PUBLIC FACES

Chapter Two
The Nature of Civic Individuality

The word moka (from the Sanskrit mukha) or face was heard frequently in conversation. It stood for a person's image before others and for his self-respect. It was one of his most important possessions.
M. N. Srinivas, The Remembered Village

Before we can understand the community of George Town and examine why it is organized in the manner that it is, it is necessary to understand the nature of civic individuality and the highly personalized nature of social relationships in south Indian society that make identity so important. Reputation, public trust, eminence, and the roles charismatic community leaders play are all associated aspects of civic individuality. What I am talking about is how relationships are formed and things accomplished, and the role that individuality plays in daily life.

Even today, when the growth of cities and the spread of bureaucracy might lead one to expect that relationships would be increasingly depersonalized and that anonymity would be on the rise, south India remains a highly personalized society. By highly personalized, I mean that, beyond the limits of kin and friends, personal relationships are important determinants of how one conducts one's life. A Tamil's success in life vitally depends on maintaining good relationships and a good reputation within one's community.

To be successful in daily endeavors in Tamil Nadu society—say in running a small business or in seeking employment—a person must be trusted. This means either being well known to others, so that they know personally what to expect of the person, or being known to the leading men and women of the community, persons whose reputations are known, so that they can vouch for the individual to those who seek to estimate him or her. For example, lenders will give credit only to those

they know repay their loans, and employers hire only those for whom they or others they trust can personally attest. This need to be known is different from American society where individuals rely much less on personal knowledge and much more on bureaucratic indicators of reliability, such as credit ratings, certification, or, as the request goes, a "picture i.d. and a major credit card."

In India, as in all societies, peaceful interaction requires that people behave for the most part in predictable and reliable ways. Law, enforced by sanctions external to relationships, is one source of reliability; trust, embodied within relationships themselves, is another. In south India, trust still has a major role. An Indian evaluates the degree of trust he or she has in the reliability of a relationship in terms of knowledge of the other party or parties in that relationship. A reputation is the public sense of one's responsibility for one's identity and actions, an assessment of responsibility for past behavior.

One important factor in such an evaluation is the reputation of the other party as a civic individual. Is the other known to be a "good person" and a trustworthy individual, for example? A second factor is the strength and nature of the relationship the parties have, including how enduring and important the relationship is to the parties involved. The underlying assumption is that if a relationship is important to a person, that individual will act in a reliable manner because it will be in his or her self-interest to preserve the relationship. Again, we see here that the Indian view preserves the sense that individuals are responsible for their own actions. Finally, a third factor is whether or not the other party is known to persons of eminence and is vouched for by them. A man or woman of eminence is a person who has an enduring and highly prized relationship with his or her community,^[1] one that is defined by the prestige and status the person has achieved in that community. The actions of such men and women are circumscribed by the collective interests that define their prestige and constrain their statuses. Consequently, it is very much in the self-interest of eminent persons to preserve their eminence by vouching only for those they trust. If a person for whom they have vouched betrays their trust, it is in the eminent person's interest never again to act as a guarantor on the betrayer's behalf. This fact is well known in India. It should be clear, therefore, that reputation, eminence, status, including offices, achievement, and responsibility for how one behaves and who one is are highly valued features of identity, defining a person's civic individuality in Indian society.

It should also be clear that eminent individuals play key roles in the regulation of trust within communities. For an ordinary person, known to a limited circle of individuals, it would prove difficult to establish ties of trust that reach beyond the circle without the validation of eminent persons. An eminent person is known much more widely. Indeed, the greater one's eminence, the greater the circle of individuals among whom one is known and can act as a guarantor of reliability. A Tamil's circle forms the context of his or her civic individuality and delimits its spatial dimension.

I remember once sitting in a jeweler's showroom in George Town, the business center of Madras City, chatting with a prominent diamond merchant, when a Viswakarma, or goldsmith, came in. This particular merchant had recently shown me a diamond and gold necklace worth, he told me, $90,000. It certainly looked like it was worth that much. It was a heavy piece with several courses of diamonds. One of his Indian customers living in New York had ordered it made by him, and he was preparing the customs papers for mailing the necklace, which was why, he said, that he had it in the shop rather than under lock and key elsewhere. The merchant had a few goldsmiths who worked at the back of his shop, but according to business custom, most of his work he put out as piecework to artisans who crafted the jewelry in their homes, some as far afield as Bombay. Goldsmiths are poor artisans on the whole, and I remember being impressed that merchants, such as this one, trusted their artisans with gold and gems worth sizeable sums, although not so much as this necklace.

The goldsmith who had entered the shop explained that he was a skilled gem-setter and was looking for work. He was very polite in his demeanor and speech, and in reply, so was the merchant. The merchant explained to the goldsmith that he had no work for him but would keep him in mind. After the goldsmith had left, the merchant said to me that he would never hire the man. I asked if he knew him personally. The merchant replied that, while he did not, he knew his name and reputation. Some time ago the goldsmith had stolen gold he had been given to make a piece of jewelry and had fled Madras. When he had spent what he had stolen and was again looking for employment, he found that no one was willing to hire him in his new locality because he was unknown to jewelers there. As his situation became more desperate, he returned to Madras, but now found that no jewelers would hire him there, either. The merchant explained that he, like other jewelers, never

hired a man whom he did not personally know or for whom another trusted jeweler would not or could not vouch. Effectively the goldsmith had been blacklisted, and it was unlikely that he would ever again work as a goldsmith.

I asked if the police had been called regarding the theft. The merchant laughed. There was no advantage in that. Calling the police meant that you would lose twice, once from the theft and a second time from the police, who would require bribes. Basically, the jeweler felt that the merchants' network offered a more effective means of dealing with theft than the police.

The jeweler's sentiment is widely shared. The effectiveness of relying on personalized relationships to regulate behavior in India's cities is attested by the low rates of urban crime, despite high rates of poverty, compared to U.S. cities. For example, with a population of about 12 million people, one third^[2] of whom live in slums, Calcutta recorded less than 100 homicides in 1990, compared to New York City's 2,200 (Karl E. Meyer, New York Times , Jan. 6, 1991). One Calcutta newspaper editor bragged, "You are safe in the parks. We've had one notorious rape there in 20 years" (Karl E. Meyer, New York Times , Jan. 6, 1991). Similarly, in Madras, with a population of 4.2 million in 1981, 15,693 crimes were recorded for the year (India Today , 66). This represents a crime rate of one crime per 267.6 persons, although the number of crimes is undoubtedly underreported. This rate compares with one crime per 5.9 persons^[3] in 1986 for the university district surrounding my own University of California campus, where underreporting is also common. In the university community the crime rate is forty-five times higher than that of Madras.

Low crime rates relate to neighborhood solidarity and stability. Urban crime rates in India are highest in the new residential suburbs of India's biggest cities. In 1961, before substantial suburbs had been developed in Madras City, the crime rate was one crime per 463 persons, almost half of what it is today.^[4] In the old residential areas of an Indian city, the individual has very little anonymity. If a man is not known, people ask who he is, or they note things about him so that, if it interests them, they can follow up on finding out about him. On several occasions I have seen an Indian note the license plate number of someone who has piqued their interest, their intent being to find out who the person is. A week later, through private enquiries, they know. I myself have found that all sorts of Madrasis have known me by my license plate number. It is clear that individuality—expressed here as being known—is con-

sidered important in urban India today, and that a reputation for good character and ability is highly prized.

The "Big-Man" Versus the Bureaucrat

It would be misleading, of course, to suggest that bureaucracy plays no role in Indian society. In fact, its role is a major one. The police, law and the courts, voter registration and elections, driver's licenses, taxes and fines, examinations and university degrees, bank accounts and applications for loans are familiar components of bureaucracy that regulate the lives of Indians. In contrast to trust based on personalized relationships, bureaucracy depends on depersonalized criteria for determining reliability. For example, a bureaucratic method of determining financial honesty is by rules of accounting that judge honesty by criteria defined in law. A man's civic individuality—who vouches for him, what his relationships are, and what his reputation is—is much less relevant, although it may weigh in decisions about hiring or about how to prosecute law breaking. However, in a trust-based relationship, honesty and reliability are evaluated in terms of who a person is. With no direct monitoring, a trusted person has considerable leeway in the use of resources that he or she might control in common trust, that is, so long as that trust continues.^[5] The point to be made is that bureaucracy diminishes the importance of individuality and of personalized relationships, whereas in a society that bases reliability on relationships of trust, the importance of individuality, namely, who a person is and what a person does, is stressed.

In a trust-based society, when persons who control desired resources, commodities, and services need to choose who is to get what, then the identity of the individual seeking a benefit and the influence of the people he or she knows can be the critical factors determining how choices are made. For example, a seller often must choose to whom among several potential buyers he will sell, and a banker has to decide to whom to lend money. Under such conditions, knowing eminent persons, people south Indians often refer to as periyar , big-men, can obviously be very beneficial. Here I use the term to include influential women, although the reader should keep in mind that their comparative number is quite small. Big-men can wield great influence among members of their communities, because people perceive them as being able to act as brokers on their behalf. The civic attention of the community focuses on the individuality of big-men when community members accord prestige to these community leaders, a public recognition that distinguishes their

singularity and agency. These are the individuals who make things "happen" for their followers. By contrast, bureaucracy is designed precisely to circumscribe the influence and agency of individuals by depersonalizing the operation of institutions.

A south Indian friend of mine, whom I shall here call Tambi, related the following story about trying to start a business in Madras that reveals the juxtaposed roles and different styles of the bureaucrat and the big-man in south Indian business and banking. The story has deeper implications than simply how to get a loan in India. It suggests that there is a close relationship between the role of the individual as an agent and the organization of south Indian society. This is a relationship I shall explore in detail in the chapters that follow.

Tambi and three of his friends met as engineering students at the Indian Institute of Technology (IIT) in Madras. When they graduated they decided to start a small industry in Madras City, manufacturing tubular aluminum furniture, an industry that depends heavily on marketing. To finance the initial capital required, Tambi went to his elder brother, Annan, who is himself a wealthy and well connected businessman in Madras. The brothers are members of the Nagarattar caste, a well-known business and banking caste in south India, which also has extensive business ties with Southeast Asia (cf. Rudner 1989). In addition to the businesses that Annan runs in Madras, he is the director of a large textile mill in Kerala and has ownership interests in rubber estates and financial institutions in Malaysia. Annan maintains close contacts with the business moguls of Tamil Nadu, many of whom are also Nagarattars. This he does in part by his active membership in the Cosmopolitan Club of Madras, a men's club that is an important meeting place for the state's business and government elite. Annan also maintains his ties by going on pilgrimage to Kerala to worship the god Aiyappan with a group of his influential Cosmopolitan Club friends every December/January. Pilgrimage groups of this sort are a common feature of the Aiyappan festival, and the shared experience intensifies the sense of fellowship among group members. Members maintain their pilgrimage group year after year.

At the time that Tambi approached him, Annan had an interest in a private lending society, Frontier Investments,^[6] a business that was owned by the father of two of his classmates. Annan arranged for Tambi and his partners to borrow Rs. 10,000 on a hundi contract, a short-term high-interest promissory note, that stipulated that at the end of three months they would return Rs. 12,000. This was an agreement that required a high rate of profits for the fledgling industrialists. Each partner also con-

tributed Rs. 1,000. The partners' plan was to use the Rs. 14,000 to finance their fixed capital, but they still needed a place to manufacture. For the time being they rented a small house on the outskirts of the city, but the structure was inadequate for their needs, and, moreover, they lacked adequate working capital.

Tambi then discovered that a new shed was being built at the Guindy Industrial Estates in Madras, which he thought would be ideal for their needs. The problem was how to win the lease. The estates are an established development run by the state government, and Tambi and his partners knew that competition for the lease was bound to be fierce. Again they approached Annan. One of Annan's friends was a deputy director in the Department of Industries and Commerce, which administered the leases in the Guindy Estates. Annan told the four young men to take their application to him. Subsequently, they succeeded in obtaining the lease, although there were several other qualified applicants for the shed.

With their fixed capital and lease secured, they now approached the branch of the State Bank of India attached to the Industrial Estate and explained to the loan officer there that they had their own fixed capital and were seeking a working capital loan. The State Bank, a state-run bureaucratic institution, administered a special scheme for industrial development. They required that applicants be university degree holders whose application demonstrate that they had a viable project. As graduates in engineering, and with Tambi's elder brother's charter accountant's help, the partners had no trouble writing a convincing application. They got the loan. Their successes up to this point were a result of a combination of personalized contacts and appearing to meet bureaucratic regulations. However, without Tambi's elder brother's personal connections, the four partners would have lacked sufficient fixed capital and the manufacturing shed, and so would not have qualified as a reliable risk for the working capital loan under the bureaucratic criteria used by the State Bank. The State Bank loan officer was also unaware of the high-interest hundi loan, which in fact meant that the partners' enterprise involved very high risks.

Surprisingly, in the year that followed the partners managed to keep their business afloat, and, seeing their success, Annan personally took over the hundi loan at year's end, allowing the partners to reduce their interest payments. Financially, things were beginning to look better, but profits were still slim. By the end of that first year, the four realized that it was going to take time to develop their market before they could expect to make any real money. All but one decided, therefore, to go back to grad-

uate school in engineering. The exception was the partner who had been managing the business. He decided to enter management school instead.

The business continued for three more years, each partner attending school and contributing to the business in his spare time. They managed the business out of their dorm rooms on the lit campus. When at last they all graduated, the manager, who had been talking to some of his management friends, decided to leave the business and to go to Bangalore to begin a new business, which he thought would earn him "big" money. The departure, however, was a tense one. The manager told his partners that they owed him back pay for his four years as manager. Shocked, Tambi told him that if he insisted on back pay, they would sue him for having ruined the business, since it was running at a loss at the time. To give force to his threat, Tambi told the manager, "Go see my lawyer." The lawyer was in actuality a good friend of Tambi's, and also happened to be the son-in-law of one of Annan's partners. The reader should notice how Tambi's access to Annan's veritable "cat's cradle" of social connections begins to emerge.

The manager decided to leave peacefully. However, unbeknownst to his partners, he went to the State Bank before he left and told the loan of-ricer there that he was leaving the business and that, since he was the only one of the four who knew how to run it, the bank would be well advised to call the partners' loan. When the partners next tried to make a withdrawal from their capital account, they found that it had been frozen and the loan recalled. The partners' financial situation was desperate.

After several months of looking for a bank that would finance a capital loan, the three partners again approached Annan. In response, Annan arranged for Tambi to meet with the chairman of the Bank of Madurai, a large private bank founded and controlled by Nagarattars. Annan knew the chairman, and like many other Nagarattars, he conducted his business banking through this same bank. Without Annan's connections, Tambi says he would never have been able to arrange to see the chairman.

Tambi arrived at the meeting prepared with an accounts summary of his business. Annan explained Tambi's situation to the chairman, and then Tambi and the chairman talked informally "across the table." Tambi described his need for a loan to pay off and replace the one that the State Bank had recalled. To a banker, of course, the State Bank's refusal to continue the loan was very damaging information because it indicated that after four years of business the bank judged the partners a bad risk. This simple fact was the reason the partners had been unable

to get another bank to consider a loan. The chairman, knowing this, also knew Tambi's brother Annan and that Annan was a reliable businessman of recognized stature in his own right who kept his business accounts at the bank and also happened to belong to his own caste. There is a common sentiment among prominent men of a caste that a leader owes it to his community on occasion to extend a helping hand to his caste fellows. Bonds are created that benefit both, the benefactor gaining a loyal supporter and the beneficiary receiving the benefaction. After a short chat, the chairman told Tambi to talk to his branch manager and that "everything would be done."

Reflecting on why he got the loan, Tambi told me that he thought being a Nagarattar was important. "Most people working in the bank are Nagarattars," he said, "and if a Nagarattar with a degree knows someone working in the bank, he can get a job there. A couple of my cousins work in the bank." Also critical was the fact that his brother had a strong, established relationship with the chairman. Had he been judged purely by bureaucratic criteria, Tambi is certain he would never have gotten the loan. Subsequently, the chairman helped the three partners with additional loans, including capital loans for machinery. As a result, the business has grown and is now quite successful.

The account of Tambi and his partners reveals not only contrasts between how bureaucratic and trust-based relationships work, but also how trust-based relationships link up with and center on men of influence, "big-men." Consider the organization of the relationships used to acquire the Bank of Madurai loan and the agency of the individuals involved. The chairman enacts the role of a preeminent big-man, that is, a man who can make things happen because he controls an institution that distributes resources and services to a clientele. He is also a big-man because he wields substantial influence among prominent men, such as Annan, who are less powerful than he. In contrast to men of Annan's stature, relationships among preeminent big-men tend to be highly competitive; they do not help each other. Instead, they try to outdo one another, each seeking to achieve the reputation of being first among the preeminent. Annan, by contrast, plays the role of one of the chairman's lieutenants; he is a subordinate big-man because, while he has influence and can affect the outcome of events, his influence is less than the chairman's. He needs people like the chairman from time to time. Indeed, because he is a beneficiary of the chairman's influence (he benefits as a banking patron and by being able to help his brother), he may be counted a member of the chairman's clientele. At the same time, as a man "with

connections," Annan also has his own constituency that incorporates Tambi and his two partners among others. In turn, Tambi is a member of the constituencies of both Annan and, through him, the chairman. Finally, Tambi's connections make him the benefactor of his partners, who in turn are members of his constituency. The relationships joining the men are in each instance unequal and are based on personal rather than bureaucratic ties.

The hierarchical interlinked nature of all of these relationships and how to work them is well understood by all the parties involved; consequently, they go largely unstated. People recognize that the chairman is a big-man because of his preeminent power and control of desired resources, and out of his earshot people will call him a periyar . By contrast, "lieutenant," "constituent," and "clientele" are my terms, which I am using to ease the reader's understanding. My informants, while they have an intimate sense of the relationships these terms label, including the subtleties of relationships involved and the degrees of power represented, do not use these terms to refer to one another. Instead, they talk about a person's connections with big-men—how good they are—and who a person knows. They also appreciate a man who helps them and imbue him with eminence because he does. Eminence is the basis of a man's prestige and all Tamils, particularly men, strive for it. And eminence is graded: the eminence of powerful persons who have large constituencies is greater than that of persons who are less influential. A direct measure of eminence is the institutions that a person heads or is an officer in and the size of the population that these institutions serve.

The Individual as Focus of Organization

In Indian society, pivotal persons such as periyar exhibit a special form of public individuality, which I call an "individuality of eminence,"^[7] because public recognition of a leader's uniqueness stems in part from the fact that the leader ranks first among his or her followers. This kind of individuality is expressed in part by the special influence and autonomy that leaders have within groups and by their abilities to distribute benefits and command others to do their bidding. The individuality of leaders, therefore, is not that of one unique person among many similar unique persons, as some characterize Western egalitarian individualism. Rather it is a type that is marked by the superiority of leaders over their followers. In the West, leaders often seek to identify themselves as one

among their constituents, as expressed in the U.S., for example, by the presidential phrase, "My fellow Americans." By contrast, in Tamil society, leaders are recognized as self-interested patrons who rank above their constituents. Leaders use their constituents who in turn use them.

The individuality of eminence is associated in Tamil Nadu with a special type of group formation, the "leader-centered group." This kind of group is an association that forms around a central figure, a man like the chairman, and his subordinate lieutenants, men like Annan, and is maintained by ties of relationship that link the central leader to all members of his group. The preeminent big-man of such a group is always the controlling officer of corporate institutions that he uses to distribute benefits and attract a clientele. A few of his subordinate lieutenants will be junior officers in these institutions and will serve only him. Others, such as Annan, lack offices in the preeminent big-man's institutions, but are persons who bring clientele into his constituency because of their own personal relationship with him. Lieutenants like Annan may have ties with several preeminent big-men, just as a big-man's ordinary clients may also maintain ties with several preeminent men. A big-man serves different sets of clientele with each institution that he controls. Consequently, the total constituency composing a leader-centered group is an aggregate of his lieutenants, the separate constituencies attracted by each of these lieutenants, and the distinct clientele served by each of the preeminent big-man's separate institutions.

