69 THE BOUNDS OF INTIMACY: INTERACTIONS AMONG WOMEN IN MADRAS, INDIA Kelsey Clark Underwood Department of Anthropology University of California Berkeley, California This paper is an interactional analysis of relationships among Tamil women in a domestic setting in Madras, South India. The dimensions of intimacy and power are explored through detailed consideration of interactions between non- kin relations, primarily a low caste servant and her high caste mistress. My aim is to demonstrate how particular soc- ial relationships are realized, negotiated, and perpetuated through interaction. This emphasis on interaction is based on the view that broad social, economic and historical processes not only shape people's lives but permeate their most minute actions; at the same time, these actions themselves constitute the larger scale social world. Through the discovery of patterns of interaction in particular situations we can find the links between the ethnographic detail of everyday life and the broader forces constraining and informing people's actions. In this way, following Mills (1961), we can gain a perspec- tive on personal biographies in the context of history and social structure. Moreover, examination of the strategies people use interpersonally enables us to discover and observe their socio-cultural assumptions or 'vocabularies of motive' (Mills 1963) and to observe how they operate. Through long term socialization in a particular milieu all people learn both the assumptions and the linguistic conventions for their use in a wide range of contexts (Gumperz 1982). Sys- tematic analysis of people's interactions enables us to find What these conventions are, what they mean to the participants, and how they relate to larger scale processes. This approach, -contrary to Giddens' charge that studies of everyday life "easily rationalize a withdrawal from basic issues involved in the study of macro-structural social forms and social pro- cesses" (Giddens 1973:15), in fact provides the observable data of how such processes operate and thereby transmit pat- 70 terns of social organization. Such analysis will enable us to look at how power is ac- tually exerted in particular situations by people of differ- ent social statuses. It will also allow us to look at how intimacy is managed and negotiated by such persons and how the limits to this intimacy are established in interaction. Power, hierarchy, and inequality are widely discussed in the literature relating to India. For example, the articles in The Powers of Tamil Women (Wadley 1980) argue that Tamil women have accrued 'power' in the structures available to them. Power in that collection is viewed as a symbolic sub- stance, frequently equated with the cosmological concept shakti, a positive female energy which women gain through maintaining order and controlling themselves within prescribed bounds. Such an approach, however, does not consider how women negotiate these bounds interpersonally, what happens when they are not maintained, or when any 'negative' power, such as pollution, comes into play. In the present analysis, power, the ability to exert one's will or influence in relation to someone else, is con- sidered in terms of its use and manipulation in interaction. One's position in the hierarchy gives one access to resourc- es, one of which is the interactionally learned knowledge of how to exercise power situationally (Gumperz 1982). I look at power, then, as it is used to perpetuate or call into question relationships of relative dominance and subordin- ation. Power is thus not viewed in this paper as an ab- stract phenomenon but as a quality of interaction -- the in- terpersonal means by which issues of influence and domination are carried out. Ordinarily one would expect that in relationships be- tween persons of vastly different social statuses, power man- ipulation would be the prevailing mode of interaction. In cases such as the one to be described below, however, in which all the participants are women, one might expect another interactional quality -- intimacy. Brown and Levinson have suggested that in many societies, "compared with women, men may assimilate to upper-class dignity and competition for power, while women, excluded from this arena, maintain sol- idarity ties with one another" (Brown and Levinson 1978:251). Other studies have developed similar arguments in discussions of Tamil women (David 1980, Egnor 1980, Wadley 1980) and American women and men (Maltz and Borker 1982). Especially in a society like India where gender activities and roles are segregated to a great extent, one might predict that fe- male interactions would exhibit a more 'intimate' quality than would male or cross-sex interactions. In their discussion of universals of language use and 71 politeness, which draws heavily on examples from Tamil, Brown and Levinson argue for the theoretical significance of anal- yzing interactional quality systematically. They suggest that such characterizations as, for instance, Radcliffe- Brown's 'familiarity' and 'respect' are crude labels for processes that can be observed and studied more precisely. They propose a fine-grained approach for the "specification of the mechanisms whereby the interactional flavour captured by such glosses is actually transmitted by the participants" (Brown and Levinson 1978:246). Accordingly, this paper ex- plores 'intimacy' and 'power' as interactional qualities among Tamil women. The