a couple of months (August’99) prior to the infamous ‘super-cyclone’ (October’99) which mainly hit the southern coastal areas of the State, but which also hit parts of the Northern plateau District of Keonjhar. Farmers at the time were seriously concerned about the certain  loss of crop. Rationales were expounded upon at length as to why the cyclone had struck with many farmers concurring with each other that it had been pointed to in the Gita or that somehow Bhagwan (generically God) had not been appeased. Others took a less metaphysical stance stating that the rains had been poor after the last couple of years and now it had come all at once. 

[4] Of all the 60 households that were interviewed regarding household labour strategies only two of the households, one of which was Raju and his son Dhumbi, described themselves as working on Kothia or Halia labour contracts.

[5] I have no wish here, however, to repeat debates concerned with the analytical opposition between ‘commodity exchange’ and ‘gift exchange’ in this context (see Gregory 1982 & Strathern 1985). Such an opposition has been rightly subject to neat criticisms (see Parry 1986 & Laidlaw 2000) given the problems of using the idea of the ‘gift’ as suitable in any sense for the Indian context, both pointing out clearly that there is really no such thing as a ‘free gift’. Central to the discussion here is the manner in which principles of exchange in two villages of Keonjhar are undergoing change not in opposition to a Maussian romantic sense of ‘gifts’ but in terms of an emerging cash-based economy and the wage-labour relationships engaged with it.

[6] It was the usual norm for women’s wage-rates to be lower whether it was agricultural labour that was being paid for or construction work. On average it usually came to about 5Rs. less than the payment for men. The rationale for this was often voiced by co-male workers and contractors in terms of the fact that men were generally believed to be stronger, more efficient and productive.   

[7] Udi is the term for half a days work from early morning until early afternoon.  A full days work was simply termed dino, literally meaning day.

[8] Another term for ‘untouchable’ castes who are not allowed to touch the water of ‘higher’ castes.

[9] Interview with Kulamani Pradhan on 17th July, 2000 who was known to me as Mausa, meaning mothers uncle, and which is a sign of both affection and respect.

[10] The term diku is not used locally simply as an expression of antagonistic difference but is also used to acknowledge any non-Munda in terms of a respected guest. 

[11] The whole question of  ‘tribal’ as opposed to Hindu identity will be dealt with in more detail in the following chapter.

[12] This was taking on a distinctly new gender dimension at the time of my fieldwork as the recently elected Panchayat ward member of the village was the wife of the previous incumbent, and was being elected under the government women quota system. In some senses this might seem to contradict my whole argument concerning the restrictions placed on women’s mobility. I would argue, however, that to a large extent it reinforces it because her election was described by villagers largely in terms of her visible access to male power rather than power she was believed to have a right to. 

[13]Anyone, potentially, from the same caste group, however, is Bandhu and there is a fine semantic line between its use in largely informal terms, being simply used to address a close ally in quasi-kinship ties, and its use for clearly demarcating potentially marriageable households and persons. I myself, could loosely be described as Bandhu, as I was on a number of occasions.   

[14] This was a very sensitive subject of discussion and it was usually not thought to be appropriate to discuss the division of property in front of a stranger, particularly when the father was still alive.