Although leader-centered groups are themselves not corporate, the institutions that the preeminent big-man controls usually are. Nonetheless, these too are the organizations of their central leader, and when that leader dies or grows too old to command, often they decline or splinter into new leader-centered institutions, each organized around a new central figure and his personal following. This pattern is typical of a range of institutions, including joint families, caste associations, community charities (Mines and Gourishankar 1990), cooperatives (Mines 1984), and political parties in Tamil Nadu today.^[8] This pattern is also true of privately owned banks such as the chairman's, if to a somewhat lesser degree, despite the direct bureaucratic control over banks exercised by the central government. For example, consider how who the chairman is (a member of the Nagarattar caste) and knows (Annan, himself a Nagarattar) has affected how he conducts business (favoring Nagarattars) and whom he employs (again, Nagarattars). Should the bank fall into the hands of a non-Nagarattar, for example, bank policies will change,

just as who the new chair knows and relies upon will change, whatever the new chair's caste may be.

A key expression of a Tamil leader's individuality is the leader's public reputation. In Tamil society, the reputation of a respected leader is somewhat paradoxical. Eminent leaders are often economically successful, reflecting their skill as pursuers of self-interest. However, to avoid accusations of venality, a leader must circumscribe his or her successes with a reputation for altruism, honesty, and a commitment to the collective good of the community. In other words, a leader must achieve a reputation for behavior that is in effect a denial of self-interest. Yet, leaders also want people to know that it is they who are altruistic. A leader's patronage is not given anonymously, but publicly, with the knowledge of his community. It is to commemorate their altruism that community leaders typically engrave their names on the stairs and walls of temples, marriage halls, charitable societies, and libraries. They also publish their names and sometimes their pictures in the posters, announcements, invitations, and book-sized souvenirs that are associated with public events in order to make known who they are and to advertise their individual responsibility for deeds done for the benefit of their public.

Achieving simultaneously a reputation for both success and altruism is difficult because the two are seen as contradictory. Consequently, it is no surprise that the more successful a leader is, the more he or she will be accused of venality and corruption. This helps to explain the highly acrimonious nature of leadership in south Indian society.

In Tamil culture, the archetypal way a leader establishes a reputation for altruism is through patronage and involvement in institutions established to serve the interests of a constituency. By acquiring offices in such institutions and rights to symbolic honors associated with them, leaders dramatize their eminence and place themselves at the center of a clientele that they seek to serve. Competition for offices and honors is a part of the achievement strategy of leaders and would-be leaders alike.

In sum, the individuality of Tamil leaders lacks the characterizing values some scholars (e.g., Dumont 1970a) have associated with Western individualism, notably equality and liberty, which are associated with the sense of the individual as an equal among equals.^[9] Nonetheless, the identity of leaders does exhibit several other features of socially significant individuality, including individualistic social identity defined by public reputation, uniqueness marked by eminence, achieved identity associated with a deliberate striving after one's own gain, dominance and prestige, and autonomy marked by responsibility for who one is and

what one does. This form of public individuality, the individuality of eminence, therefore, rests on a recognition of achievement and of the individual as an agent.

Kasi’s Story: A Tamil Big-Man Explains His Civic Identity

I interviewed "Kasi," Kasiviswanathan, in his home in 1979, in the town of Trichengode, which is located in the western dry zone of interior Tamil Nadu. The town is a pilgrimage site because of its famous Saivite temple, dedicated to Ardhanariiswarar—Siva represented in a half male, half female form. The temple offers a dramatic sight, located on the top of a giant rock at the town's edge.

Born in 1912, Kasi was sixty-seven years old at the time of my interview and still very much involved in public life. He belonged to a Tamil weaving caste, the Kaikkoolars, and was well known in the region as a caste and political party leader.

As a younger man, Kasi had been a follower of the famous Tamil independence leader, "Rajaji," C. Rajagopalachariya, and until his forty-fifth year, was the president of the taluk (subdistrict) Congress committee. The Congress was the leading party of India's independence movement and the dominant political force in Tamil Nadu until the mid-1960s. I had been told by acquaintances of Kasi's that I must go to Trichengode just so I could meet him. Kasi was recommended to me in this manner because his friends considered him an important man, a man of accomplishment, one of the big-men of his caste. In their recommendation, his acquaintances described to me features of his individuality: his importance, offices, and reputation for past actions. Stressing Kasi's eminence, his friends emphasized his individuality.

When I interviewed Kasi, we met at his house. It was after lunch and the atmosphere was quiet; most in the household were taking their afternoon naps, as is the Tamil custom. On the wall was a signed photograph of Jawaharlal Nehru, dated 1929. Kasi had the relaxed manner of someone who had been interviewed many times before, but he seemed happy to be of service to me for a few hours.

When I begin a life story interview, I usually start by asking my informant about family—his or her parents and siblings, date of marriage, children, their education and what it is that they are doing today. Completing this section of the interview, I next ask my informant about himself or herself.

Kasi had a precise sense of what it was he needed to tell me so that I would know who he was and what he did and had done. His approach was identical to the way I had heard others describe themselves when they, like Kasi, had achieved some degree of eminence. He related to me first his status as the head of his family and family businesses. Next, he listed his current civic offices, and then he described the institutional offices that he had held earlier in his lifetime.

He said that he was the proprietor of all his family's businesses, meaning by this that he was still the head of his own joint family. He and his wife lived with their second son, while his three other sons lived separately. Nonetheless, all five men together managed the several enterprises that they jointly owned under Kasi's proprietorship. Kasi also ran a textile sizing business, which he owned in partnership with one of his brothers.

Kasi next described to me the offices he currently held in various civic institutions. At the time of the interview, he was headman (naattaanmaikaarar ) of the Kaikkoolar caste council in Trichengode and the joint-headman of the caste's territorial council for a region incorporating seven cities and their surrounding village hinterlands, known as the Seven City Territory (Eeruurunaadu). Trichengode was the territory's headquarters. In other words, Kasi was telling me that he was one of two preeminent caste leaders for the seven cities area. Next, Kasi said that he was the president of seven religious institutions and the president of a committee that collected construction funds for the Trichengode Government High School. He told me the committee had already built eleven rooms for the school, three in his family's name. When Kasi spoke of his family's name, he meant his name as family head. He had also built under his leadership and in his family's name a marriage hall for his caste, which was located in front of the rock temple, an auspicious location. He said ten to thirty marriages were conducted there by his caste fellows every year. Finally, he said, he was also a member of the Trichengode Rotary Club, an association that kept him in touch with all the associations in town. He kept his finger on the pulse of local events, he said, and remarked that his second son had been the joint secretary of the Junior Chamber of Commerce the previous year.

Next Kasi described civic offices that he had held in the past. Until 1977, he had been the vice president of the statewide handloom weaver's cooperative society, Cooptex, an important post. A friend, caste mate, and close political ally was the society's president. Kasi had also been a member of the state's Handloom Board for ten years and director of

three cooperative spinning mills, located in Salem, Tirunelveli, and Srivil-liputtur. And he had been the director of the Cooperative Union in Tamil Nadu, which is an advisory board to the government on production cooperatives. Finally, as mentioned above, when he was younger and had political ambitions, he had been the president of the Congress Party Taluk Committee.

At the time, I remember realizing that Kasi's purpose in describing his institutional offices was not simply to list his accomplishments for me, but primarily to describe his role as an institutional leader with multiple constituencies, some small, some large. His description reveals the contextual nature of civic identity and its spatial dimension, which expands and contracts in relationship to the size of a leader's constituencies. Over time, the size of the populations Kasi had served had changed substantially. Once he had been a state-level leader. Until his retirement from Cooptex in 1977, Kasi had served a statewide constituency that he had helped build. The handloom textile cooperative movement took its inspiration from Mahatma Gandhi and was until 1977 controlled by members of the Congress Party, which used it for the distribution of largess and the building of grassroots support. In those days, the handloom weaving industry was the largest category of employment after agriculture in the state, and so Cooptex appealed to what was theoretically a significant pool of potential political supporters. Throughout the state under Cooptex auspices, local leaders founded or managed production cooperatives, which tapped into the statewide association, enabling them to enact roles as patrons within their local weaving communities. Kasi told me that when a weavers' housing colony was built in Trichengode with Cooptex support, it was named after him, although he himself had favored naming it after Rajaji. As president of the Congress Party Taluk Committee, therefore, and vice president of Cooprex, Kasi was for a time a patron of statewide influence and eminence.

In 1979, when I interviewed him, Kasi was a leading townsman and regional caste leader serving a much smaller constituency. As co-headman of the Seven City Territory, he shared his preeminence as a caste leader with only one other man. That night, following the interview (see Mines 1984), I had the opportunity to observe Kasi in his role as headman during a council meeting that the leaders and representatives of the subcouncils of the territory attended. Kasi sat on a raised dais surrounded by the subordinate headmen and representatives of the Seven City Territory, and when I entered the meeting room, I had an immediate sense of his power. He was a chief among his lieutenants, a man with

the power to outcaste or impose fines on any local council that failed to adhere to Kaikkoolar caste rules regarding the governing of councils. That night the council of one of the towns that had been outcaste by a previous action came to plead for readmission. Symbolic of its submission, throughout the discussion of the case, a servant of the council lay prostrate on the ground, his arms outstretched before Kasi and his fellow Eeruurunaadu Council members.

Yet the caste council was only one of the many institutions within which Kasi held a preeminent role at the time of the interview. The other associations he described himself as heading were what Tamils term charitable institutions. The constituencies of these institutions—the seven religious institutions, the school building fund committee, and the caste marriage hall—were much smaller than that of the caste council. Each of these institutions, however, enabled Kasi to play the role of patron. As patron, Kasi cast himself in the role of an altruistic leader dedicated to the service of his community, his clients. He told me that he had donated Rs. 30,000 to the school building fund. Through these many institutions Kasi built for himself a grassroots political constituency, based on his personalized relationship with the circles of clients associated with each of the many institutions he managed.

Listening to Kasi describe himself, I realized that his intent was to make clear to me his public self. He was telling me that his eminence, his institutional roles, and his influence among others distinguished him. These features define Kasi's "civic individuality." Kasi even described his identity within his family in these terms, stating that he was the head of his joint family and proprietor of all his family's businesses. It also became apparent that having control over decisions affecting himself and others, an ability which is associated with statuses of eminence, was an important feature of his individuality because control gave him autonomy.

When a person such as Kasi describes his civic identity, his telling is filled with implied authority, but it is also always impersonal, revealing little about the inner man. This impersonality is a distinguishing feature of the way Tamil men express the civic dimension of their individuality. It was later, when I asked Kasi about his personal goals and dreams that he told another story, the story of his motivations, hopes, and disappointments. In his telling, he revealed himself as an idealistic young man who had entered politics in the heady days of Mahatma Gandhi, Rajaji, and the Indian Independence Movement. Kasi personally bad known Rajaji. He told me that politicians then were motivated by ideas and ideals, while today politics is "all personalities." Politicians no longer

care about ideas. He regretted the years he had spent as a politician after Independence was achieved, and he was disgusted with modern politicians. He felt betrayed by them.

The layered way in which Kasi reveals his identity is typical of men who have achieved some public eminence in their lives: First they tell of their civic identity. It is later, when they are questioned about their goals and aspirations, that they reveal the private dimensions of their lives. Civic lives and private selves are not, then, two separate identities. Kasi has a sense of himself within society and a sense of how others know him. He also has an inner awareness of himself. Each of Kasi's senses of who he is affects the other. Personal talents and abilities, private desires, hopes, and disappointments drive motivations and give force to achievement in civic life. They energize its drama and pathos. They lie behind the manifest public individual. (Tamil descriptions of the inner self are the subject of chapters 7 and 8.)

Finally, it must be noted that Kasi did not describe his individuality in opposition to groups, but rather in terms of who he was and what he did within groups. In this context, it is important to distinguish between Kasi's civic identity and what Dumont labels a collective identity. In the latter case, the person has the identity of his or her group. In Kasi's case, his identity is his own, but it is an identity that is built on his eminence within groups, an aspect of his contextualized identity.

The Community and the Individual

The examples of the goldsmith, Tambi, and Kasi illustrate the importance in Tamil society of personalized relationships and of leader-centered groups, configurations of interaction that require a close connection between organization and individual agency. But just how important are the agency of leaders or the actions of ordinary persons as influences on the way urban communities are organized and integrated? Part of the answer lies in how civic individuality is valued. I have suggested that eminence implies a community that shares values, as does reputation. What, then, is the relationship between these values, which are features of the interaction between individuality and community, and the formation and dynamics of an Indian urban community? Or, put differently, what does an Indian urban community took like, and how does its appearance reflect this interaction between shared values and individual interests and agency? And finally, just how important are communities, castes, and individual identities in a large cosmopolitan Indian city today? What are

the motivations and involvement of ordinary citizens in community, and what are the currents of change?

I argue in the chapters that follow that the relationship between individual and community has been historically dynamic in India, and that, as circumstances change, individuals rework their relationships with their communities. In the process, they transform the meaning of both citizenship and community. Leaders are important agents of this community transformation. Forever competing to enlarge their constituencies, they must from time to time redefine what they consider their own interests in light of changing circumstances and adjust how they make their appeals to ordinary citizens to suit.

The roles and motivations of individuals are reflected not only by the organization of urban relationships but also by an urban community's very structure and layout. I take as the focus of my study the old walled town of Madras City in Tamil Nadu, south India, the area of the city that is today known as George Town. I argue that the organization of George Town reflects the history of a series of Indian solutions to the problem of how, in the highly competitive context of commerce, to guarantee trust and reliability in relationships. What the organization of the urban community shows is that the freedom of the individual is always circumscribed by an ethic that sets the standards by which the civic individual in different roles must behave in order to establish trust-based relationships. Individuality is not eliminated by this imposition of constraints on behavior but, rather, is emphasized, although emphasized differently from the way it is in the West.

Chapter Three
Institutions and Big-men of a Madras City Community
George Town Today

George Town and Fort St. George, just to its south, are the oldest parts of Madras City (see map 2). One can still see fragments of the old wall fortifications that once circumscribed and protected George Town, while the fort still has its walls and traces of its moat. The fort must be entered through heavy iron-studded doors, which remind the visitor of the uncertain peace of former times. Within the old walled area of the Town the streets are narrow, often choked with every sort of vehicle, cycle rickshaws carrying boxes of freight, small children on their way to or from school, and sometimes men holding great sheets of metal, which they have purchased for purposes unknown to the observer, slowly being pedaled among the crowds; there are brightly painted lorries, hand-pulled carts, bullock carts, bicycles, scooters, automobiles, and auto rickshaws buzzing about the Town's outer edges, trying to avoid getting stuck in the traffic jams. Pedestrians clog the streets' remaining spaces.

The Town is the financial center of Madras. American Express has its office here, and so does the Madras Stock Exchange. The Town is where the descendants of Madras City's oldest banks are located; it is also a center of informal banking. The big moneylenders, the "shroffs" of old, dealers in bullion and cash, operate here, so too, the Cashiers, the Kasukaarar,^[1] the accountants of the big banks and enterprises. George Town is also the business center of the city. Madras harbor lies just across the railroad tracks on the Town's eastern edge. Once there was no harbor, only the open sand beach on which small wooden boats departed and landed through the persistent surf, ferrying goods and people to and

Map 2.
Contemporary George Town and Fort St. George, Madras City.

Figure 2.
A view along First Line Beach Road. The photo was taken from the
roof of the Beach Railway Station. The large building on the left is the local head
office, State Bank of India, formerly the Imperial Bank of India during the colonial
period. George Town, 1992.

from ships anchored at sea in the "Madras Roads." Today, Madras has an artificial harbor, which is fitted to receive the giant container ships that ply international waters, although one can still see ships like tiny dashes on the distant horizon, "standing" in the roads waiting their turn in the harbor. For convenience, the Custom House and the General Post Office, the city's main post office, are located near the harbor on North Beach Road, now renamed Rajaji Road.

As the city's business center, the range of enterprises, large and small, located in the Town area is extensive: jewelers, gem dealers, book sellers and publishers, moneylenders and bankers, wholesalers and retailers

of iron and steel, nonferrous metals, textiles, fancy feathers for the fly-fishing industry, fireworks, electrical goods, leather, pharmaceutics, locks, and metal hinges—the list is enormous. Several of the oldest business houses, houses whose names reflect their British colonial past, have home offices here: Binny's, Best and Crompton, Castrol, Parry's, Gordon Woodroffe. This area of the Town, just across from the harbor, is called Muthialpet. Mannady, the adjacent area to the southwest, is considered by old residents to be an integral part of the Muthialpet neighborhood. Popham's Broadway, running north and south, marks the western edge of this eastern section of George Town.

Across Broadway, inland and to the west, is the old community of Pedda Naickenpet, although, today, this area, too, is subdivided. The northern section is still called Pedda Naickenpet, but the southern section is now called Sowcarpet. North Indian Hindu and Jain merchants live here in large numbers, and south Indians say that they feel transported to north India when they step into these streets. Sowcarpet is where many of the big moneylenders, gem dealers, and some of the Tamil-speaking goldsmiths live and operate. The buildings are mostly old houses to which one or two stories have been added over time. On the narrow veranda of one, a goldsmith and his three assistants manufacture 22-karat-gold bangles and necklace "chains." Nearby, a Jain gem dealer sits resting against a long white bolster, talking quietly to another man, his wooden desk cum money-box by his side. Large sums of money change hands in these modest surroundings.

At its heart, Sowcarpet incorporates the city's main green grocers market, Kothawal Chavadi, where I once bought one hundred limes for Rs. 2, about U.S. $0.26 at that time. Business is so intense here that wholesalers rent shop space by the hour. A single shop may have several tenants operating in series, each for a few hours, through the course of the twenty-four-hour day.

Just to the south of Sowcarpet is Park Town, famous for its jewelers and iron merchants and the location of George Town's most popular temple, the Muttu Kumaraswami, or Kandasami temple.^[2] This is the temple to which I was going on a rainy night in September 1985, to begin my fieldwork in George Town.

It was my plan over the next ten months to study leadership and individuality among two of the Town's prominent castes, the Beeri Chettiars, one of the foremost merchant castes of the Town, and the artisan caste once named the Kammaalans (Smiths),^[3] now usually called Acharis or Viswakarmas. The Beeri Chettiars had been close associates

of the British, from the founding of Madras in 1639, up to Indian Independence in 1947. I had chosen the Viswakarmas because they, like the Beeri Chettiars, had once been a leading caste of the left-hand section of castes in Madras City. Their historic rivals were the castes of the right-hand section, especially the famous merchants, the Komati Chettiars, the community that owned and controlled Kothawal Chavadi. While these once bitter rivalries and the left-hand/right-hand moiety division that framed them are now memories, and although today's descendants of these castes pride themselves on their friendly relations, the history of Madras is marked and shaped by riots and competition between the two factions. The area east of Popham's Broadway, Mannady and Muthialpet, was once the territory of the left-hand castes, the area to the west, Pedda Naickenpet, the territory of the right-hand castes. What legacy had this old division of the Town left behind to shape the civic communities of today?

A few days before my first night of fieldwork, I had been discussing my research intentions over tea with an acquaintance, C. Gourishankar, and his friend, "Babu." Babu, recently retired, had volunteered that one of his former colleagues at All India Radio was a Beeri Chetti who also happened to be the head managing trustee of the Kandasami temple in George Town. I asked him if he could help me contact this man. I thought he might make a good starting point for my research.

Babu made the arrangements. We were to meet with his former colleague, Tiru P. Balasubramaniyam ("Bala"), at the temple office a few nights later. Babu had explained my interest to Balasubramaniyam, and he, "Bala," had indicated that he would try to have present at the meeting a few knowledgeable older men of the community.

Although it was beginning to rain, the three of us—Babu, Gourishankar, and I—managed to find an auto rickshaw willing to take us to the Town. It was dark and wet when we arrived at the temple. We left our sandals at the sandal concession next to the temple's main entrance and entered through the Raja Gopuram, the kingly tower that faced onto the street. High on the tower's face a neon sign depicted in Tamil the word "Ore," the primeval sound of meditation. Inside, the temple was stone and cement, cavernous; we made our way toward the back, past a video shop selling religious films, to the office. Several men, all Beeri Chettiars, sat around a table. Along one wall were large steel cabinets, marked with sandalwood paste and vibuti , symbols of Siva.