juxtaposition of their extreme dis- tance in terms of caste, economic and occupational factors and their hypothetical closeness as women gives rise to the question of how they negotiate these potentially opposing qualities. The cast of characters includes primarily the women in the household where my husband and I lived as renters and adoptive family members. We lived in the upstairs but inter- connecting and interdependent half of the house, while land- lord and landlady lived downstairs. My landlady and Tamil 'mother,' whom I will call Kamakshi (another form of the goddess Parvati, wife of Siva, and titulary deity of a fam- ous temple in South India), is a childless woman in her six- ties, an orthodox Tamil Brahmin, English educated, and upper .middle class by Indian standards. She had been married fifty years to a high court judge, Ganeshan, who had retired from that career and worked part-time as a bank consultant. As a Brahmin (Iyer) Kamakshi is of the highest caste, measured in terms of relative 'purity' and 'pollution.' For the previous three to four years Kamakshi had hired a very poor local woman as her servant. I will call her Kannamma, which I intend to be glossed as meaning 'woman whose eyes are in pain.' Kannamma also worked as my servant. She belongs to a sudra caste which she called by a different name than the one Kamakshi used. While Kamakshi referred to Kannamma's caste as Padiacchi, an old and somewhat derogatory term for a caste of cultivators in Tamil Nadu, Kannamma re- ferred to herself as a Naicker, denoting a caste of rich cul- tivators. The use of a more prestigious term by any caste for purposes of upward mobility, a well known phenomenon in South Asia (Srinivas 1966). is notable in this context because Kannamma's new caste name is the same as that of the man who led the anti-Brahmin movement in Madras in the 1920's (Dr. XPeriaswamy Naicker), and who is considered the father of the Dravidian 'reclamation' of self-worth and rights. Caste re- lations along these lines continue to be extrememly politicized in the state of Tamil Nadu, for even though the 3% of Brahmins have been removed from political power for decades, they re- main the intellectual, religious, and social if not economic 72 elite. Thus in the different ways Kannamma and Kamakshi man- ipulated these highly charged labels of caste terminology, we see the negotiability of definitions of self and others as part of the larger picture of changing power relationships in twentieth century Madras. Kannamma was 36 at the time of my research, married to a factory worker whom she said earned 250 rupees per month (about $25), and she was the mother of six children. She lives in a mud hut about nine by twelve feet large, five min- utes' walk from Kamakshi's house. She speaks a different but mutually intelligible variety of Tamil from that of Kamakshi, and is illiterate. Given the respective castes of the two women, Kannamma's work as a servant fell within certain conventional boundaries. Her particular chores for Kamakshi followed traditional caste prescriptions. As a non-Brahmin, Kannamma could not touch or prepare food that would directly be eaten by Brahmins, but she could do work on food before it was cooked by a Brahmin, such as clean the rice. Her basic daily chores were to sweep and mop the floor, wash the clothes, and wash the vessels. Since only Untouchables are conventionally expected to clean toi- lets, Kannamma was not asked to clean those in Kamakshi's house (ironically, since no Untouchables were readily avail- able, Kamakshi did this herself). Kannamma's salary in 1981- 82 was the same as it had been when she was first hired, thirty rupees per month, about $3. The wage for this level of servant had gone up in that neighborhood of Madras to 40- 50 rupees per month (approximately $4.50 to $5.00), but Kamakshi had not raised Kannamma's. In contrast to the lack of salary increase, the tasks Kannamma carried out had greatly increased over the years to include a fair amount of Kamakshi's shopping and numerous errands. Kamakshi had had a Brahmin cook, whom I'll call Uma, an old widow who for fifty years had lived with the family both as beloved servant and Kamakshi's main companion until her death shortly after I arrived. About six months later another Brahmin widow, Gauri, a distant poor relation of Kamakshi's, came to live with the couple and help in the cooking and other domestic work. The distinctions and overlap between the Brahmin servants' and Kannamma's tasks caused friction. Even this minimum of information leads us to ask why Kannamma continued to accept such a low wage and especially why she allowed more and more responsibilities to be heaped on her. Her husband's and her salaries together equaled under $30 per month, and they have six children, while child- less Kamakshi and Ganesan have more than ten times that in- come (and still live very frugally). Given the financial terms of Kamakshi's employment of Kannamma and their respec- 73 tive positions, I want to look first at the behaviors which demonstrate the boundaries keeping them in these positions. Maintenance of these boundaries exemplifies their social distance. The first place to observe social distance is in the women's physical movements in relation to each other. As an orthodox Brahmin it was important to Kamakshi that she not be touched by Kannamma. When Kamakshi would give Kannamma money for shopping, she would not hand it to her but would put it on the table or even drop it on the floor for Kannamma to pick up herself. The "don't touch" rule was infringed upon part- ly due to my presence. As Kannamma and I got to know each other she would touch me in gesturing, for emphasis, or from affection. About this time she began to touch Kamakshi oc- casionally in the course of animated conversations. Further- more Kannamma was expected not to stand too close to Kamakshi when they were talking, and she begain to transgress this as well. Kamakshi reprimanded her for these transgressions by saying "You keep your distance. Kelsey may let you touch her, but that doesn't allow you to touch me!" The proxemics of pollution were of course noteworthy at the time of Kannamma's menstrual period. She would tell Kamakshi when it arrived -- Kamakshi's husband Ganeshan joked that the arrival of Kannamma's period was announced over All- India Radio. Then for three days Kannamma was not allowed to enter the Brahmins' household or cross their immediate paths, but she was still required to wash the vessels outside the back door. As I learned about this I said to Kamakshi, "then when I have my period should I stay out of your,part of the house also?" Her response was "No, it doesn't matter, what I don't know won't hurt me." This rule was negotiable in relation to me but not in relation to Kannamma. One incident further demonstrates this. Since I did not maintain the men- strual taboo with Kannamma, she was one day (during her period) at work in my part of the house, when Kamakshi, know- ing Kannamma was just on the other side of our interconnecting door, opened the door. She then lashed out at Kannamma, saying she had crossed Kamakshi's path, thereby causing her to have to bathe herself all over again. In these examples it was Kamakshi who reprimanded Kannamma and reasserted the distance between them. In the first case, however, Kannamma had crossed the boundary towards unacceptable intimacy by touching or standing too close, but in this latter case she was blamed for the transgression even though she had not actually violated the rule. Given this degree of social distance between the two women and the power Kamakshi was able to exert over Kannamma, other examples of their interaction show occasions when they 74 stretched or crossed the boundaries toward greater intimacy. More than once Kamakshi gave Kannamma an interest free loan of her salary for a ten month period, thereby enabling Kan- namma to make major purchases that would otherwise have been impossible. Kannamma's other loans from the money-lender were taken at exorbitant interest rates. No one else Kan- namma had worked for had ever given her loans, which indi- cates the unusual degree of trust Kamakshi had that Kannamma would continue to work for her. This shows that a kind of generalized reciprocity existed between them as a measure of their intimacy even though they are not close in social rank. While it may be argued that this resembles the tra- ditional inter-caste jajmani relationships of mutual depen- dence, the important point here is that in an urban, cash- based relationship these individuals went beyond the 'bal- anced reprocity' characteristic of the jajmani system. Un- like traditional caste relations, Kannamma was a wage earner and could quit her job; that she would continue as Kamakshi's servant did not follow automatically from their respective caste positions. It is especially significant that in the case of the loan Kamakshi crossed the social boundaries in an act of relative intimacy. Intimacy between Kamakshi and Kannamma was also evident in verbal exchanges. They would engage in playful banter, long confidential conversations, gossip, laughter, and tears. For instance, almost every afternoon after Kannamma's work was finished, her two little girls would come to the house delivering milk and they would all play with a stray puppy they had adopted, laughing and talking animatedly. On these occasions Kannamma would squat on the ground while Kamakshi would sit on the front steps, thus slightly elevated above the others. Though these relative positions predominated be- tween them, there were also many times when Kamakshi and Kannamma both sat on the floor (at the same level), espec- ially during long talks. The confidentiality in these conver- sations, however, was much more on Kannamma's part than on Kamakshi's, thus giving Kamakshi access to considerable in- formation about Kannamma's problems. This further reinforced Kamakshi's control over Kannamma since she could use the in- formation later to manipulate her in certain ways. The pains Kannamma told about her life were taken seriously by Kamakshi, though, and there was enough intimacy that Kannamma would cry bitterly in her presence. Though Kamakshi would not physi- cally touch Kannamma to comfort her, she exhibited great ten- derness through her eyes, face, and intonation. The question raised by these examples is: how far can intimacy be taken before it breaks down and distance is again asserted? Or conversely, when can the boundaries that main- tain distance be stretched, laid aside, or stepped over? There are clearly situational factors known to the actors in- 75 volved which define the appropriate contexts for the use of relatively more distant or more intimate styles of inter- action. Particulars of time, place, persons present and topic define these contexts and provide for the shifts in tone, intensity, and communicative