Over the next two hours, Bala and his associates spoke rapidly, sometimes singly, sometimes several at once, so that it was all that I could do

Figure 3.
Rasappa Chetti Street showing the Kandasami temple entrance tower
or Raja Gopuram. George Town, 1992.

simply to follow what was being said, let alone take the kind of notes I wanted to. In those two hours, this small group of men outlined what they thought I should know about the Beeri Chettiars of George Town. They told me about the Kandasami temple, its history, and why it was important to the Beeri Chettiars of George Town. They told me the names, families, and brief particulars of who in their memory had been the leading men of the Beeri Chettiar community before families had begun to move away from the crowds and congestion of George Town.

Figure 4.
Kandasami temple roof line.

The list included bankers, the cashiers of important British companies, a lawyer politician, and several very wealthy merchants. A few leaders had belonged to families in which sons or sons-in-law had succeeded fathers and fathers-in-law as influential men. I could tell that everyone present knew of these men, including the non-Beeri Chettiar friends who had come with me, even though some of the leaders had been dead for decades. Among those present, there was a shared sense of the George Town Beeri Chettiars as a community distinguished by who its leading citizens had been. It was also clear that several temples and especially the Kandasami temple figured importantly in both the identity of the community and its leading men. Bala then told me who he was and described the schools and charitable institutions that he had helped found as a leader within his community. Next the men told me a little of the history of the Beeri Chettiars in George Town. It was their sense that the community vitality that had prevailed in their youth was weakening. Families had moved away, and now people often did not even know who their neighbors were. Bala said it was a goal of his, as head trustee of the

Kandasami temple, to reestablish some of the old sense of community. Finally, the men described some of the relationships the Beeri Chettiars had with other temples and other neighborhoods in Madras and surrounding towns. What they were describing to me, albeit in outline, was the identity of their civic community—a community that they saw as threatened, a community that was identified with its preeminent men, its temples, and its institutions—and the relationship to that community that Bala and some of the other leading members of the caste held as individuals. These were the topics upon which my research was to concentrate. They were giving me a quick overview. There are several features of this overview that told me much about the nature and role of civic individuality in Madras today.

When I left the meeting at about ten that night, it was raining heavily and the city was dark. The city lights had failed in the storm and the streets were flooded in places. The normally congested streets of George Town were almost deserted except for a few men wading slowly through the deep puddles.

Exploring the Nature of the "Institutional" Big-Man

Later, thinking about what Bala and his friends had said, I was struck by how personalized their description of the Beeri Chettiars in George Town had been. Names had flown about the room as they had described the caste's past and present distinguished men, the caste's periyar .^[4] They had told me which houses had belonged to these men, who they were related to, what they had done, and what their roles in businesses and civic institutions had been. My friend Gourishankar, aged sixty, obviously had known who they were talking about. He had known some of the men when, forty years before, he had attended Christian College, which in his day had been located in George Town. Reflecting on what had been said, I realized Bala and his friends had described their community to me in terms of its big-men and their institutions. I realized that if I were to share their sense of familiarity with their community, I needed to understand what it meant to be a Beeri Chettiar big-man.

In the months that followed, I was to hear informants use a number of terms denoting leaders. I have listed some of these in my previous chapters. In addition to the term periyar , I learned that periyadanakaarar (big-gift-giver) is a related term in a class of similar terms for informal leaders, including talaivar (headman) and ejamaanan (master, headman).

These terms are used to mark the preeminence of persons within their communities and, in some cases of special fame, within the larger society. Mahatma is a similar Sanskrit term, familiar as a designation for Mohandas K. Gandhi, and lexical equivalents of the Tamil terms for big-man are general to Indian languages. In fact, Pedda Naickenpet, the name for the area of George Town west of Popham's Broadway, conveys this same sense of big-man. Pedda is Telugu for "big," and Pedda Naickenpet means the place (pet ) of the "Big Naicken," the title of the headman who used to control this part of the Town in the early years of British East India Company trade. During those early times, in the shadow of the Telugu-speaking Vijayanagar Chiefdoms, the Naicker caste was politically dominant in the locality.

As the head trustee (dharmakarttaa ) of the Kandasami temple, Bala is himself an "institutional" big-man among the George Town Beeri Chettiars. As indicated in chapter 2, I add the "institutional" qualifier to the Indian big-man concept because men like Bala attract followers and enact their roles as generous leaders through the "charitable" institutions that they control.^[5] For south Indians, a "charitable institution" is a highly personalized leader-centered association or group that is designed by its founder or leader to benefit a particular clientele, constituency, or community. The leader attracts a following by the benefits he provides. In return, his followers reward him with prestige and an eminent reputation, attributes that give him great influence and discretion among his constituents. When his eminence is great and his institutions particularly effective, then a big-man's reputation also spreads among outsiders. People will know that he is a man responsible for getting things done among his constituents. Like Annan's banker of the previous chapter, an important big-man is a very valuable person with whom to have connections. In Bala's case, the Kandasami temple is one of several community institutions that he heads, but he founds his reputation and so his leadership in these other institutions on his role as the head trustee of the Kandasami temple.

Institutional position is a necessary condition for the viability of the Indian big-man, but it is not sufficient. In keeping with their highly per-sonalized nature, Indian institutions, including temples, expand and contract in popularity and membership depending on the idiosyncratic charisma of their heads. Although it is often supposed that local Indian leaders can depend on ready-made caste and kin constituencies, the fact is that even hereditary leaders have few followers when they lack charisma and skill. Indian society is salted throughout by these "hollow

crowns"—would-be leaders, the heads of associations who have little or no following (see Dirks 1987). In the case of the Kandasami temple, townspeople told me that they thought Bala a particularly effective leader and that the popularity of the temple has greatly increased during his tenure as head trustee.

As I reflected further on my hosts' naming of their community's past big-men, I began to be aware that knowing the names and offices of many people is a concomitant of leader-centered group organization. I sensed that if a person wishes to navigate a fruitful public life in a society composed of such highly personalized institutions, then it is important to know who heads and is responsible for what, as well as who has connections with whom. This is one of the reasons why Tamils keep tabs on so many people. A person who can claim good connections with influential people finds it is easier to accomplish social objectives and to influence others. Knowledge of this sort and the ability to claim connections is valuable and can be misused by dishonest persons. Reflecting this, when much later in my fieldwork I was collecting Bala's genealogy, he suddenly remarked that the information I was gathering could be valuable to unscrupulous persons who might use the information to claim close ties with him. He asked me to be careful to whom I showed it.

Why, then, did Bala and his friends list the past leaders of their community? It was clear to me from the way they spoke of these men that they did so in order to stress that who the George Town Beeri Chettiars are today is in part built upon their community's connections with men of prominence and responsibility from their recent past. They see their community as a composite of families and other associations that big-men head and the connections that exist among these elements, a community centered on eminent persons, their institutions and groups, and on the connections that link them, forever in the making.

Temple, Trustees, Donors, and the Civic Community

A Tamil villager once told me that a community without a temple was unfit for residence. The temple, he said, indicates that the community is graced by the presence of God and that its citizens form a moral community. A community identifies and is identified by others with its temples. Bala and other Beeri Chettiars went to considerable length to explain to me the nature of their community's identification with the

Kandasami temple and the role that big-men played as its patrons and managers. The temple and its functions symbolize the caste community, and publicize its leading associations and who its leading men are. Stories about the temple's history and endowments reveal as much.

That night in the temple office, Bala and his friends described the Kandasami temple as a "denominational" temple, meaning a temple controlled and managed by a single caste community, in this instance, themselves, the George Town Beeri Chettiars. What makes the temple a particularly important institution of big-man leadership among the Beeri Chettiars today is that it is the primary and by far the wealthiest institution controlled by the caste as a whole and, in much the same manner that villagers use their temples, it is used by leaders to represent the caste as a civic and moral community to the world at large. It is also importantly a charitable institution, the caste's central repository of resources that exist for the benefit of the caste as a civic community. An individual who is elected to the managing board of trustees of the temple is elected, therefore, to a position of leadership within the Beeri Chettiar community with control over its main assets. Among the five trustees of the temple, the head trustee, the dharmakarttaa , is preeminent. It is Bala who holds this position. He is the periyar , the big-man, and he is a preeminent figure in his community.

Until 1980, the electorate of the Kandasami temple included only male Beeri Chettiars who lived in Muthialpet, Mannady,^[6] and Park Town, but because by that time increasing numbers of families had moved to other parts of the city, caste leaders changed the bylaws of the temple to include in its congregation male Beeri Chettiars living or doing business anywhere in greater Madras City (Madras High Court records). The Madras Beeri Chettiars, therefore, today form the temple leaders' constituency. And this constituency, as a group, constitutes the caste's civic community defined most broadly. But even with this change in bylaws, the Park Town-Muthialpet Beeri Chettiars constitute the core congregation of the temple, and temple trustees have always been selected from among the caste's George Town leaders. It is these leaders who are and always have been the principal donors to the temple and sponsors of temple functions, and it is because of them, and because of the location of the temple, that George Town remains the geographic heart of the Beeri Chettiar's sense of their civic community in Madras City.

Because popular temples such as the Kandasami temple are important institutions of civic leadership, control of them is often contested. In George Town, the leaders of several castes would like to gain special

rights in the temple, and some conspire to dislodge the present temple trustees with this aim in mind. These contenders pursue a variety of strategies, among them bringing lawsuits claiming that members of other castes have made donations to the temple and so, since the Beeri Chettiars are not its sole financiers, they should not be its exclusive managers.

In and out of court, Bala and his allies have countered these pleas, asserting that the caste's right to exclusive control of the Kandasami temple is based on what they argue has been more than three hundred years of unbroken management and on a legend that the temple was founded by two old friends, Velur Mari Chettiar, a Beeri Chettiar, and Kandappa Achari, a Viswakarma man. According to this legend, the two friends were on their monthly pilgrimage to worship Lord Murugan at Tiruporur, fifty-six kilometers away, when they miraculously discovered the idol of Kandasami hidden in an anthill and brought it back to Madras. There, on an auspicious day in 1673, they installed and consecrated the deity in a temple dedicated to the elephant god, Vinayakar, located in the garden of one Muthiyalu Naicken of Pedda Naickenpet. Subsequently, when Mari Chettiar sought to build a temple for the deity, funded in part by his wife's generous gift of her jewelry, Muthiyalu donated the Park Town lands on which the temple now stands. When Mari completed the temple, he handed its management and that of its financial trusts to the "eighteen group" Beeri Chettiars, the eighteen named clusters (gumbuhal ; sing., gumbu ) that composed the Town Beeri Chettiar community at that time. In commemoration of his services, the Beeri Chettiars installed a statue of Marl Chetti near one of the temple's sanctums, where he is worshiped today as a god. Here we see an individual, Mari Chetti, being commemorated for what he had done.

Aside from this legend, lists of donations, and a few undocumented stories, little specific historical detail is known of the temple. Nonetheless, challenges by covetous leaders of other castes to the exclusive control of the temple by the leaders of the George Town Beeri Chettiars have been unsuccessful so far. What historical evidence there is of Beeri Chettiar control has been too strong.

From archival materials, endowment records, and stone inscriptions in the temple we do know that Beeri Chettiar control of the temple is at least two hundred years old. F. L. Conradi's 1755 map (map 3) of "Madraspatnam," as the city was then called, depicts a small unnamed shrine at what is the temple's location today (Love 1913, 2: endpocketmap). We know that the temple was renovated and sanctified as a brick temple in 1780 by the "eighteen group Beeri Chettiars." We know that

Map 3.
Madras in 1755. Based on "A Plan of Fort St. George and The Bounds
of Madraspatnam" by F. L. Conradi.

about 1865 the temple was rebuilt of stone in its present form, and that in 1869 a generous man named Vaiyabari Chettiar donated Rs. 66,000, an enormous sum in those days, to establish a trust to fund various temple functions. He also built the large temple car (teer ) in which the processional idol of the god is carried on the seventh day of the main annual festival (brahmootsavam ). We also know that around 1880 Akkamapettai Govinda Chettiar and Narayana Chettiar donated the land next to the temple for the purpose of building a large community hall, the Spring Hall (Vasantha Mandabam). And we know that in 1901 Kali Rattina Chettiar, a wealthy businessman and the father-in-law of Bala's father-in-law, donated Rs. 50,000 to build the temple's entrance tower (gopuram ) and, in a dramatic gesture still spoken of with awe, gave a cup of diamonds for jewelry to decorate the idol.

Richly endowed, the temple today owns more than sixty houses, most located on prime urban land. The head temple priest's house, which is rented to him by the temple, gives an idea of values. In 1986, its worth was estimated at Rs. 10-15 lakhs (1 lakh equals 100,000 rupees), $100,000 or more at a 1986 rate of exchange. A few of the endowments have been especially grand. The previously mentioned gift of a cup of diamonds and Rs. 50,000 by Kali Rattina Chettiar is one of these, a gift that in those days was worth many times the value of a house. Another is Bala's uncle Venugopal Chettiar's gift to the temple of ten grounds of urban land, which are today the site of a one-thousand-student Beeri Chettiar grammar school founded under Bala's administration. Other endowments sponsor particular festivals; they buy flowers and textiles for rituals and clothes for the deity. Yet others fund building, renovations, and cultural events. Today the temple controls hundreds of millions of rupees in assets.

The rich endowment of the temple reflects its popularity and the affluence of the Beeri Chettiars. As a measure of the temple's lively appeal, the temple concession that looks after the sandals of worshipers annually earns about Rs. 36,000 by charging customers a small fee of ten paise^[7] for safeguarding their footwear while they go barefoot into the temple. Given that many locals leave their sandals at home when they are going to the temple and avoid the charge, this sum equates with at least 360,000 individual visits to the temple each year. This popularity is especially evident during the Spring Festival (Vasantha Brahmootsavam) when the god, garlanded in flowers and bedecked in gold and diamond jewelry, is taken on lengthy nighttime processions. On these nights, when the processions are longest, crowds gather in the streets for

miles. These crowds are the audience before whom the trustees play out temple pageantry depicting the trustees' role as patrons and the wealth and importance of the Beeri Chettiar civic community.

The temple's rich endowment also reflects the sense each endowment donor has of his or her civic individuality, since endowments state something about who each donor is in relationship to the Beeri Chettiar community: that he or she is an acknowledged member of the community and makes his or her gift in the interest of the caste's collective good. Through his gift, the individual achieves for posterity a respected reputation within the community congregation. Over the centuries Beeri Chettiars, singly and as associations, have made numerous donations, both large and small, slowly building the temple's wealth. Individual donors without children who have left houses and property to the temple are commemorated in inscriptions and posters that list donations. Without descendants, their donations must preserve their identity within the civic community and keep alive a memory of who they were.

A Viswakarma once told me that a person with lots of gold has abundant strength and fertility. So, too, a community. On the night of our initial meeting, demonstrating the temple's wealth, Bala first described the gold ritual vehicles and processional objects possessed by the temple, and then, with the others at the meeting, we left the office to examine them in their locked sheds. He made clear the connection between donors and sponsors and particular ritual objects: the Beeri Chettiar Iron Merchants Association donated the gold crown worn by the processional idol and all the gems that encrust it; the Town's betel leaf^[8] merchants donated approximately 2.5 kilograms of gold for the gold peacock processional vehicle (vahana ); the shroff (bankers and dealers in bullion) merchants donated 4.0 kilograms of gold for an elephant processional vehicle. And the Town Beeri Chettiars as a community donated 3.0 kilograms of gold for the processional palanquin of the supernatural warrior-hero and ally of Kandasami, Surabatman.

The temple also has a silver-plated car, strung with colored lights, constructed with temple funds. Temple cars have a pyramidal form, ornately decorated with carvings, temples on wheels. Some are huge juggernauts, towering twenty to thirty feet or more. Electric wires obstruct the passage of these biggest of cars, and today only a few are taken on procession in the city. Others, smaller, are designed to pass below the city's electrical lines. The garlanded and jewel-bedecked idol is carried in the car-shrine during processions, often with priests sitting before it in order to accept and present offerings submitted by worshipers. In

Figure 5.
Gold-plated temple car (teer ) with the Kandasami gopuram in the
background. The processional idol of Kandasami may be seen riding in the car.

1984, in his role as head trustee of the temple, Bala himself built for the temple with temple funds a gold teer plated with 7 kilograms of gold, one of twelve in southern India, and a significant new expression of the temple's claim to importance.

In their opulence, each of these ritual objects declares to all who see or hear of them the vitality of the George Town Beeri Chettiars and ide-

alizes the altruistic commitment to the civic community of the associations and leading citizens who gave. Of course, everyone recognizes that the objects also make great advertisements for the donors and the Beeri Chettiars as a community and boost reputations. Bala and the others were showing me that the Kandasami temple is regarded by the Beeri Chettiars as a key institutional symbol not only of their community as a whole, but also of its leading citizens and associations.

Temple Processions, Community, and the Eminence of Leaders

If the temple is a symbol of the community, then the temple's builders and the officers of the temple and its events are key agents of the community. They take responsibility for the community's image and in doing so draw attention not only to the community, but also to their own individual importance within it. Without leaders, there is no community; without a community, there are no leaders. Civic leadership and community go hand in hand.

During festivals and important times of worship, which are collective community events, temple trustees and other leading citizens among the Beeri Chettiars enact ritual roles that single them out and publicly dramatize their eminence as instrumental individuals within their community. During a special worship, for example, the sponsor stands closest to the deity among the spectators and acts the role of a dignified but humble host to important guests and spectators (fig. 6). South Indian culture values the leader who is generous, who acts on behalf of his followers. The leader's actions and position near the god mark him as responsible for the event and indicate that he is important. It will be known that he is responsible for accomplishing things in other social arenas as well. Should he show special respect to others while enacting his ritual roles, this recognition confers on them public prominence. The sponsor's civic individuality, therefore, is expressed partially by his eminence and his seemingly selfless agency, his status as a temple trustee or as a sponsor of a ritual event. After the worship he will receive tokens of special respect from the head priest, marks of honor, such as a large quantity of prasadam (food "left" by the god) that he can distribute, in recognition of his instrumental role as sponsor. When a temple event is important, great prestige is attached to these symbols of respect because they indicate a man's significance within his community, and for all the outward appearance of altruism, the reality is that leading men compete bitterly to achieve or protect rights to them. By the mechanism of his own re-

Figure 6.
Tiru P. Balasubramaniyam, head trustee or dharmakarttaa of the Kandasami
temple, shown here with the decorated processional idol of Kandasami, who is flanked by
his two consorts. The humble demeanor of Bala is characteristic of a sponsor of worship.
George Town, 1992.

distribution, the recipient of honors confers respect on others as well. In a few instances at Kandasami festivals, leaders belonging to other castes are also publicly marked with ritual honors, indicating their status as respected but subordinate allies.

The most important of the Kandasami temple festivals is the annual twenty-day Spring Festival. During this festival the god, Kandasami, is treated like a king, engaged in the activities in which a real king might be engaged at the onset of spring. He leaves his palace, the temple, and goes out on excursions (mandaga padi ), and, symbolizing the annual renewal of life, he reenacts his marriage to his two wives. His love marriage to his second wife, Valli, a mythical tribal princess encountered on one of his excursions, is an especially popular ritual event among women.

The twenty days of the festival are a period of celebration among the Beeri Chettiars, a time of spectacular entertainments and grand ritual events. Under a huge pandal , or thatched roof, funded by the Vaiyabari

endowment and directed by Bala, the temple and Kandan Arts, an association founded and headed by Bala, sponsor public concerts in the street that runs before it, and the crowds that come out to celebrate and worship are large. The temple processions reach out like tendrils, entwining the constituencies the temple serves, providing leaders associated with the events opportunities to be seen publicly in roles that declare their individual importance as the community's civic leaders. Ideally, the more prominent a leader's ritual role, the greater his fame and influence. In reality, the more important a leader's ritual roles are, the more apt his claim to eminence is to be contested. The politics of temple leadership reflect the competition of local leaders for prominence as individuals within their community. Temple offices and ritual roles are their "badges" of prestige and honor, badges that in most cases must be individually achieved.

Temple processions go where the trustees and sponsors have clientele, persons who will be attracted to the event and among whom the leaders wish to publicize who they are. For example, over the years, north Indians, who do not participate in such festivals, have gradually replaced the south Indians who used to live around the non-denominational Ekambareeswarar temple, just to the north of the Kandasami temple (see map 2). As a result, the trustees of that temple have modified its processional routes so that they leave the area behind and enter the residential area of south Indians. There the roles of the trustees are still appreciated. However, processions cannot go anywhere a temple's management might like to go because other temples and their leaders mark off their own domains. A temple's domain is the constituency area of its leading men, and if a procession were to enter uninvited the domain of a temple controlled by a different set of leaders, the incursion would be seen as an act of aggression, an attempt by the encroaching temple's leaders to poach on the clientele of the other temple. In George Town, where prominent castes control their own denominational temples, an elaborate processional etiquette exists, which historically has been much contested and negotiated. Reflecting the old left-hand/right-hand moiety division that separates Muthialpet and Mannady (the domain of the left-hand Beeri Chettiars) from Pedda Naickenpet (the territory of the right-hand castes), even today no Kandasami temple procession enters or passes through Pedda Naickenpet (map 4).

The Kandasami temple has several processional routes (see maps 5 and 6), which declare the influential reach of the Park Town Beeri Chettiars. One circumscribes the block in which the temple is located, the

four streets known as the Chinna Maada Viithi (Little Temple Streets). A second begins like the first but then goes north to the Ekambareeswarar temple along Mint Street, then returns along Nyinappa Naicken Street, Evening Bazaar Road, returning along Mint Street. These streets are known as the Periya Maada Viithi (Big Temple Streets). A third route, now defunct, used to take the idol to the Angalaparameeswari temple for the benefit of the Beeri Chettiars who live in Sulai (Choolai), in west Madras. A fourth route proceeds along N.S.C. Bose Road, Thambu Chetti Street to the Kachaleeswarar temple. A fifth, which used to go to the Harbor Beach, now travels further south to the Marina Beach because the harbor has been closed off, making its beach inaccessible. The sixth runs the length of Thambu Chetti Street, leaves the Town area, and continues north deep into Royapuram. This processional takes Kandasami out beyond the northern border of the Town to what once was countryside. On its return, it travels along Lingi Chetti Street before returning to the Kandasami temple. Bala is now talking about developing yet another processional route that will take the idol to the Kapaleeswarar temple, a major temple located in Mylapore, several miles to the south.

The Community Center and Its Galaxy of Institutions

In the recent past, the Kandasami temple was only one of several religious institutions that the leaders of the George Town Beeri Chettiars used to organize their community. Although the community is no longer so tightly knit, caste members still understand the organization of their community in terms of these institutions and the framework for leadership they provide. To conjure a picture of community and leadership, therefore, it is necessary to describe these institutions and the relationships that once characterized them.

Although it is today the principal remaining community institution, in fact the Kandasami temple was once merely an important satellite temple of a central aggregate complex of institutions located in Muthialpet that the Town Beeri Chettiars used as institutions of leadership and community during the eighteenth, nineteenth, and early twentieth centuries. That the Kandasami temple is among the last institutions with castewide significance in George Town is an indication of just how much community solidarity among the Town Beeri Chettiars has declined since the turn of the century, a decline that has accelerated since the 1950s.

Map 4.
Kandasami temple long processional routes and destinations.

Map 5.
Kandasami temple processional routes to Choolai and to Marina Beach.

Until the late 1950s, caste leaders controlled a big-man centered galaxy of ritual institutions, which together formed a mandala, each point in the layout of which was a satellite that served its own constituency and was headed by a different mix of leaders.^[9] In the case of the George Town Beeri Chettiars, this galaxy framed a minipolity composed of the caste's leading members, including caste headmen (sing. ejamaanan ), and their constituencies, called "clusters," which together constituted the civic community of the caste in the Town.

Map 6.
Kandasami temple processional routes to Kachaleeswarar temple and to Royapuram.

In addition to the Kandasami temple, caste leaders controlled two other major kingly temples in Muthialpet, the Kachaleeswarar temple, built around 1725, which, prior to the twentieth century, was the caste's central temple, and the nearby Mallegeeswarar temple, first referred to in British records in 1652 (Love 1913, 1:119n). Each of these temples was an important satellite in the caste's institutional galaxy (see maps 2 and 3), although in the 1950s the caste lost control of both.

Figure 7.
The north end of Lingi Chetti Street, with the Mallegeeswarar temple
and mandabam . The tin shed to the right of the Raja Gopuram houses a large
temple car. Muthialpet, 1992.

In addition to these temples, other satellite institutions filling out the Beeri Chettiar's George Town mandala were a large agraharam , or residential street of priestly Brahmans, associated with the Kachaleeswarar temple, which was a kind of priestly center to the Town Beeri Chettiar's civic domain; a caste monastery, located within the agraharam , the

Figure 8.
A view of part of the agraharam that surrounds the Kachaleeswarar temple on
three sides. The first building on the right is the Abanatha Dharma Siva Acharya Madam,
the Beeri Chettiar's caste madam or monastery. Muthialpet, 1992.

Abanatha Dharma Siva Acharya Madam, which housed the caste's celibate Brahman guru and was where leaders of the caste community met in council in the nineteenth century; the caste death-rituals shrine and tank, the Attipattaan Kulam, located some distance from the caste's residential streets, and a host of lesser temples and "public" institutions and locations in the Town area and its surroundings within which caste leaders maintained rights. One of the more important of what I am labeling here "lesser" temples included the Kaalahasteeswarar temple, which the Beeri Chettiars built in 1640, one year after the British founded Fort St. George at Madras (Love 1913, 3:388). The temple is located on what became Coral Merchant Street. There are numerous others, some fairly large, some small, built by the caste.

Institutional galaxies, such as the Beeri Chettiar's, are loosely integrated, since each of its satellites can conceivably stand alone and is typically characterized in the public's mind by its current preeminent leader

and lieutenants. These institutions are, therefore, what I am calling big-man or leader-centered institutions, not only because locals sometimes refer to their top leaders as periyar or periyadanakaarar (big-gift-givers), but also because these leaders use the institutions they control to attract their followers (see Mines and Gourishankar 1990; Mines 1992). It is the pyramidal overlapping leadership and constituencies of these institutions, a layering of alliances and tacit agreements among big-men and ordinary caste members, rather than any overarching system of administration, that makes them satellites in a galactic polity. However, as I describe in chapter 4, in the nineteenth century, the principal headmen of the eighteen named gumbuhal of the caste did meet in council, the Periyagramam, to jointly administer caste affairs and to enforce caste codes (ILR, 135). But at that time each of the eighteen gumbuha1 of the Beeri Chettiars also formed a distinct constituency under its own head-man and handled its own affairs with the recognized proviso that when differences between gumbu actions and Periyagramam dictates arose, the gumbu headman deferred to the Periyagramam.

In those days, it was the Periyagramam council that selected the trustees of the Kandasami temple, while some of their members separately controlled major ritual institutions that extended the influence of the Beeri Chettiar leaders beyond the Town. For example, one Beeri Chettiar headman was the suroodiriyamtaar^[10] of a major temple in Tiruvotriyur, just to the north of the Town, a status that gave him an inheritable right to the temple's office of head trustee. His descendant is the suroodiriyamtaar and head trustee today. Another headman was the hereditary trustee^[11] of a major temple in Tiruvanmiyuur, the Murugeeswarar temple, located on the way to Mahabalipuram to the south. Yet another controlled a temple near Poonamallee to the west, while the trustees of the madam also administered a shrine and shelter (mandabam ) in Tiruvaalangadu, not far from Poonamallee at the site where caste elders believe the original Dharma Siva Acharya Madam was located.^[12]

The reader should notice in this description the mandala pattern that begins to emerge, which extends the leadership of Beeri Chettiars well beyond George Town. In fact, according to one informant, himself a descendant from a headman family, when a headman family sponsored an auspicious event, such as a marriage, it was still the custom in the 1920s to prepare twelve offerings of betel nut and leaf (a traditional offering of hospitality, called a taambuulam ) to give to persons who attended as representatives from outlying locations where the Town Beeri Chettiars

maintained Important ritual rights. A taambuulam would then be given to the chief visitor from each locale for distribution at their home institutions, a kind of symbolic announcement to the community of the family's auspicious event.^[13]

Portraying the Center:
Muthialpet’s Kachaleeswarar Complex

Until the end of the nineteenth century the Kachaleeswarar temple was the central temple of the George Town Beeri Chettiars. Today, despite the popularity of the Kandasami temple and the superior wealth of their locale, the Park Town Beeri Chettiars still portray their former status as a satellite of Muthialpet during the Kandasami temple's Spring Festival when, on the morning of the eleventh day, they take the idol of the god on procession. The god, Kandasami, enacting the role of a living king, travels to the Kachaleeswarar temple to celebrate the renewal of life that the winter rains have brought and to float on a raft in the temple's tank. Today, unfortunately, the tank is dry because the Town's water table has fallen, so, as one informant said with a twinkle in his eye, the idol goes instead "for the stroll."

At the tank, Kandasami is joined by the idol of Kaligambal, the benevolent female deity who gives her name to another nearby temple on Thambu Chetti Street (see map 2), this one controlled by the Viswakarmas, or Smiths, who have carried her here this morning in procession. Like Kandasami, Kaligambal's idol has come to "float" on a raft in the Kachaleeswarar tank and to pay respect to the more powerful god-king, Kachaleeswarar. The two deities coming together like this expresses the friendly relationship that has historically existed between the Beeri Chettiar community and the Viswakarmas who live among them, but it also suggests that the trustees of these two temples were once the subordinate lieutenants of the trustees of the Kachaleeswarar temple.

In British times, older informants say, the Viswakarmas were poor compatriots of the Beeri Chettiars and were often employed by them. One informant told me that forty years ago it was common for young Smith girls to work as domestic servants in Beeri Chettiar households. In yet earlier times, when the temple was built in the eighteenth century, the two castes were leading members of the left-hand moiety of castes that lived east of Popham's Broadway and were allies in competition with the right-hand castes that lived to the west. This is the old division of George Town, which, as we have seen, has left its legacy in the local

organization of the Town and in legal rights in institutions. The Float Festival, a day-long celebration that occurs during the annual Spring Festival, continues to symbolize the old alliance between the two castes.

The Float Festival is also associated with two of the Town's most venerable legendary Beeri Chettiar big-men, Thambu Chetti and Lingi (sometimes Linga) Chetti. Two of Muthialpet's main streets today bear the names of these former leaders. In the eighteenth century, the Kachaleeswarar temple was sometimes called "Tambi Chetti's pagoda" (Love 1913, 2:541-2), and according to British East India Company records, the temple and its tank were built by Thambu Chetti in 1725, when he was one of the Company's chief merchants and the premier headman or chief of the Town Beeri Chettiars (Love 1913, 3:387-8, 391). He built the temple on land that had belonged to Lingi Chetti, a contemporary headman-merchant and in the mid-1700s the mint contractor of the British East India Company that ruled Madras at that time (Love 1913, 2:312-13n). Lingi Chetti was a major employer of Smiths in the mint and his relationship with the caste may well have been the source of the connection between the Beeri Chettiars and the Smiths that continues today.

One of the first stories Bala told me, a story others often repeated to me, concerns an attempt by the trustees of the Kandasami temple during Thambu Chetti's reign to carry on the "float" festival independently of the Kachaleeswarar temple. My informants' purpose in telling me the story was to describe the power and control that a preeminent big-man had in former times, power which enabled him to dictate relationships among leaders within the community polity. The story goes that the trustees of the Kandasami temple decided to hold the event at the Kandasami temple itself rather than go to the Kachaleeswarar temple. When Thambu Chetti heard of the plan, he immediately ordered the Town officer to instruct the Kandasami trustees to bring the deity to the Kachaleeswarar tank as tradition dictated. Hearing of Thambu Chetti's displeasure, the persons responsible for the day's festival rushed to him to apologize and beg forgiveness. Thambu Chetti chided them for their presumption and ordered them to renovate the Kachaleeswarar temple palanquin, used to carry the god in processions, as an expiation for their action.

This story, of course, also tells us that the Kandasami temple is seen by today's leaders as once having been a satellite of the Kachaleeswarar temple and that the Beeri Chettiars of Park Town were once under the jurisdiction of Muthialpet's leaders. My informants thought that

the story clearly illustrated the great importance of individual leaders, especially the premier big-man of the community, and his dictatorial power.

F. L. Conradi's map of "Madraspatnam" (see map 3) supports this view of the satellite status of the Park Town Beeri Chettiars. The map depicts the Kachaleeswarar temple in 1755 as centered in what we know from records was the heart of the Beeri Chettiar community. The temple is grand, surrounded on three sides by the tree-lined streets of a Brahman agraharam . There is no other temple in this section of the Town as large. Only the Ekambareeswarar temple in Pedda Naickenpet, the section of George Town then controlled by the right-hand castes, is comparable. There is a noticeable symmetry in the size of these two major temples, a balanced representation of each moiety's importance and power. By contrast, in 1755 the Kandasami temple is depicted as little more than a small shrine. Kachaleeswarar's size aside, what gives the temple its centrality in the caste's institutional galaxy is its agraharam and, located within the agraharam , the Beeri Chettiar's caste madam . The temple, the madam , and the agraharam together constituted the central institutions of civic leadership within the Muthialpet area until the end of the nineteenth century.

The agraharam once housed a Brahman priestly community, members of which were supported at least in part by Beeri Chettiar "charity," reflecting both the caste's spirituality and wealth.^[14] Endowing Brahmans, for example, by feasting them—sometimes the Beeri Chettiars fed a thousand at a time—was a traditional way of symbolically maintaining the relationship between the wealth and power of leading individuals and the caste community on the one hand and the cosmic order on the other. This public gifting of Brahmans and the large agraharam dramatized, therefore, not only the honorable reputation and eminence of the caste civic community and its leaders who endowed Brahmans, but also represented the civic community as an orthodox ethical order. At the center of this urban cosmos was the temple-palace housing the god-king, Kachaleeswarar. Next were the agraharam streets that encircled the temple. Finally, came the civic community itself, which was made part of the cosmic fabric by the processional streets of the temple that wove through the community.

As noted, the third institution of the Kachaleeswarar temple aggregation is the caste madam , the Abanatha Dharma Siva Acharya Madam, located within the agraharam . Until 1876, the madam housed

the celibate Brahman spiritual head of the caste, the Dharma Siva Acharyar, whose duty it was to instruct Beeri Chettiars on their moral codes. In the nineteenth century, the madam was also importantly the administrative center of the George Town Beeri Chettiars and the meeting place of their caste council, the Periyagramam, a council of between twenty-one and twenty-two headmen in the 1880s (ILR, 135). This council remained active until near the end of the nineteenth century. In the uses of the madam , the reader should notice again how the Beeri Chettiars harmonized and dramatized worldly power and wealth with ceremonials and a ritualized depiction of their position in the cosmic order.

Today, the trustees of the Kandasami temple still see themselves as part of the Muthialpet Beeri Chettiar community, but, even so, the George Town community acknowledges few leaders and has nothing like the cohesiveness that it once had. There are still descendants of head-man families living in George Town, and among them there are some who are influential public figures. But, as one informant explained, most today are merely ordinary men. Descendants of many of the old families have left business for other lines of work and, no longer so dependent on connections, have moved out of George Town. As they have, the residential composition of Park Town and Muthialpet has changed dramatically. "You do not even know who lives in the next house," one old-time resident told me.

In the late 1950s, quarreling among themselves, Beeri Chettiar leaders lost control of the Kachaleeswarar and the Mallegeeswarar temples to the state run Hindu Religious and Charitable Endowments Department, widely referred to by Tamils as the HRCE Department. In the 1960s, the Kandasami temple was also threatened with takeover by the HRCE Department. First in 1964 and again in 1969 the Department appointed managing boards to the temple. For a time the caste collectively controlled no important big-man institutions.

But in 1967 things began to change slowly. Several Park Town residents petitioned the HRCE Department to allow the caste again to elect their own temple trustees, and in 1976, several residents, including the head trustee of the caste madam and some of Bala's allies, filed a civil suit to regain the legal right to do so. As a consequence, pending a final decision by the Madras High Court, elections were again held in 1978, the first since 1957, but with the new proviso mentioned early in this chapter: In order to increase Beeri Chettiar involvement in the temple, the temple's charter ("scheme") was modified to extend voting rights to

Beeri Chettiars living or doing business anywhere in Madras City. While Muthialpet still had its core of old Beeri Chettiar families, as it does today, by this time most had relations living elsewhere in the city. One elder Beeri Chettiar estimated that today there remain only about one thousand members of his caste still living in Muthialpet. The decline of the headman system, changes in education and in economic opportunity, high real estate prices, family partitioning, and the desire for more open spaces had dispersed the once compact community.

"It used to be," my elder informant said, "that everyone knew each other, but now often your neighbors are strangers." He then gave me an example to illustrate the extent of the change: People used to lend their jewelry to the temple to be used to decorate the idols during festivals. Donors knew that they would get their jewels back. Now, because the community is so loosely knit, no one lends jewelry, and it is unlikely, my informant thought, that they would get it back if they did. The high level of trust that went with a community organized and constrained by personalized face-to-face relationships has disappeared. To a considerable extent the need to establish trust through enduring personal ties has been replaced by impersonal contractual relationships, law, and governmental bureaucracy.

Reflecting on the institutions that the Beeri Chettiars once controlled in Muthialpet, it is easy to imagine how prior to 1959 the processions of the Beeri Chettiar's three main temples crisscrossed and circumscribed Muthialpet and Park Town. Caste leaders used these temples, in conjunction with the madam , their Periyagramam council, and the agraharam , to define the caste as a community and locality-based political domain. Viewed by all who saw them, the temples and their processions were grand and opulent expressions of the Beeri Chettiar's power and wealth. They stated who the Beeri Chettiar's leaders were, that the caste was a moral congregation, and that Muthialpet and Park Town were their neighborhood domains. Of course members of other communities subordinate to the Beeri Chettiars also lived among them, including the Smiths (Viswakarmas), as we have seen. But the caste that historically had been the Beeri Chettiar's chief rival, the Komati Chettiars, lived with their allied castes west of Popham's Broadway, where they had their own temple domain.

Clearly the Beeri Chettiars saw themselves and were seen by members of other castes as the dominant caste of Park Town and of Muthialpet and Mannady. They controlled a more elaborate ritual structure in those neighborhoods, including denominational temples, than did any other

caste, and so they controlled the public symbols of status and prestige associated with these institutions.

To Royapuram and Back, the Singularity of Leadership

My fieldwork journal for January 30 reads: "Bala tells me that he is now trying to reestablish some of the old etiquette associated with festivals. His aim, he says, is to reinvigorate the sense of community that he feels has been so greatly weakened among the Beeri Chettiars, especially over the last forty years." Prior to the Spring Festival, he had a temple servant personally deliver announcement-calenders of the festival to the heads of some of the old leading families of the Town, showing them special respect and inviting them to be present and honored during the events. Taking the Kandasami idol on procession through the Town, Bala still seeks to weave a community constituency from its residents (fig. 9).

On the night of the fifteenth day of the Spring Festival in 1986, Bala, as he has done every year in his role as head trustee, plans to take the idol through the length of George Town in a procession that will last until dawn. The procession will begin in Kasimedu, a part of Royapuram, the area of Madras City that lies just north of George Town. Here the idol is said to have "rested" during the day at a site owned by the temple, Suriya Narayana Chetti Choultry, named after its donor. The Choultry was once a wooded garden on the outskirts of the city, and the idol of Kandasami has been brought here in procession the night before to enjoy the spring forest^[15] and to fetch his bride-to-be, the tribal princess Valli, whom he will marry as his second wife on the sixteenth night of the Spring ritual cycle.

Returning to the Kandasami temple from Kasimedu, a journey of several miles, the procession will proceed down Mannarsami Koil Street, where thousands of celebrators are gathering to watch and make offerings to the god as he passes. When it reaches George Town, the procession will then pass down the full length of Lingi Chetti Street on the way back to the Kandasami temple.