style. The time when social distance was most operative was when domestic chores had to get done, while more intimate exchanges tended to occur after the work was finished. The switching between intimate and distant styles could be abrupt and take place quickly; if Kamakshi noticed an unfinished task in the middle of chatting as Kannamma was leaving to go home, she would switch immediately from the intimate to the distant style. The places where rules of distance were most applicable were the kitchen and the puja (worship) room, the areas most susceptible to pollution. When Kamakshi's Brahmin cook was alive, she never allowed Kannamma to enter the kitchen or puja room. After Uma's death, however, Kamakshi came to rely on Kannamma for more help and thus she began to carry out tasks in these previously forbidden locations. Uma, though also a servant, had distanced herself from Kannamma and de- monstrated her social proximity to Kamakshi by enforcing Brahmin orthodoxy even more forcefully than Kamkshi did herself. After the second Brahmin widow, Gauri, came to the household, the expanded tasks and areas of Kannamma's work overlapped with Gauri's, and there was animosity between them from the start. The locations most amenable to intimacy were those most peripheral to the sacred, pure, or private areas of the house (kitchen, puja room, and bedroom). The more intimate ex- changes thus took place on the front verandah, in the door- ways, or outdoors. All these interactions occured within the walls of Kamakshi's compound. It would be inconceivable for Kamakshi to go to Kannamma's house. Certain encounters were constrained both by the-location and the persons present. Occasionally whe Kamakshi would pass by Kannamma's hut in the car with Ganeshan, they would stop to give some instructions to Kannamma or receive a par- cel. These encounters, in full public view, followed the most rigid rules of distant behavior. In a public situation like that, all participants presented themselves in the most conventionalized modes of displaying social distance and hier- archy, for not only the location but the range of social ranks constrained them severely. At home, the presence or absence of Kamakshi's husband was also a major factor influencing the interaction between the women. The more distant styles pre- vailed when he was at home; and just as he spoke in authori- tative tones to Kamakshi, so she spoke to Kannamma. The pre- 76 sence of other men, visitors or strangers similarly kept in- teractions more formal. When one of the Brahmin female ser- vants was present other constraints and conflicts came into play, since there was ambiguity and overlap in the different servants' work status though not their caste status. Thus Kannamma developed less intimacy with either Uma or Gauri than with Kamakshi. Each of these actors knows the socio-cultural rules de- termining the specific contexts where greater intimacy or distance can be communicated. Each person as well can use a range of interactional styles, depending on how the situation is interpreted. Yet it is Kamakshi's economic hold over Kannamma that in a sense allows her to develop the affective bond they share. Kamakshi has more control over determining where the bounds of intimacy will be set and when they can be stepped over. In this sense, then, the affective bond between the women serves to feed the economic bond. One person's power over another derives from her structural pos- ition but is lived out through active and ongoing negotiation between them. NOTES Acknowledgements. This research was carried out in 1981-82 with the generous support of a grant from the National Science Foundation and the American Institute of Indian Studies. For helpful comments and discussion on aspects of this paper, I would like to thank John Gumperz, Daniel Maltz, Charles Underwood and Doug Campbell. This paper was presented at the Kroeber Anthropological Society Meetings of May 14, 1983. REFERENCES CITED Brown, Penelope and Stephen Levinson 1978 Universals in language usage: Politeness phenomena. In Questions and Politeness: Strategies in social interaction. Esther N. Goody, ed. Cambridge: Cambridge University Press. pp. 56-289. David, Kenneth 1980 Hidden Powers: Cultural and Socio-Economic Accounts of Jaffna Women. In The Powers of Tamil Women. Susan Wadley, ed. Syracuse University: Maxwell School of Citizenship and Public Affairs. pp. 93-136. Egnor Margaret 1980 On the Meaning of Sakti to Women in Tamil Nadu. In 77 The Powers of Tamil Women. Susan Wadley, ed. Syracuse University: Maxwell School of Citizenship and Public Affairs. pp. 1-34. Giddens, Anthony 1973 The class structure of the advanced societies. London: Hutchinson. Gumperz, John J. 1982 Discourse Strategies. Cambridge: Cambridge Univer- sity Press. Maltz, Daniel N. and Ruth A. Borker 1982 A Cultural Approach to Male-Female Miscommunication. In Language and Social Identity. John J. Gumperz, ed. Cambridge: Cambridge University Press. pp. 195- 216. Mills, C. Wright 1961 The Sociological Imagination. New York: Grove Press. Mills, C. Wright 1963 Situated Actions and Vocabularies of Motive. In Power, Politics and People. Irving Horowitz, ed. New York: Ballantine Books. Srinivas, M.N. 1966 Social Change in Modern India. Berkeley: UC Press. Wadley, Susan 1980 The Powers of Tamil Women. In The Powers of Tamil Women. Susan Wadley, ed. Syracuse University: Maxwell School of Citizenship and Public Affairs. pp. 153-167.