Just prior to the procession, the principal sponsors and participants sit on mats chatting with their guests in a temporary enclosure erected to house the god. Priests dress the god and decorate him with a dense garland of flowers. I want to photograph the process, but after some discussion the priests decide that the deity should not be photographed in

Figure 9.
Kandasami on procession.

an undressed state. When the dressing is complete, the deity is carried to the front of the enclosure on his palanquin and the procession forms up. Bala, as the head trustee of the Kandasami temple, stands with a dignified demeanor at the head of the procession. He beckons me to stand on his right, a position that shows me special respect. The betel nut merchant, who is the chief sponsor of this night's festival, will walk with his middle-aged son on Bala's left. In his role as head trustee of the temple, Bala's eminence and authority is expressed by his position at the head

of the procession, and it is in his power to distinguish his allies with symbolic marks of respect such as inviting me to stand by his side. Normally, the other members of the temple's board of trustees would also walk with Bala, but on this occasion they choose not to attend because of an ongoing dispute for control of the temple. They are waiting to see what happens.

Before the procession leaves the "garden" it pauses, and spectators rush to the god to make offerings of fruit, incense, coconuts, and money. Sitting on the palanquin at the feet of the idol, priests take the offerings from the worshipers and present the offerings to the god, then, keeping the small sums of money and a portion of the rest of each offering, they return to the worshiper what remains as prasadam .

Entering a broad Royapuram street, the procession pauses again. Traffic has been diverted so that the night-long journey back to the Kandasami temple can be made. The residents of this section of Royapuram are poor. Many are fisher folk, someone tells me. Tens of thousands have come out for the god's passing and they fill the streets. As we stand before the god, a young fisherman approaches Bala and, addressing him as if he were a lord, shows him his shoulder, explaining that it was broken when he was beaten by the police. He asks Bala if he could help him to have his shoulder mended so that he will be able to work again. Observing this, I turn to my research assistant, who is walking with me, and remark that people are treating Bala like a king. "He is king!" my assistant responds.^[16]

The god next enters George Town and passes down Lingi Chetti Street. The idol pauses at street corners and before the houses of prominent men who are singled out when the god stops before their houses and the priests perform aaraati to them, giving them a coconut and fruit offering. The honored individual makes an offering of money in return. In this manner, the procession stops at the house of the hereditary trustee of the Tiruvotriyur temple. It also stops at the house of the head trustee of the Kaligambal temple, a Viswakarma. Proceeding further down the street, Kandasami again pauses to give respect to the goddess at the Kaligambal temple and then continues on his journey back to his temple. The following night, under the sponsorship of the betel nut merchant and his son, the marriage to Valli is to be staged.

Beeri Chettiars tell me that this grand procession and those of other festivals have been repeated each year for as long as anyone can remember. Sometimes new processional routes are established, which extend the reach of the community into new areas of the city, as Bala is

now attempting to do. Other times, old processional routes are given up or lost when a section of the city loses its significance to the caste community. Each procession provides its sponsors and the temple trustees a chance to portray themselves as their community's eminent men. Without community, their roles and sponsorship would be meaningless. The converse is also true: without leaders and their institutions, there would be no community. Not only is civic individuality an important expression of who a person is, but also the very structure of George Town is an expression of civic individuality.

Chapter Four
Making the Community
George Town in Social History

One of the things that surprised me when I began my research on Madras was how little the local organization of George Town had changed since the start of the eighteenth century. Early maps, such as the one commissioned by Thomas Pitt around 1707-8 (not reproduced here), when he was governor in council at Madras, and that of F. L. Conradi, "Madraspatnam," drawn in 1755, reveal this simple fact quite clearly (see map 3). True, in its early days the Town area had far fewer buildings. In those days, much of the area was garden and open land, a sharp contrast to George Town's density and near complete lack of open spaces today. But already in these early maps the division of the Town into eastern and western moieties is one of its dominant features. In both maps, the Town's two sections are clearly divided by a canal with gardens on either side. In the Conradi map, a line, straight as a ruler, divides the western gardens from those of the eastern section. This can also be seen in J. Talboys Wheeler's sketch map of Madras in 1733 (map 7). Later, Popham's Broadway replaces the canal and gardens as the dividing line between the two sections (map 8).

The maps reveal that even in these early times, the local organization of the Town reflects the division of castes into right-hand and left-hand moieties. This is a distinction that is no longer made in Madras today, although it still affects the Town's layout and the distribution of caste residences. And there are older inhabitants still living who can remember disputes over temple rights founded in the division. As we have seen, the left-hand division to which the Beeri Chettiars belonged lived pre-

Map 7.
Madras in 1733 (based on Subrahmanya Aiyar's enlargement of J. Talboys
Wheeler's map).

Map 8.
Contemporary central Madras City.

dominately in the eastern section of the Town, to seaside, while the right-hand castes, headed by the Komati Chettiars, lived in the western section. Today, should a person stroll the streets of the two areas, the division is still apparent in the location of temples. In the eastern section of Town, temples are Saivite, because the left-hand castes were predominately Tamils, the great majority of whom are followers of Siva. By contrast, Vaishnavite temples predominate in the Town's old right-hand section because its residents were Telugu speakers, a group composed largely of worshipers of Vishnu. But the right-hand/left-hand division was not sectarian based. There are Telugu-speaking Beeri Chettiars who are followers of Vishnu, and while the majority of the Telugu-speaking Komatis consider themselves Vaishnavites, nonetheless Komatis claim a Saivite goddess for their caste deity, Kanyaka Parameeswari. Her temple, located in Kothawal Chavadi market is the counterpart of the Beeri Chettiar's Kandasami temple.

On the basis of these observations and on the evidence of the maps and the stroll through the Town, it might, nonetheless, appear that the forces determining George Town history and local organization are primarily those of caste and religion. After all, this is what the maps and caste-based residential areas seem to reveal. One would have good company making such an assumption; the supposition that caste and religion are, in fact, primary determinants in Indian social history has long dominated anthropological and historical discourse. Yet, when listening to Bala and his Beeri Chettiar friends explain who they are, it is apparent that they stress the actions and motivations of individuals, not caste and religion, to explain why things are the way they are. To listen to Bala, George Town is organized as it is because leaders have made it that way and because ordinary townspeople have been motivated by self-interest to assist in that construction. Bala's explanation and interpretations that stress caste and religion as the primary agents in Indian history are, therefore, fundamentally contradictory. Why do Bala's and his friends' views contradict those of Western scholars? Might not Bala's explanation of agency be a product of the late twentieth century, when individual responsibility is much more stressed in law and in social life generally than it appears to have been in the past (e.g., Béteille 1987)? In this chapter, I explore the social history of George Town and the role that civic individuality has played in its construction. My aim is to see what explanations persons living in different time periods gave as motivations for their behavior, and why, from their point of view, the Town was organized in the manner that it was. Looking back in time

helps us, as observers of present-day George Town, to understand what the interrelationships between religion, caste, and individual interest were and why the community of George Town today is framed in the manner that it is.

The Social Setting

What follows, then, is an analysis of the character and role of individuality among the George Town Beeri Chettiars and its role in the construction of the Town during three time periods: 1652-1708, 1717-1816, and 1876-1890.^[1]

At the outset it is necessary to call attention to the fact that Madras City was officially founded by the British when they gained the right to locate Fort St. George at the site for the purposes of trade in 1639 (Arasaratnam 1986:21). It is wrong, however, to assume that from its inception Madras was socially a British city, or that it was framed by a dominant Western-style mercantile economy (Washbrook 1988). On the contrary, "Black Town," the Indian section of the city, was Indian in its organization and reflected the social character of the surrounding region (Lewandowski 1985). Similarly, the organization of trade and production was also Indian. The dominance of Indian ways is hardly surprising, given the small number of English East India Company servants, about twenty-four "exclusive of Apprentices and Soldiers" in the second half of the 1600s, the weakness of British forces, and the Company's policy of depending on local merchants to conduct their trade and manage production (Wheeler [1861-2] 1990, 1:38-39). During the reign of Charles II (1660-1685), for example, when the number of English soldiers in Madras was increased, they still numbered less than two hundred. Rather than imposing Western ways and economy, Company agents tapped into existing systems (Arasaratnam 1986:3; Washbrook 1988:62; cf. Wink 1990:4) and adopted local Indian customs of public interaction, patron-age, and display.

For example, during his stay in Madras from 1699 to 1704 Thomas Salmon described how temple dancers

make up part of the equipage of a great man when he goes abroad; for every man of figure in the country, I observed, had a number of these singing women run before him; even the Governor of Fort St George was attended by fifty of them, as well as by the country musick when he went out. (quoted in Love 1913, 2:75)

The Company Council also made donations toward the maintenance of priests, supported a "water choultry" (a roofed public well and rest place for weary travelers) within the Town, and contributed to ten to twelve temples, including particularly the Company Church, the "Town Pagoda" (Chennai Kesava Perumaal temple), and the Triplicane Pagoda (the Paartasaarati temple) (Love 1913, 2:578). Public giving of this sort marked the preeminent leadership roles Company officers sought for themselves within the local Indian community (see Mines 1984:149). But in these activities the British mimicked the many prominent Indians, whose model for displaying eminence they were following.

As I demonstrate subsequently, Company-Indian interaction had a characteristically dialogic quality (Irschick 1989:490; 1994). In the early years, the Company's main effect on local society was that it offered new opportunities for trade, and Company representatives clearly fanned competition for access to these by favoring different Indian merchants. Consequently, from the beginning competition among merchants was intense.

From the first, Beeri Chettiar merchants were one of the main trading castes of Madras City. Their arch-competitors in the early years of Black Town were the Balijas and the Komati Chettiars, who were the dominant trading castes of the surrounding countryside. This competitive opposition reflected the organization of castes into rival right-hand and left-hand moieties in this region of post-Vijayanagar influence. The Komatis and Balijas were both Telugu-speaking, Vaishnavite castes be-longing to the right-hand caste division, while the Beeris were predominantly Tamil-speaking Saivites, belonging to the left-hand section.

1652-1708

During the early period of Madras history, the East India Company relied on prominent Indian traders, called "Company merchants," to contract their trade.^[2] When a merchant was awarded the Company investment, or "Prize," his control of contracts and cash advances meant that he stood at the head of an extensive intercaste network of weavers, artisans, service castes, and traders, who produced, processed, and ultimately delivered the goods listed in the Company muster to the Company ships. Consequently, when the Company Council in Madras awarded a contract to an Indian merchant, the contract meant not only that such men stood to become wealthy, but that appointment singled these persons out as men of power who could deny or reward access to

trade to anyone wishing to participate. This sanction of denial enabled them to act as arbiters of trust within their communities (Rudner 1989). Essentially, such men were brokers and agents, who through their dispersal of the investment, controlled the logistics of trade from the point of manufacture through to the boatmen, who took the goods to the ships anchored in the Madras Roads (cf. Pearson 1988).

It is evident, therefore, that the mercantile-manufacturing sector of Madras society at this period contradicts a Dumontian interpretation of India on two grounds: first, because the society is integrated by market trade rather than cooperative distribution, which Dumont contends characterizes the Indian economy and is a concomitant feature of India's caste-based division of labor,^[3] and second, because intercaste relations are organized by key individuals in pursuit of self-interests. Dumont offers market trade no causal force in Indian history, arguing that economics remain "undifferentiated within politics," which is itself subordinated to religion (Dumont 1970a:164-6). What he means by this is that it is religion that determines the order of Indian society, not politics or economics. Yet in south Indian society the importance of trade is attested by its antiquity (cf. Hall 1980; Gittinger 1982), the receptivity of local polities to European trade, and the ready response of traditional Indian merchants and manufacturers to the opportunities created by the East India Company. Essentially, the Company tapped into what was already a highly developed manufacturing and trading culture, which was, of course, what attracted them in the first place (Arasaratnam 1986:3; Chaudhuri 1985:92-3; Washbrook 1988; Wheeler [1861-2] 1990, vol. 1). Further, the Indian merchant headmen exhibited an individuality that indicates that they cannot be dismissed as mere representatives of collective interests and so lacking real social significance as unique persons.

The individuality of these merchants was especially marked by their preeminence as the heads of intercaste constituencies that they formed around themselves in their capacity as the brokers of trade. Expressing the headmen's preeminent roles, the British referred to them as the heads of caste, commonly meaning by "caste" either the right-hand or left-hand moiety, and held them responsible for the behavior of their followers. The headmen's ability to control behavior depended on their pivotal roles as the contractors and financiers of trade, roles that enabled them to demand compliance from allies in exchange for their patronage in trade. In addition, individuality Was marked by achieved reputation , which depended importantly on the personal ability and successes of a merchant. Finally, individuality was distinguished by au-

tonomy , stemming from the headman's achievement of preeminence (he was responsible for who he was) and reputation as a broker-financier (he was responsible for what he did). Headmen exercised this autonomy in their relations with the British as well as in their intercaste and intracaste relations.

As we shall see, religion played an important part in the organization of Black Town society, but religion was not the cause of organization, competition for control of trade was. Relations between the two moieties were determined by competition for trade, and their relative positions in Madras society directly reflected the successes and failures of merchant-headmen, especially in their relationships with the Company.

Reputation, Achievement, and Individuality

For its own part, the Madras Company Council maintained its power over merchants by awarding contracts on an annual basis. The continuity of a merchant's power, therefore, depended on his ability to maintain good relations with members of the Council and in particular with the Council governor. By the early eighteenth century, such relations depended upon the merchant's ability to finance his own trade and also upon his entrepreneurial skills and reputation for being able to meet the Company muster. If a merchant was publicly accorded honors ("tasheerifes," from Arabic, tashriif , "honoring," conferring honor^[4] ) by the Council, and especially by the governor, then his reputation among his production and trade network remained strong. His ties with the Council were confirmed. Conversely, if in public a merchant failed to receive respect from the members of Council, it meant that the favor of the Council had shifted to his competitors. Thus, the standing of Indian merchants within their own communities depended on public knowledge of their individual relationships with the governor and his Council. Respect and ritualized honors singled out individuals and were the barometers of power that signified that their recipients had the connections that made for profitable trade.^[5]

Competition among leading merchants was fierce, and the propensity of the British on occasion to favor the lower-ranked Beeri merchants over right-hand Komati and Balija merchants led to heated disputes. The riots of 1652 involved all of these elements. The governor in Council, Aaron Baker, was on the beach embarking on a leave from Madras when he made public his preference for Beeri merchants by refusing the gift of fruit right-hand merchants offered him. Instead, Baker honored the

"principalls of the left hand," Beeri merchants, with "tasheerifes," and "after them the Bramanees also, and hee comforted said principalls of the left hand with the Bramanees, that they should have patience two yeares, in which tyme hee would take course to Content them" (Love 1913, 1:123). A Balija, who was agitated by Baker's favoritism toward the Beeris, called a Beeri to his face and in the presence of the English, a man not worth a "cash," and tit for tat, the Beeri labeled the Balija worth less than two cash. Incensed, the Balija went to Town and raised members of the right-hand to attack and plunder the left-hand community, cutting off the ears of two men in the process (Love 1913, 1:120). The Company had only twenty-six English soldiers at the time, and they were unable to prevent the fighting that ensued. Both sides brought in men to carry on their fight.

The aim of the Balijas was to drive the "Chettis" (the Beeri Chettiars) from Town (Love 1913, 1:122). Significantly, this was a trade dispute, although ostensibly the dispute was over "'who should be the more honnourable Cast and have Presidency of the other'" (Love 1913, 1:120). If Baker's public preference for the Beeri merchants were allowed to stand, it would have signaled that they were also to be given precedence in trade. Baker intimated as much at his departure (Love 1913, 1:123). The Balijas struck, therefore, with the aim of undoing all this.

In the end, as had also occurred in the 1640s, the dispute was settled by dividing the streets of the Town, assigning those in the east to the left-hand castes and those in the west to the right-hand. The purpose was to separate the moieties to prevent riots from occurring during ritual occasions, when the leaders of the rival communities claimed ritual honors that expressed their conflicting claims to precedence and public reputation, claims that would single them out, individualizing them as leaders (Dirks 1987; Presler 1987; Mines and Gourishankar 1990). Territorial separation was a standard Indian approach to settling such disputes (cf. Buchanan 1807, 2:268; Appadurai 1974; Mines 1984:45) and indicates that ritual precedence, far from encompassing politics and economics, was limited by these. Separation, however, was a temporary solution because the Company Council created an untenable situation when it contracted exclusively through the merchants of one alliance. Riots were to occur again when the right-hand Komatis were once more excluded from trade.

The next great dispute occurred in 1707.^[6] According to the right-hand Komati perpetrators of the riots, the point at issue was again one of ritual precedence. But the actual cause of this great conflict was the Com-

pany Council's decision to break off ties with "Sunca Rama and Iapa Chitty," head merchants of the dominant right-hand Komatis, and to award the year's "Prize" to two left-hand Beeri merchant-headmen, "Calloway and Vinkattee Chitty," the "best Chapmen [merchants]" (P.C., July 31, 1707). The Council's reasons were that the finances of the Komati men looked shaky, while Calloway and Vinkattee Chitty had the "money at command to carry on the buisiness [sic ]" (P.C., Aug. 9, 1707:92-3).

Discounting Komati complaints that the riots were caused by Beeri infringements on their rights to precedence, the Beeri merchant-headmen explained the reasons for the riots to the Company Council as follows:

The design of the Right hand Cast was no other thn to impede the dispatch of the Companys Ship for England Expecting there was not a sufficiency of Goods in the Godown [warehouse] to Load on thm, which they thought would occasion the Company displeasure against the Governor and councill and utter ruin to the Left hand Cast, as soon as it could come to their [the Company's] knowledge, and that was their [the right-hand caste leaders'] sole aim. (P.C., Nov. 6, 1707:152)

The importance of the Beeri Chetti contention is that it reveals that the Komatis are contesting ritual precedence as a political strategy, revealing that ritual hierarchy is not the cause of Komati-Beeri Chettiar intercaste relations, as a Dumontian analysis would lead one to expect, but that competition for control of trade is. Thus, ritual practice is made part of the politics of trade.

Dirks (1987:261) has criticized Dumont's contention that in India religion has always been a separate and superior force to politics. By contrast, he argues that religion was an integral part of political authority in the ancient regimes of India but became separated from politics during colonial times, when the authority of caste headmen, which had been invested by local kings, declined in the face of British rule. Under colonial rule, temple honors continued to signify local prestige, but the recipients no longer had any political authority. The result: religion was separated from politics. The disputes of 1652 and 1707 support Dirks's assessment of the political uses of religion and indicate that the East India Company was still too weak to force the separation at that time. In fact, as will become evident, one wonders whether at the local level this separation ever occurred fully.

It took the highly vexed British a year to settle the 1707 dispute. For the first several months riots disrupted the peace. At one time the leading Komatis led a withdrawal of right-hand castes from Madras City to

San Thome (P.C., Aug. 20, 1707), while members of the left-hand castes locked themselves in their houses (P.C., Aug. 22, 1707). Life was made extremely difficult for both the British and Indians of the city. Water carriers no longer brought water to Fort St. George. Parayah laborers refused to work, boatmen refused to carry passengers and cargo to and from the ships anchored in the Madras Roads, fishermen refused to fish, and artisans engaged in textile production refused to process cloth (P.C., Aug. 27, 1707). To top it off, the Council determined that one of their number, William Frazer, was in league with the rebel Komatis, trying to engineer his own trading advantage to the detriment of Governor in Council Thomas Pitt (P.C., Dec. 2-6, 1707; Wheeler [1861-2] 1990, 2:41-61).

The manner in which the right-hand traders had achieved all this mischief was straightforward enough. They paid or promised to pay the providers of services, the handicraftsmen, Parayah workers, boatmen, washermen and women (bleachers and preparers of cloth), and fisher-men, for the pay they would lose while they were on strike (cf. P.C., Aug. 27, 1707). In other words, society was organized by the market and self-interest—not by cooperative distribution, as Dumont would contend.

The response of the Company to the disputes was multiplex. First, it threatened to use force: "[T]he gov' propposd the attacking of St Thoma [San Thome] the 26[th] at break of day ... and [there to] put as many of them ['our Parriars'] as posible to th Sword, but no Inhabitants of any Cast else ..." (P.C., Sept. 24, 1707:128-9). But this "allarmd Severall" inhabitants of Madras, especially the Armenians and Moors, who feared the Company would not succeed and the "Parriars would flye" into the country (P.C., Sept. 25, 1707:129). Likewise, the "Pedde Naique," the big-man head of the western section of Town, pressed the Company Council not to attack, arguing that it should use milder means to achieve a settlement (P.C., Sept. 25, 1707). He said he knew the governor of San Thome and intimated that he would try to use his influence with him to persuade the "Parriars" to return to Madras. Notice in all these ploys the importance of connections and of key individuals in the organization of early Madras society.

Second, the Company sought to gain control over the loading and unloading of ships by contracting to build twelve of their own shore boats, and by recruiting boatmen from outside the Town, who were independent of the right-hand headmen. Third, they offered to finance the removal of service caste headmen, who had cooperated with the right-hand merchant-leaders, in order to make room for new headmen who

promised to show first loyalty toward themselves (P.C., Oct. 20, 1707). Next, they attempted to counter Komati tactics by offering Company support to service castes in substitution for the support the Komatis had provided or promised, guaranteeing employment and protection, in return for the promise of loyalty and a return to work (e.g., P.C., Dec. 2, 1707). Further, in order to control the Kaikkoolar weavers and the "Oylemen," who had been switching allegiances back and forth, causing trouble to both the right-hand and the left-hand castes, the Council required that these castes declare themselves for the left-hand and the right-hand respectively (Love 1913, 2:29). Here we see the British attempting to place lower-ranked castes under the headmen of the leading castes of the right-hand and left-hand moieties so that they could be controlled.

When, in January of 1708, the Council learned that Sunca Rama, one of the displaced Komati merchant-headmen, with the help of the Indian governor of nearby San Thome had coined money ("coin'd a parall of fanams and cash") to finance his efforts, they banned the circulation of his money in Madras, under threat of severe punishment (P.C., Jan. 15, 1707/8). Obviously, the British thought that the loyalty of the right-hand intercaste alliance was financed by the dispossessed Komati merchant-headmen. Much of the British effort to end the conflict, therefore, was directed at countering the financial power of the right-hand leaders. In effect, many Company efforts were recognizably counterbids designed to buy loyalty for the Company.

But the most effective policy of the Council was its use of the pivotal role of these leading individuals. The Company held the leading caste headmen of the right-hand directly responsible for the uprisings, citing them as the "Chief Instruments in raising the ... rebelion" (P.C., Aug. 27, 1707) and repeatedly arrested the headmen of both sections to force them to make a settlement (e.g., P.C., Aug. 25, 1707). Sunca Rama was prominent among the rebels of the right-hand, while Calloway and Vinkattee Chitty were two of the four headmen of the left-hand. In effect, while struggling to control these headmen, the Company was also attempting to rule indirectly through them—without great success, it might be added.

In the end, the headmen of the moieties themselves reached a settlement involving the redivision of Black Town into right-hand and left-hand sections. After 1652, the old residential division had gradually been forgotten and members of both moieties had come to live in the streets of the other. Now the streets of the moieties were redefined and the ownership of upwards of five hundred houses were transferred. As in 1652,

the purpose of dividing Black Town was to confine ritual activities, which expressed the ritual precedence of headmen, to the localities of each moiety. In essence the purpose of separation was to limit the significance of ritual precedence by dividing it equally between two separate but rival headman territories, reflecting competing domains of merchant-headman authority. It is this division that frames the local organization of the Town area to this day. Its continued importance attests to the renewal of the separation down to the twentieth century.

Like the earlier riots of 1652, this dispute and settlement reveal several features about the organization of society at this time. It reveals, as we have seen, that the headman-merchants had great social authority, and that the British relied on headman-brokers to manage their trade and to regulate society. The Company participated in establishing authority by awarding contracts to headmen, which in effect defined the strength of their political and economic influence. In turn, headmen created the local organization of Black Town, and they divided it into two domains as a compromise solution that allowed them to preserve their preeminence and so their reputations among their separate constituents. The separate domains of the moieties, therefore, reflect the merchant-headman basis of organization. It was not religious hierarchy, then, but competition and the broker system of trade and finance that were the causes of local organization and caste relationships. The acquisition of wealth, reputation, and power were key motives behind organization. Finally, the division of the Town into domains of precedence indicates that local organization centered around the pivotal role of headman-brokers, rather than around a single encompassing ideology of hierarchy. This is apparent because separation offers no solution to contested hierarchical caste rank. However it does offer a solution to headmen threatening each other's preeminence among their constituents because it distinguishes constituencies as distinct territorial domains. Under conditions of separation, conflicts would only arise in situations where headmen either usurped the precedence of their rivals within their domains, or where they attempted to penetrate their rivals' domains without acknowledging their preeminence (Mines and Gourishankar 1990). Indeed, as we shall see, these are precisely the kinds of causes of conflicts that characterize the Town through most of the eighteenth century.

To summarize, when the role of headmen is understood in the organization of society, the causes of disputes and the local organization of the Town become apparent. Disputes were generated by headmen competing to control production and trade for their own interests and re-

flected the pervasive influence of the market at all levels of society. Town organization was created by the headmen themselves, in order to perpetuate their own roles and reputations as key individuals among traders and producers. The headmen resolved the dispute of 1707 by demarcating exclusive constituencies for themselves, dividing the Town into distinct territorial units. Their aim was to give themselves permanent social arenas within which they would stand as the central figures. These territories formed their constituency domains, and it was within them that their individuality as the leaders of their communities was now to be defined.

1717-1816

Conflicts led by the headmen of the two moieties did not stop with the resolution of the 1707-8 dispute. Riots (Love 1913, 3:385; P.C., 391A, Mar. 6, 1812) or the threat of riots occurred again in 1716-17, 1728, 1750, 1752, 1753, 1771, 1789, 1790, 1795, and 1809, and then, throughout the nineteenth century and into the twentieth century with less frequency. But from around 1716 the scope of disputes in Black Town changed. Altercations now arose for a variety of more limited reasons than those that had led to the all-out "wars" of earlier years. They occurred when members of one moiety threatened to enter the domain of the other with their ritual processions (e.g., the disputes of 1728, 1752-53), when the caste members of one moiety usurped the symbols or rights of a caste of the other (e.g., the caste flag disputes of 1771-90), or when members of one moiety created innovative ritual displays that were seen by their opponents as overreaching their own markers of preeminence (e.g., the dispute of 1790, occurring when the Viswakarmas created an innovative "brass disc" temple cart, and in 1809, when they introduced the grand pretension of using five calasams [pots] on their funeral biers). The instigators of all these disputes were the headmen whose reputations were threatened.^[7]

By eighteenth century's end, the role of the Company in local society had also dramatically changed. As a result of the Company victory in the Carnatic Wars, most of south India now came under its rule. In the early years of the nineteenth century, the Company consolidated its administration and developed bureaucratic principles that directly affected the role of headmen in society. When the riots of 1809 disrupted the peace of Madras City, the Company resolved no longer to mediate such disputes but to treat breaches of the peace as matters for the police (P.C.,

391A, Mar. 6, 1812:1444). This is the historical point at which the police, as a bureaucratic arm of Company policy, effectively disrupted the power of headmen to manipulate public peace for their own political ends in the name of caste tradition. Nonetheless, headmen continued to play important roles as brokers and arbiters of social trust among their own constituents.

Reflecting the importance of caste headmen, the eighteenth century was a period when a variety of leaders and their caste constituents built new temples in Black Town (Lewandowski 1985). Both Appadurai (1981) and Dirks (1987) attribute this proliferation to the displacement of Indian kingly political authority by the Company. They reason that ritual honors received during temple events are important public expressions of the relative status of local leaders. Therefore, as an extension of their own authority, kings would have prevented the proliferation of temples controlled by different castes in order to avoid confusing the ladder of local authority.

It is possible, however, that Appadurai and Dirks have misread the reason for the proliferation of temples in Madras when they take it to indicate the fragmentation of governing authority. True, temple honors did single out individual merchants and establish their relative dominance within a temple community. But the goal of merchants who controlled honors was not political rule, it was to use temples to establish their own public reputations and influence in order to facilitate their en-terprises (Haynes 1987; Rudner 1989; Mines and Gourishankar 1990). The division of the Town into moieties and of the moieties into caste locales, each with its own set of temples, created limited constituency do-mains, as we have seen. This enabled a different set of prominent individuals to lay claim to temple honors in each domain. Opportunities for establishing reputations by building temples and public works, sponsoring festivals, and administering community occasions were consequently multiplied. Indeed, the multiplication of temples in urban settings is characteristic of south Indian commercial towns prior to the coming of the British, and the division of towns into right-hand and left-hand sections is also well known.

In Madras, therefore, the proliferation of temples represented several things: the replacement of what had been a small number of Company merchants by a much larger number of men of diminished prominence, the narrowing of social domains so that they corresponded more closely to caste domains (Bayly 1984), and the decreased reach of headman constituency networks. Therefore, what the proliferation of caste temples in

Black Town indicated was less that central political authority was diminished, than that the number of merchants and prominent individuals seeking to establish public reputations had increased.

The Company Merchant system was abolished in 1771. Its function was divided among several types of prominent men. Primary among these were merchant-headmen, who organized around them constituents drawn from their own trading castes; "dubashes," who were powerful personal agents usually drawn from non-trading castes and employed by British individuals and enterprises to mediate interaction with Indians (Neild-Basu 1984), and "shroffs" (sarrafs ), the cashiers of banks and companies, and the financiers of Indian and often British enterprise. Shroffs and dubashes were influential both among Indians and the British. For Indian entrepreneurs, they were the points of access to business with the English, while for the British, they brokered and guaranteed the Indian merchants and businesses with whom English companies contracted.

Should an Indian merchant fail to fulfill his obligations, the Indian dubashes and shroffs could and did deny further contracts. In this manner, these key individuals monitored the reputations and achievements of individual merchants and enforced public trust. For each of these types of pivotal figures, maintaining good public relations with the English was an important determinant of his influence. Nonetheless, since dubashes (Irschick 1989:477-8) and shroffs served competing interests, the British considered their loyalty highly problematic and used a variety of techniques designed to encourage them to make British interests their first priority, including requiring sizeable bonds, holding out to would-be miscreants the threat of dismissal for life, and guaranteeing to those who loyally served them the employment of a son upon retirement. The link-age, therefore, between individual identity and the trust or predictability upon which business exchanges necessarily depended is obvious, as is also the fact that social relations depended on individual reputations and achievements. Among merchants, dubashes, and shroffs individuality was expressed in the importance of who one was and what one did and was a determinant of social relations both within the British and Indian communities.

Both the increased control of the British over Indian society and the displacement of Company merchants narrowed the scope of authority and influence of merchant-headmen. Headmen were no longer responsible for maintaining law and order within their moieties, but as merchant-headmen they retained the power of patronage. For example,

if a caste member sought employment or a merchant sought a loan, then the word of one of these men was sufficient to establish his trustworthiness or credit or the lack thereof. Consequently, in their roles as broker-guarantors headmen continued to establish among their constituents the conditions of social trust needed to conduct business.

In sum, during the eighteenth and early nineteenth centuries headman-broker leadership continued to be important, and authority, though continuing to be equated with ritual status within caste communities, was closely linked to the quality of relationships a headman established with the British. If a headman were able to claim the respect of the British, his ability to do business was enhanced; he was in a position to claim apical ritual status within his own caste community, and his influence consequently earned him immense authority and power among his caste fellows. The actions of Collah Singana Chetty, the great Komati headman of the early nineteenth century, are illustrative.

Collah Singana Chetty (alias Rownappa Chetty) was the brother of Collah Moothoorama (d. 1804), the founder and head trustee of the Sri Kanyaka Parameeswari temple and Kothawal Chavadi, Madras City's wholesale vegetable market. On 22 February 1816, Collah Singana made a permanent loan of Rs. 40,000 to the Company Government for the maintenance of four choultries (rest houses) in return for a suitable show of appreciation—that is, an expression of public honor—from the Company. The Company Board of Revenue agreed to Collah Singana's wishes, reasoning that accepting the gift might encourage others to make similar "benefactions" and that the offer publicized the honorable reputation of the Company among locals. Notice the dialogic nature of Indian-British interaction. Both needed the other. The Company needed the recognition of big-men like Collah Singana to establish its reputation, and Collah Singana needed the Company to establish his eminence and prestige. The symbolic forms for initiating and expressing this public recognition were, given the context, necessarily Indian: Collah Singana's generous gift and the governor in council's public giving of honors.

In appreciation of Collah Singana's gift, the Board of Revenue recommended that he be given "some approbation of the Right Hon'ble the Governor in Council" (Madras District Records). And on 13 September 1816, the Company publicly honored him with gifts of a gold medal and chain, a palanquin, and a "cowle" (from Arabic, kaul ; an Anglo-Indian legal term meaning a "lease or grant in writing" [Hobson-Jobson 1986]), the latter framed and still preserved by his descendants in Kanchipuram (field notes, 23 Mar. 1986). Later, he was made the suroodiriyamtaar of

Enadur; that is, he was given the right of management and maintenance of temple land in Enadur. Honor received and a strong relationship established, the Company seems subsequently to have facilitated Collah Singana's purchases of land in Madras, while among his own community he established himself over a rival faction as head trustee of the caste temple (field notes; Madras District Records).

The continued importance of the headman model of preeminence also remains evident in the behavior of some English Company officers at about this time. In other words, rules for establishing identity and individuality as a leader are still Indian. Thus, during the 1790s, Lionel Place, the Company collector of the district surrounding Madras City, commanded great pomp at major festivals in nearby Kanchipuram. Adopting

the role of an indigenous king ... Place would call for "all the dancing girls, musicians ... elephants and horses" attached to the temple at Conjeevaram, a temple city in Chingleput District. "Attending in person, his habit was to distribute clothes to the dancing girls, suitable offerings to officiating Brahmins, and a lace garment of considerable value to the god." He also used Company troops and sanctioned the prayers of Brahmans "to propitiate the deity for a good harvest or for good trade." (Irschick 1986:12)

Place's purpose in acting the role of a headman was to establish his public reputation as the preeminent figure of the district that he administered as collector. He sought a public identity both as commander and patron of the people.

1876-1890

In the mid-1870s the headman-centered organization of Beeri Chettiar caste leadership came under attack from within. This occurred when some of these men, whose personal interests were no longer served by the merchant-headman system of organization, were outcaste for breach of caste codes. In defense, the outcastes responded by setting out to destroy the dictatorial autonomy headmen enjoyed as the arbiters of public trust and conduct within the Beeri Chettiar caste.

Between 1876 and the first years of the twentieth century, these rebels successfully undercut the headman system of authority and administrative style. They did this by limiting the autonomy headmen enjoyed as administrators of caste affairs. The rebels used the civil court to circumscribe with law the power of headmen to enforce caste behavior codes with outcasting; they made offices in caste temples and institutions contestable by election, thereby weakening the hold caste headmen had

on the institutions of leadership, and public reputation. And they made headmen accountable under law for their management of institutional funds and assets—the community's resources—in effect substituting audits for reputation as a determinant of trustworthiness. In other words, the rebels forced a redefinition of leadership which curtailed it and made it conform to limits set by bureaucratic standards of administration. Here again we see individuals, motivated by their own self-interests—in this case the desire to relax caste conduct codes—initiating actions that led to the transformation not only of caste leadership, but also of the organization of the Beeri Chettiar caste.

In the 1870s the Town Beeri Chettiars were divided into "18 or 20" named endogamous subcastes, which informants today call gumbuhal (clusters), each headed by one and in some cases two headmen (ejamaanan ) (ILR, 150). According to Beeri Chettiars today, the member-ship of each gumbu was small, numbering a few hundred to a few thousand families. Gumbuhal were named after village-centered locations (e.g., Poonamallee gumbu , Illatuur gumbu ), after the headmen who founded them (e.g., Periyaswamy Chetty gumbu , Kasappa gumbu ), after their relative status (Koruvu Viidu [lowly house] gumbu ), or after a trade specialization (e.g., Tin-Sheet gumbu , Araca-Nut gumbu , Tobacco gumbu , Kasukaarar gumbu ). The gumbuhal were ranked by wealth, and the headmen of wealthy clusters, such as the Poonamallee gumbu , were the most prominent among the Beeris.

Numbering twenty-one or twenty-two in 1876, the headmen of the gumbuhal met to administer the affairs of the caste in the council called the Periyagramam (ILR, 135). These council meetings were held at the caste madam , the Abanatha Dharma Siva Acharya Madam. Headmen also conducted separate subdivisional meetings to decide the affairs of their own cluster constituents, but the Periyagramam meeting held the power to override divisional decisions. Gumbu headmen enforced caste codes for behavior but rarely expelled anyone from caste before the 1870s. They had no need to because, as the monitors of public behavior, headmen ensured that the cost of deviating from caste norms was high: loss of reputation. Therefore, incentives favored compliance with headmen in return for the benefits of a good reputation vouched for by one's headman. Without the word of headmen attesting to their reliability, individuals were at a definite disadvantage in economic and social ex-changes. For example, people would hesitate to employ such a person, give him credit, or arrange a marriage with him or his family members.

The Indian Law Reports (ILR) describe the clusters in 1886 as very much like guilds and the Periyagramam as like a federation of guilds. In the ILR, the Periyagramam is described as "concerned at least as much with enforcement of social observances and of conduct deemed to be for the good of the community—and possibly for regulation of its trade ...—as with the punishment of ceremonial offenses" (ILR, 150-51). The comments of the judge here reflected his awareness, gained from witnesses, that by monitoring public behavior headmen played an important role, ensuring a high level of public trust within the George Town Beeri Chettiar community.

Prior to 1876, only two cases of outcasting among the Beeri Chettiars of the Town were known, one in 1864 involving the son of Mylapore Kandasami Chetti, who was expelled when he became a Christian, and one in 1874-5 involving Pakam Sabapathi Chetti, expelled for an illicit sexual relationship with a low-caste woman (ILR, 138). The absence of prior outcasting suggests that the customary sanctions wielded by headmen were sufficient to punish deviants without resorting to expulsion.

By 1876, these circumstances had changed and the first of a series of outcastings occurred. In that year Ratnavelu Chetti, a member of the Madras Civil Service, was outcaste upon his return to Madras for having travelled to England. Overseas travel was still an outcasting offense at this time. But opinion about the outcasting was divided. Ratnavelu's father, Ramasami, a headman of his cluster and a member of the Periyagramam, opposed the outcasting, causing a division within his cluster between those who supported readmission to caste and those favoring the outcasting. The split quickly spread to members of other clusters, and the headmen of the caste were soon divided among themselves.

The first casualty of the dispute was the caste guru. From court records and modern informants it can be surmised that proponents saw the out-casting as legitimated by the caste guru, who was employed in part to interpret the religious basis of caste codes (O.S., 12). Opponents, it seems, countered that the guru had no authority in the decision. Caught between the two factions, his position as a servant of the Periyagramam was untenable. He left the caste madam and site of Periyagramam meetings on pilgrimage. He never returned and the Beeri Chettiars of the Town were never to replace him, although they continued to maintain the madam building as a caste institution. Informants today explain why the guru position has been left vacant: they oppose a priestly definition of caste codes.

Next, in January 1877, Ramasami, Ratnavelu's father, was expelled from caste for associating with his son, then readmitted, and reexpelled in 1878 and 1881 respectively. In 1879, Subraya Chetti, the headman of the Periyagramam, considered twenty-eight men outcaste (ILR, 138-9). Suits and countersuits in the civil courts followed.

There are a group of court decisions in particular that unraveled the authority of the Beeri Chettiar headmen. The allies of Ramasami brought suits to gain control of the caste madam , which as the site of the Periyagramam and the dwelling of their spiritual teacher, represented both the spiritual identity of the George Town Beeri Chettiar caste and their corporate center. This identity was expressed in part by the caste codes for conduct sanctioned by the Periyagramam under the spiritual aegis and teaching of the caste guru. The aim of the suits was to force

an account being taken of properties constituting the endowment [of the madam ], for their being secured for the benefit of the institution, for a scheme of management being settled for plaintiff No. I [one of Ramasami's outcaste allies, Krishnasami Chetti] and other competent persons being appointed trustees, and for adequate provision being made for the due administration of the endowment [of the madam ]. (ILR, 135)

The plaintiffs' purpose was, of course, to control the administration of the madam . Control meant it would be they who were first honored during madam events, and they who scheduled Periyagramam meetings and defined who was invited. Imposing the madam management scheme, therefore, was designed to upstage the moral authority of the conservative headmen and to nullify their efforts to exclude the rebels from participating in caste assemblies.

In 1883, as an outgrowth of these suits, the High Court framed a new administrative scheme for the madam in which the rebels had a role (O.S., 4, 11-12). This was challenged by the conservatives, who argued that the rebels had no right to bring the suit because they were outcaste. But in 1886, the outcasting was itself invalidated by the court on the principle that the rebels had been absent from the Periyagramam meeting that had expelled them and so had been unable to face their accusers as was their civil right, under the maxim, aude alteram partem (ILR, 145). Western law here defined the civil rights of individuals set against the legal corporate right of a caste to expel members.

By wresting control of the madam from their opponents on the Periyagramam and eliminating the caste guru, the rebels were in the position to validate their own more liberal codes and to successfully challenge the prerogative of the conservative headmen to enforce caste codes of be-

havior. Reflecting on these events today, aged informants recall their elders' attitudes and put the matter bluntly: educated men were no longer willing to accept the reactionary authority of uneducated caste headmen (field notes).

These lawsuits mark the end of the headman system of caste organization among the Beeri Chettis in Madras City, but they do not mark the end of caste leadership nor the end of the role leaders as individuals play as a cause of caste organization. In chapter 6, I continue the story of these issues and the changing nature of individuality in Madras society.

In the last quarter of the nineteenth century, the rebels had achieved a rapid and fundamental shift in social organization from a headman-centered system of caste, well suited to a mercantile lifeway that benefited from a high level of social control, to a caste society organized by general bureaucratic principles of law, better suited to a more open, less personal society. For educated civil servants and professionals the head-man system of organization had become too restrictive.^[8]

The degree to which the caste organization of the Beeri Chettis has been that of headman sodalities is made clear by the fact that the demise of headmen has also meant to a considerable extent the end of the overall corporate organization of the caste within Black Town. This demise should not be surprising, since the institutions of corporation were the Periyagramam and the clusters, which were orchestrated by headmen. Today, caste members still have rights in the skeleton institutions through which headmen once implemented their leadership: the caste madam building, the one remaining caste temple—the Kandasami temple-the caste funeral rites tank, and the endowments associated with these. These legally constituted rights are the legacy of the bureaucratic management schemes the rebels and their associates implemented at the end of the nineteenth and the beginning of the twentieth centuries. But no one considers that either these institutions or their elected officers rep-resent the identity of the Beeri Chettiar caste as a community today.

Conclusions

Do individual uniqueness and achievement play a role in the history and organization of Indian society? The history of the Beeri Chettiar caste strongly indicates that they do. During each of the time periods considered, key individuals play a primary role in determining the organizational form that caste society takes. Further, recognizing the importance of individual achievement in Indian society allows the highly competi-

tive and pragmatic nature of much Indian interaction to find its rightful place in the portrayal of Indian public life. Thus, the role of achievement is readily apparent among the Beeri Chettiars when the highly competitive nature of intercaste and intracaste relationships is recognized as stemming from the competitions of merchant-headmen spurred by individualistic desires to achieve wealth, power, and prestige.

In the pursuit of these goals, headmen created the organization of Black Town society, and as circumstances changed, modified old organization and created new. This appears to have been true to such a degree that one might contend that among the Beeri Chettiars at each period of history caste organization was identical with the organization that headmen created in the course of their leadership. During the period 1652-1708, the moiety division of castes reflected the primary competition of individual headmen-brokers, each of whom stood at the apex of an intercaste production and trade network that he organized as a trader-financier. This headman-broker system of organization was set against a legacy of Vijayanagar rule that gave social precedence to its chiefly allies, the right-hand castes, and subordinated in social esteem the traders of the left-hand. During the period 1717-1816, headman-traders segmented the Town into constituency domains. This enabled the heads of the moieties to compete within their own territories for economic and social preeminence, while to a considerable degree avoiding the issue of relative precedence between the two main trading castes, the right-hand Komatis and the left-hand Beeri Chettiars. Finally, during the period 1876-1890, their interests no longer served by the headman system of maintaining social control and spurred by their expulsions from caste, leading men garnered allies and forced a restructuring of the Beeri Chettiar caste, curtailing the autonomous authority of headmen in the civil courts and imposing bureaucratic standards of administration that deflated their power. The history of the Beeri Chettiars, therefore, directly contradicts Dumont's characterization of Indian society as ignoring the individual. On the contrary, individuals are the agents of organization and their interests and achievements play clear roles in the transformation of Madras society.

How then are we to understand and characterize individuality in Indian civic society? Certainly throughout the history under consideration, Beeri Chetti headmen have striven for and have achieved a type of socially significant individuality based on preeminence. Equally clearly, this has been an individuality unsupported by the values of equality and liberty that Dumont stresses in his characterization of Western individual-

ism. What Beeri Chettiar history has illustrated is that individuality may take a different form than Western individualism, involving, nonetheless, a mix of traits, some of which Western notions of individuality also share, notably: (1) individualistic social identity that is recognized, evaluated, and considered instrumental within the context of groups, a contextualized individuality that is spatially defined; (2) an identity defined by eminence within groups, an individuality of inequality; (3) uniqueness marked by public reputation; (4) achieved identity associated with a de-liberate striving after one's own gain, dominance, and prestige; and (5) autonomy marked by responsibility for who one is and what one does. Because context defines the spatial dimension of a person's individuality, and it is within particular contexts that features of identity are known, as noted in chapter 1, I have proposed the label "contextualized individuality" to characterize this particularly Indian type of politically significant individuality.

Recognizing the essential role individuality plays as a cause of organization and as an integral feature of leadership greatly affects how Indian society is conceptualized. The societal order that individual pursuit of advantage begets, while it exists within the context of castes, expresses a distinctive form of traditional organization in which the paradigm of society is not caste hierarchy, nor conceptualizations of purity/impurity, nor even priestly versus kingly modes of behavior, but that of constituencies , which form around dominant individuals and their patron-age. It is these groupings that I have labeled "leader-centered" or "head-man associations." Throughout history and at every level of Indian society leaders have achieved public recognition as individuals. It is they who have created, managed, and transformed the institutions that give Indian society its corporate frame and Indian history its political and economic drama.

Chapter Five
A Portrait of Change

The 1870s and 1880s are too distant in the oral histories that George Town informants relate for me to flesh out the lives of Ramasami and his son Ratnavelu, the headman and his son, whose outcasting led to the nineteenth century lawsuits just described. I wonder about the specifics of their motivations and what their goals were. Were they merely concerned with reversing the Periyagramam's decisions to outcaste? Or was their vision more far-reaching? Did they personally feel that gumbu headmen and the caste council no longer served the interest of men like themselves?

I am fairly certain that they were motivated by both reasons. Their actions show that they wanted to nullify the decisions that had outcaste them. And clearly they found the rule of the headmen too restrictive. Circumstances had changed. Education and broadened opportunities had created a new social order and a need to travel and form alliances beyond the locality. I find evidence that corroborates these inferences in more recent experience. Very similar changes occurred in the lifetime of an elderly informant of mine, Natarajan,^[1] a member of the weaver caste (the Kaikkoolar), who began his adulthood in Kanchipuram, an urban hinterland of Madras City some fifty miles to the west. In towns located away from Madras, the kinds of changes Ramasami and his son experienced occurred many years later. I present Natarajan's story here as a cameo study of these kinds of changes.

Informants had told me that I should interview this man, an elderly lawyer practicing in Madras City. They told me that he was an important leader of the Kaikkoolars, who in his youth had been at the center

of a controversy that led to what my informants described as progressive changes in the organization of his caste.

Successful lawyers like Natarajan are busy men. I arranged to meet with him at his office in his home one night at seven o'clock. A bald, round-faced gentleman with wire-rimmed glasses met me at the door. Age slowed his movements and he seemed tired, but he was keenly interested in telling his story, and his animation during its telling was demonstration that he cared deeply about the principles he had spent his life living and advocating. Sometimes, when he related an episode in his life, he would fall silent for a moment, reflecting on what had happened. There was a sadness about him.

Describing his youth, Natarajan (b. 1907) told me how his future father-in-law, Murugeesan,^[2] an engineer and architect in Madras, approached his father to seek his, Natarajan's, hand in marriage for his daughter. The year was 1932, and at the time Natarajan was away from his family in Kanchipuram studying law in Madras. What made the proposal unusual, indeed revolutionary, was that his father-in-law's family belonged to a different locality-based endogamous subgroup. In 1932, locality endogamy was still the custom among the Kaikkoolars in Kanchipuram, as it was for the caste throughout the state.

Kaikkoolars adhered to the custom of locality endogamy not because they feared "excommunication" (outcasting), but because the custom corresponded with their own individual motivations to maintain their community connections and their family's good-standing within the locality. After all, it was on these that their successes and reputations depended. Nonetheless, since it was the caste headman's role to act as guarantor of his constituent's behavior, everyone also understood that if anyone did transgress the locality marriage rule, they could expect the headmen of the subcaste to outcaste the culprits. According to the locality rule, a Kaikkoolar who belonged to the Kanchipuram subcaste might move anywhere he wished, to Madras City, for example, but when he married or his descendants married, they were required by custom to make an alliance with someone belonging to the Kanchipuram subgroup. Murugeesan's marriage proposal, therefore, was a rebellious bid aimed at overthrowing one of the central rules of Kaikkoolar marriage custom.

Natarajan said that in the early 1930s, there were only thirty to forty Kaikkoolars statewide with university educations. One knowledgeable informant told me that as recently as 1927 there were only fifteen Kaikkoolars with bachelor degrees. Natarajan himself was one of the first degree-holders from Kanchipuram. Acting through a newly founded

statewide Kaikkoolar caste association and its publications, this small body of university-educated men strongly advocated discarding locality endogamy rules. Their motives for doing so were twofold. First, with university educations their occupational opportunities extended far beyond the locale, and they sought new extralocal connections that would help them to exploit these opportunities. Marriage offered them a way to start. In south India, a man's wife's relatives are supportive kin, and marriage ties create strong bonds upon which to build new relationships of trust. Second, and this is the reason informants stress, they wanted to arrange marriages with educated families, who shared their perspectives. They saw themselves as "progressives" and as "advanced" thinkers. But, because there were so few educated men in the state, as Natarajan said, it was necessary for them to "come out of their localities" in order to find educated families with whom to arrange marriages.

At about this time, 1932, the caste association held a statewide meeting in Kanchipuram. Natarajan's university education qualified him as a member of his caste's small progressive elite, and, eager to serve his caste, he was elected secretary of the association. He was also introduced for the first time to his future father-in-law. During that meeting, some association leaders renewed the call for "breaking out" of the locality restrictions that prohibited free intermarriage among Kaikkoolars. It was in this atmosphere of "advanced" ideas that, after meeting Natarajan, Murugeesan approached the young man's father and mother. He found them receptive to his marriage proposal. Natarajan's father was himself a progressive who wanted to arrange a marriage with an "educated and advanced family," knowing full well that this would mean breaking locality endogamy rules. Natarajan's mother and father also believed an alliance with one of the leading members of the Kaikkoolar community in Madras City would be to their son's professional advantage.

Natarajan related that it was his "desire in life" to become a lawyer. "I wanted to lead an independent life and to have scope for doing social services and joining public activities." The Great Depression was in full swing when he told his parents of his career wish. At first his father opposed the idea because, as Natarajan recalled, he "thought the lawyer's profession an unsteady one." In his mother, however, Natarajan had a strong advocate. She persuaded his father by threatening, if need be, to sell her jewelry to finance her son's law studies.

Natarajan must have been exaggerating slightly when he described his father as having been a poor man. At the time of the proposal, he ran a small jewelry shop in Kanchipuram, while earlier he had been a small-

scale silk dealer. Kanchipuram is famous throughout India for production of handloom silk saris. But certainly Natarajan's father was not wealthy. Nonetheless, by local standards he must have been reasonably successful because he was a respected member of his Kanchipuram community. An indication of this is that, prior to Natarajan's marriage, he had arranged the marriage of his only daughter to the brother of the caste headman. Thus, although his economic status may have been modest, he was well connected within his own caste locality. Accepting Murugeesan's proposal meant for him, therefore, putting in jeopardy his family's community standing and straining his ties with the headman, a man he could be sure would oppose the marriage. Perhaps he believed he could dissuade the headman from outcasting his son because his daughter was married to the headman's brother. Whatever the reason, Natarajan's father and mother's individual motivations for accepting the proposal had to have been strong.

Natarajan says he himself was personally amenable to the marriage but thought it would be necessary to put it off until he had finished school. He knew his father was too poor to support both Natarajan and a wife while Natarajan was studying in Madras. However, after being made aware of the problem, Murugeesan offered to provide board and lodging for the couple for as long as Natarajan needed to finish his studies and establish himself as a lawyer. Murugeesan's offer made the marriage possible.

Natarajan was the first Kaikkoolar from Kanchipuram to attend law school. When he graduated, he was to join one of a small handful of Kaikkoolar lawyers in the state. But because law in the 1930s was dominated by Brahmans and law practices required connections, Natarajan needed his father-in-law's support and social network to survive his early years as a lawyer. In the end, Natarajan says, his father-in-law supported him for seven or eight years with occasional help from his own father. During the interview, Natarajan spoke about his father-in-law and wife, both of whom were dead, with deep affection. It is apparent, therefore, that while Natarajan's and his father and mother's individual motivations for accepting the proposal were complex, ultimately they rested upon the fact that, given Natarajan's desire to be a lawyer, advantage lay more with aligning with an educated progressive Madras family than with a locality-constrained conservative Kanchipuram family.

Natarajan says his father-in-law's motives for proposing the marriage were of this same sort. He wanted an educated and "advanced" son-in-law who shared his progressive views. Education was what mattered,

not wealth. After meeting Natarajan at the Kanchipuram statewide caste assembly, Murugeesan apparently deemed Natarajan an ideal choice.

But there were problems. For all the liberal talk by the young caste association intellectuals supporting free inter-subgroup marriages, according to Natarajan, the reality in 1932 was that interlocality alliances were very rare and widely opposed. When his family indicated publicly their intention to go ahead with the marriage, as expected, his father was told by the Kanchipuram caste headmen that, if he did so, Natarajan would be outcaste.

In the end, however, Natarajan said, he was not outcaste. He was threatened, but his father's influence in Kanchipuram saved him. The marriage took place at his father-in-law's on Pycroff's Road in Madras. But, said Natarajan, "even close relatives refused to attend the wedding." Natarajan paused and became very still for a moment, thinking about his marriage. I could tell there was something in this boycott of relatives that still hurt him. When I asked him who these close relatives might have been, at first he hesitated, and I thought he might not want to tell me. But after a moment, he explained that the absence that mattered the most to him was his sister's. It was clear that all these years later, her absence still disturbed him. His sister's husband's brother was, of course, the reigning headman of the Kaikkoolars in Kanchipuram at the time.

Although tensions ran high for a time and the marriage divided the Kanchipuram Kaikkoolars into progressive and conservative camps, Natarajan's marriage, nonetheless, did constitute a successful rebellion. With a law degree, Natarajan's statewide connections were more important to him than those of his locality, and his father-in-law more helpful to him than the caste headman. Natarajan says his marriage spelled the end of the office of headman. His singular defiant marriage reflected a new reality, namely, that the influential educated members of the caste—the men who many thought were to be the future leaders of the caste—no longer supported the headman's authority. There was never another Kanchipuram Kaikkoolar headman.

Several important conclusions can be drawn from the story of Natarajan's marriage. For all the headman's seemingly dictatorial^[3] power, his authority proved surprisingly fragile. This was in part because his authority was based less upon sanctions than it was upon the incentives that motivated individual compliance with his bidding. Individuals complied as long as they saw it as in their interests to do so. This fragility was also in part because the headman's authority was a feature of community and depended upon the community locale continuing as the pri-

mary arena of social interaction. The headman, in other words, was a consensus leader whose arena of influence was limited to the locale. That consensus was abruptly withdrawn when Western-style education expanded opportunities beyond the town. Suddenly Kanchipuram was too small a pond for the fish that education had created.

Motivated by new possibilities for achieving wealth, prestige, and influence, individuals with the support of their families restructured the organization of their caste (subgroup locality endogamy was abandoned) within a very short time-frame, reinvented the arena of social interaction (the statewide caste community replaced the local community), and, completing the revolution, displaced the dictatorial headman. In fact, education, and the new occupational opportunities being opened by education, were creating a new elite within the caste. Natarajan and his father-in-law saw the new circumstances. Like other "progressives," they recognized that to succeed in the new arena, new kinds of alliances would be needed. In this new arena, they also hoped to "serve" their caste community, that is, to play the role of big-men. Their intent was to develop the statewide caste association as an umbrella institution that would engineer the transformation of locality groups into a statewide caste constituency. The resulting galactic polity would be composed of the local satellite caste associations that made up the membership of the statewide association. In fact, they and their fellow leaders were to prove remarkably successful (Baker 1976:199; Mines 1984).

Natarajan's 1932 election to the secretaryship of his caste's statewide association was his first step toward establishing his civic identity as an eminent leader within this new arena. At the same time, he was also designated the honorary editor of the association's magazine, the Senguntha Mittiran , a position he held until 1960, when he was elected vice president of the statewide association.

The influence and eminence of the civic individual expands when, as an officeholder, he or she extends control to other satellite institutions, each appealing to its own constituents whom the officeholder seeks to serve. In 1953, Natarajan became the secretary of a devotional society, the Sri Balasubramaniam Baktha Jana Sabha, located in the Royyapet section of Madras City and dedicated to the worship of Murugan (map 8). As I have noted earlier, participating in the management of a religious institution helps validate a leader's altruistic concern for constituents. Since Kaikkoolars consider Murugan their caste god, Natarajan used the society to further establish his singularity as a leader among Madras City Kaikkoolars, particularly those living in the south-central

part of the city where the society was located. In 1960, Natarajan was elected the society's vice president, and then, in 1972, its president. In 1970, he also became the president of the Siva Subramania Baktha Jana Sabha, another devotional society dedicated to promoting worship of Murugan, this one located in Pudupakkam, just to the northeast of Royyapet. From 1953, he was also president of the Pudupakkam Permanent Fund, Limited, a private lending and banking company. While this was a position for profit, the company may also be seen as a satellite institution in the galaxy of institutions upon which Natarajan built his role as a leading civic individual. Finally, in the 1970s, Natarajan was designated executor of the Annammal estate, a trust that ran a middle school of about two thousand students in Saidapet, a section of Madras City that lies to the south. Located in the south-central and southern parts of the city, these institutions and offices constituted Natarajan's institutional galaxy, the mini galactic polity within which he constituted his identity as a civic individual and as a lawyer. In 1976, when the statewide association met again in Kanchipuram, Natarajan was its associate president (tunai talaivar ) and a recognized and honored pioneering member of his statewide caste community. By that time the caste headman was all but forgotten. By contrast, there were numerous Kaikkoolar big-men in attendance at the meeting.

Chapter Six
The Decline of Community and the Roles of Big-men

I want to return to Bala and George Town in order to explore what has happened to the Beeri Chettiar community and civic individuality during the twentieth century. But in order to put George Town into perspective let us look for a moment at another section of Madras, which, after all, is a city of about five million. My purpose is to demonstrate how the forces that wed neighborhood and community, that motivate the union of big-men and their constituencies, have determined the organization and character of not just George Town, but of other parts of Madras City as well.

As we have seen in previous chapters, times change and what it is that people want and leaders can accomplish changes with them. Big-men and their clients respond by renegotiating their relationships and by redefining their roles. In the process, the character of the neighborhoods of the city change also. In the second part of the chapter, 1 return to George Town to describe its evolution since the nineteenth century lawsuits. There, today, Bala is trying to pull together a neighborhood that is coming apart. The forces of change that Bala confronts are broad-based and are impelled by shifts in the needs and motivations of ordinary people.

Looking back over the past one hundred years, I see three major changes in community leadership and in the integration of George Town, each of which is associated with changes in the meaning of civic individuality. First is the period of gumbuhal , headman, and the head-man council, the Periyagramam. Second is the period of a looser form

of leadership, a period of big-men, their institutions, and constituencies. Third is the present day. Big-men still exist, but there are fewer of them, and they are redefining their roles and relationships among a shrinking clientele. Neighborhood localities are losing their community integration. An evolution is occurring that affects the entire city, every neighborhood, old and new. Madras City has begun to change rapidly. It is entering a new period.

Today, a drive down any major street in Madras City reveals clear signs of a vigorously developing metropolis. Consider Nungambakam High Road, which cuts through one of the city's old residential districts: vehicular traffic is congested; diesel exhaust hangs heavily in the air. Lining the road are apartment buildings, a multistoried, "five-star" Western-style hotel called the Taj Coromandel, new businesses, restaurants, and shopping centers. Then, turning down Krishnamachari Road,^[1] a side street, the driver enters an older world. At first the driver passes old houses set in large garden compounds. A few of these are posted with signs indicating the businesses they now house. Traveling further, the driver suddenly finds himself in the center of what was once a village, the "village" of Nungambakam. In the small bazaar, street vendors sell vegetables and rolled sets of banana leaves to be used as dinner plates. A small village-style temple dedicated to the elephant god, Vinayakar, is located there (see fig. 10). A second structure not far away is a devotional society dedicated to the god Ram, and not far from that is a third small temple.

When I first visited the village with a Nungambakam friend, it was September, just prior to my beginning work in George Town. We had gone because it was a festival day, Vinayakar cathurthi (Full-moon Day). In the street near the Vinayakar temple, children clustered around a man selling small, brightly colored paper umbrellas, symbols of the god. Next to him a vendor was molding and selling clay copies of the god for people to take home for their private worship. Later they would drop the mud idol into their household wells or into the sea, the proper methods for disposing of these clay images. The central event of the temple festival occurred when the bronze temple idol, garlanded in flowers, was taken in procession, carried on his palanquin through the village streets.

The names of these streets recall the castes and the big-men who once lived here: Krishnamachari Road, a Brahman Iyer name; Kumarappa Mudali Street, the name of a high caste agriculturist; Mangadu Sami Aiyar Street, another Brahman Iyer name (Nungambakam is known as a Brahman Iyer village); and Rama Naicken Street, the name of the mere-

Figure 10.
The Vinayakar temple in Nungambakam, Madras City, 1992. The
photo shows the village-like quality of the neighborhood, which is located in the
heart of the city.

ber of yet another dominant agriculturalist caste. Here the houses are not set in large compounds; instead, their verandas line the streets so that their inhabitants can sit out in the evenings and talk with passers-by, much as in a village or small town. In inner-city villages like this, there may even be an agricultural plot or two. Called "gardens," these bits of agricultural land, now behind walls, are remnants of the time when the area really was a country village. Madras was once known as the

"Garden City" because of these plots, and in the 1960s when I first lived there, urban agricultural "gardens" were still fairly common. Now, land values and urban growth have eliminated most of them.

Should the driver continue his short journey through the center of the "village," he will leave these traces of a rural past. Suddenly, just as abruptly as he entered, he will pass a wall and emerge once again onto a major road, Nungambakam Village Road, and into the bustling metropolitan present. The village, after all, lies in the heart of Madras City.

What the driver has experienced on his short drive is evidence that Madras's urbanization has been a process of the city spreading and engulfing the villages and towns that once surrounded it. Today Madras City is an agglomeration of villages and towns, each of which retains characterizing features of its own. J. Talboys Wheeler's map of Madras as it was in 1733 (compare maps 7 and 8) depicts the villages and small towns that today lend their names to the municipal districts surrounding George Town at its core: Egmore (Elambore), Nungambakam (Nungambaucum), Purasawakam (Persiwaukam), Vepery (Wepery), Triplicane, San Thome. Unlike the new outlying residential suburbs of the city, where neighbors are strangers and proximity is a product of class, these urban "villages" preserve their small-scale atmosphere. Residents know one an-other—if not always intimately—and strangers are noticed when they enter. Caste, too, remains an important determinant of residence. The identities of these localities have yet to be lost to the homogenizing forces of urbanization. As in George Town, it is in the context of places like Nungambakam that civic individuality finds expression. It has always fascinated me that these urban villages still influence the atmosphere of the city and leave traces in people's lives.

The experience of a graduate student of mine, who comes from Nungambakam, provides one sense of this influence. Her family has lived in Madras for nine generations. Recently her father and his brothers sold the family's ancestral home in the old "village" center. My student's father, mother, and brother were the only ones living in the house at the time. The father's brothers, who jointly owned the house, had for some years lived elsewhere, in Bombay and Delhi. In old age and in a cooperative spirit when they sold the house, the brothers bad all agreed that the time to partition family holdings had come. Like many inner-city houses, it has now been converted into a business. When the house was sold, my student's father and mother rented a nearby house, owned separately by one of the brothers. There, today, they live modestly, a seemingly ordinary middle-class family, neither poor nor rich. They own a

scooter, but frugality dictates that they rarely use it. However, if, as observers, we change our perspective so that we see them, not as a single family, but instead as a household at the center of an extended circle of kin, our understanding of the status of the family changes dramatically. My graduate student's extended family is well known in Madras and claims within its circle several highly placed and influential members both in its past and at present. It is no ordinary family.

One summer my graduate student returned to Madras to marry. When she came back to the university in the fall, I asked her about her wedding. "How many people came?" I asked. "Oh," she replied, "it was strictly a family event." "How many attended?" "About a thousand," she said with a smile. Curious about the extent of her family ties and knowing that her family was widespread in India, I asked her further, "If you were to walk down the streets where your father and mother have lived, how many houses would you still pass in which relatives of yours reside?" After reflecting a moment, she said, "About seven." Later, I learned from her father that about 1,200 relatives and close friends had attended the wedding. When I mentioned this to another Tamil friend, he had replied matter-of-factly, "Yes, that's her family."

Among other things, what this story reveals is that kinship ties are still important in Madras and its inner-city village localities. Until relatively recently, locality and kinship were overlapping categories for all castes in Tamil Nadu. For many they still are. When personal connections are important to social success, extensive kinship networks are an outcome, and community localities such as Nungambakam and George Town tend to persist. It is within the contexts of kinship circles and community localities that civic individuality finds expression. Without such contexts, there would be no civic individuality, and without a need for knowing who people are, the ties that unite a community, including those of kinship, need not be strong. It behooves us, therefore, to understand the forces that have contributed to the maintenance of these contexts and to their change so that we can understand how the role and nature of civic individuality is changing in the twentieth century.

Four factors in particular reinforce the association of kinship with locality or contribute to the continuity of urban localities as communities, notably: (1) locality-based marriage rules, (2) the roles that personalized trust and connections (including kinship ties) continue to play as determinants of social and economic success, (3) the community-making roles that big-men play, and (4) the presence of "charitable" community institutions—"big-man institutions," such as temples, production cooperative

societies,^[2] schools, and scholarship funds—which are the institutional contexts in which local big-men define their roles as community leaders and shape their followings. Like kinship and trust, big-men and their institutions are defining elements of community and social locality. Because big-man institutions have geographic location and often a distinct and sometimes impressive architectural form (e.g., temples, marriage halls, schools), informants describe them as key features of the physical layout of their own community and as distinguishing features of other locales. When institutional management is "denominational," that is, when it is controlled by a single locality-delimited caste, the connection between institution, locale, leadership, and kinship is further emphasized, even when the clientele the institution serves is not caste exclusive, as few are. But keep in mind that all of these forces and connections are weakening in the face of changes that are diminishing the role that communities play in people's lives in Madras City today. When community ties weaken, the spatial context community provides for expressions of civic individuality loses its significance, and the role that civic individuality plays changes.

Consider the nature of the four forces strengthening the association of kinship and locality and of those countervailing forces that are weakening community. First, locality-based marriage rules: Even today within Tamil castes most marriages still occur within recognized locality-based endogamous subgroups. This is true despite the fact that locality rules are no longer enforced and inter-subgroup marriages are increasingly accepted, especially among elite families. Smartha Brahman (Iyer) elites in Madras, the Iyers of Nungambakam included, were among the first to discard subgroup marriage rules. However, working to counter trends toward interlocality marriages, many south Indians still prefer marriages that maintain kinship ties by uniting related individuals: Tamil culture specifies cross-cousins and sister's daughter/mother's brother marriages. In fact, the mother's brother's daughter is called "murai pennu " by her father's sister's son, a term that means the girl over whom the son has a right [of marriage]. Genealogies that I have collected reveal that in fact Tamils often do marry relations, if not always these specific ones. My own assistant married his elder sister's daughter in 1986, for example. In some communities, the majority of marriages are among relations (Mines 1972). When I seek motivations for these preferences from informants, south Indians have ready answers: they stress the desirability of arranging marriages with families with whom they share custom and outlook, the better to avoid incom-

patibility and conflict. However, even when great care is taken to ensure suitability, marriage is fraught with difficult adjustments. Families are naturally concerned about the welfare of their children, especially daughters, who will leave the family when they marry. Consequently, mothers and fathers feel more secure when their children marry into families that they know intimately, and they like their children to live close by so they are able to see them frequently. In part, therefore, these marriage patterns reflect matters of trust, a much cited concern of informants. It is easier to trust people you know and with whom you have continuing relations than those who lack these traits.

But as we have seen, the need to establish trust through personal knowledge is diminishing in India today, and reflecting this in south India, community ties are weakening and strict subgroup endogamy rules with them. The disbanding of subgroup endogamy rules is closely associated with the spread of Western-style education and modern occupational diversity that has led to occupational mobility. Family heads prefer to marry their children to spouses with similar educational back-grounds, to avoid matching a person with a progressive outlook, for example, to someone whose outlook is narrowly circumscribed. Opportunity and territory are closely linked, and education expands opportunity well beyond the locality. Members of educated families tell me that sub-group endogamy rules work against their alliance-making interests because they restrict the field of suitable families too severely. Consequently, as soon as families in a caste begin to educate their children, they begin to chafe at locality-based subgroup endogamy restrictions. If there are no other reasons for restricting marriages to their locality, elite families soon begin to arrange interlocality marriages, although the majority still arrange marriages within their subgroups. In some instances, full-blown intercaste marriages begin to occur as well, but the number of these marriages is small. Usually these are "love marriages," arranged by the partners concerned.

Since the 1960s, my data shows that the numbers of intercaste marriages have increased in Madras City; by 1986 every genealogy I collected in George Town included one or more. What is important is not the number of such marriages, which is still small, but the change in attitude toward them. Twenty years ago families cut off ties with inter-caste couples and were reluctant to discuss them. Today people ac-knowledge them as part of life, something with which kin have had to learn to live. But informants privately tell me they still dislike them. They say their caste community disapproves of them, and families become

quite upset when they occur. In George Town, the family priest, purohit , who conducts marriages for the Beeri Chettiars, told me that twenty years ago if he had conducted an intercaste marriage, he would have been boycotted by the caste. Intercaste marriage ceremonies were then small private events. Today he performs intercaste marriages at five-star hotels. However, despite such evidence of growing tolerance, continued reliance on personalized trust among business and artisan castes acts as a countervailing conservative force that works to preserve subgroup caste marriages. So do the personal preferences of individuals. In fact, I found among the Beeri Chettiars of George Town the vast majority of marriages still unite families that belong to the same gumbu subgroup.

Also reinforcing the association of kinship with locality are the roles that personalized trust and connections continue to play as determinants of social and economic success: locality integrity is maintained because castes whose members still rely on personalized social trust to regulate their economic relationships also tend to see advantages in maintaining close community ties. Business and artisan communities are of this sort, although there is considerable variation among them. Take the case of the merchant. Community membership facilitates business because it enables the merchant to achieve a reputation for trust and establish en-during social and economic ties. Within his community he may achieve a reputation for credit-worthiness, for example. This may enable him to enact verbal contracts among his fellows at will. "His word is money," is highest praise for the Tamil businessman because it attributes great integrity, and large sums are sometimes exchanged on the basis of such a word. A merchant's reputation in India is worth money. Further, knowing the members of the community, a merchant has a sense of whom to trust, mistrust, or be cautious about. Of course, any merchant will also tell you that you can never really tell about anyone. But, among people he knows, or knows indirectly through big-men, the merchant who maintains strong community ties has the advantage of greater information with which to assess risks compared to one who lacks such ties.

Another factor strengthening the association of kinship and locality is the community-making roles that big-men play. In particular, this includes the various ways in which big-men and personal relationships are put to use to support reputations, to determine social trust, and to organize the ways and means of accomplishing ends. These uses together create forces that contribute to the preservation of neighborhood localities and the urban village phenomenon. Communities are preserved because people preserve and use personal ties as means to achieve an ordered so-

cial life and other important social ends. Knowing this, whenever I set up household in India, I make an effort to obtain most of the household services I need through a big-man or woman. This helps ensure for me the reliability of the service providers, because they choose not to abuse their relationship with me lest they jeopardize their continuing relationship with the big-man/woman through whom I made my contact.

Of course, relations with big-men and women are two-way streets. Such leaders also make deals on behalf of clients who seek to negotiate their relationship with me. I remember once, while conducting research among Muslim merchants, wanting to do a survey of Muslim women in their houses. I checked with a number of the community's leading Muslim men before beginning, explained to them that I had hired a female Muslim social worker to administer my survey—purdah , the seclusion of women, precluded my personally conducting the survey—and began the survey when I thought I had the support needed. I had been conducting fieldwork in the town for over a year and thought I knew my way around. Two days into the survey, the religious head (moulvi ) of a small but conservative faction objected strenuously to the survey and demanded that it be stopped at once. I went back to the Muslim big-men with whom I had first consulted and asked their advice. In a most friendly manner, they explained to me that while they personally thought the idea of the survey was fine, after I had gone back to the U.S., they were going to have to live with the conservative members of their community. Perhaps it would be better, they suggested, if I gave up the survey. I did.

The fourth factor reinforcing the association between kinship and locality is the presence of charitable, big-man institutions. The big-man institutions that mark and delineate distinct communities reflect the important roles that big-men play in these social formations. However, when people rely on education and new occupations, such as government service, to achieve success, their need for big-man connections to establish their trustworthiness or to act as brokers on their behalf lessens. Bureaucracy, measured in terms of standards established by certification, law, and codified procedure (e.g., diplomas, the courts, elections), grows in importance as a mechanism for maintaining reliability and provides alternatives to the wrath of big-men for countering transgressions and restoring order. Consequently, people are less willing to submit to the personal dictates of big-men and there is less reason to maintain community membership. Concomitantly, reputations and community identity become less important, and individuality is muted. When in association with Western-style education caste members also

pursue new occupations monitored by bureaucratic controls, then the roles of big-men are challenged and diminished at the same time as sub-group endogamy rules are abandoned. There is even less motivation for individuals to maintain their community ties.

To understand the impact of the historic spread of bureaucracy and the decline of community, a distinction needs to be made here between "big-man," which I am using as a generic term, and "caste headman" (ejamaanan ). A caste headman is merely one type of big-man, the recognized head of a caste subgroup. As we have seen, many castes in Madras City and elsewhere once had, but no longer have, headmen, although members of these castes do continue to recognize men and women whom they consider big-men and big-women. A consequence of the growth of bureaucracy is that over the last one hundred years, communities have been rebelling against the dictatorial control that caste subgroup headmen once were able to wield. For castes such as the George Town Beeri Chettiars that had families who turned to education early, rebellions began in the nineteenth century. For others, rebellions began later. A few castes still have recognized headmen, although their power is consider-ably curtailed.^[3] Recall Natarajan: Since the power and authority of a headman depends on consensus, once that authority has been success-fully challenged—by even one individual—the end of the office of head-man is at hand. This is because without consensus, anyone can disregard the headman's commands. The way former headmen explain the disappearance of their role is that they are no longer given "respect," the respect that they need to command others. However, this does not mean that big-men no longer have a role in community life. On the contrary, if caste members continue to pursue occupations that rely upon high levels of personal trust, then subgroup endogamy may persist and, while the dictatorial headman will disappear from the community stage, big-men will continue to be influential arbiters of trust and manipulators of connections. As such, they continue to play similar roles to those that headmen once had, but without the dictatorial authority to command others to do their bidding. In Madras City, since the nineteenth-century lawsuits, individuals have sought to limit the personal autonomy and power of big-men, using bureaucracy to curtail their independence. They have done so by limiting the big-man's control over institutional offices by converting these into elected positions with defined terms of office. And they have circumscribed the big-man's independent control of institutional funds by implementing bureaucratic checks such as public audits and governmentally instituted and court-reviewed management schemes. Insti-

tutional big-men who used to have great entrepreneurial independence in the way they used their institutions have now been entangled in the web of law. As a result, the importance of their civic individuality has been diminished.

The Decline of the Beeri Chettiars’ George Town Community

In order to understand the role and nature of civic individuality in George Town today, it is necessary to understand how the community has changed. Although, as described in chapter 4, the nineteenth-century lawsuits mark the end of headman leadership among the George Town Beeri Chettiars, the caste community remained strong and gumbu subdivisions, which headmen had controlled, continued as an important feature of social identity and community organization. Beeri Chettiar informants say that it was not until the late 1950s and early 1960s, when inter-gumbu marriages started to be common in the Town, that gumbu distinctions finally and precipitously begin to lose their social importance. Among the Town Beeri Chettiars, therefore, transgressions of endogamous cluster distinctions were unusual until seventy years after rivals had successfully challenged the dictatorial powers of the Beeri Chettiar headmen in the High Court lawsuits of the 1880s. Today, while informants can still recall gumbu names, and a few elderly people still use the initial of their gumbu as part of their name, most people agree that memory of them is fading and that they are no longer socially very important. Why, then, did gumbuhal persist for so long after the demise of their headmen and of the Periyagramam council, their collegial body? And what finally brings gumbu endogamy to an end? The answers to these questions help explain the decline of the Muthialpet-Park Town Beeri Chettiar community at the end of the 1950s and the changing importance of individuality in the Town and city. In hindsight, the answers also reveal the historic relationship between caste organization and agency among the Beeri Chettiars. A caste is a mutable association, reflecting symbiotic interaction between leaders and members, whose organization reflects the purposes to which it is put.

One reason the Beeri Chettiar gumbuhal persisted is that the economic and social value of gumbu membership continued unabated in the George Town Beeri Chettiar community until the late 1950s, long after the disappearance of the Periyagramam council and headman offices. It is easy to understand why this was true. For one thing, the Muthialpet-Park

Town Beeri Chettiars were still predominantly a business community operating on connections and trust. Each headed by its own big-man, a gumbu was a small, tightly knit grouping composed largely of interrelated business families. The strong sense of gumbu identification and the need to maintain one's good name helped to create the high levels of public trust that elder informants still talk about. The nineteenth century lawsuits had left this linkage between family and business unaffected. Continuing gumbu -locality endogamy ensured, therefore, that relationships among families were enduring and that the trust and connections that facilitated the conduct of business were maintained.

The son of Natesa Chettiar (Natesa Chettiar [b. 1890 - d. 1964]; Kandasami temple trustee 1918-24, head trustee 1931-36), himself a descendant of Ramsami and Ratnavelu, two of the progressive big-men of the nineteenth-century lawsuits, told me that until the late 1950s Beeri Chettiars married street to street, with no intermarriage between Park Town and Muthialpet, even when families belonged to the same gumbu . People liked to marry people they knew intimately and with whom they had enduring relations. Known to one another, these were families that would have something to lose if they were to act dishonorably, namely, their reputations and the trust of others, losses that would carry high costs in business and social networking.

Informants also describe a strong desire for the families linked to them by marriage to share their interests and lifestyles. But the Town Beeri Chettiars were a heterogeneous community. In the twentieth century, Park Town had the highest concentration of prominent business families. Muthialpet also included business families but was occupationally and socially more diverse, counting among its numbers families in prominent service with the English in business and government as well as increasing numbers of new professionals. Complicating marriage matters, in Muthialpet, several prominent families were also considered to have traces of English blood, the result of generations of close association with the British. In the eyes of many, this tinged their eminence with a strong sense of ambiguity, and Park Town's elite avoided intermarrying with such families. Some members of these families are said to have had blue eyes and light skin, and, as several informants remarked, they preferred tea to coffee, as their descendants do now—"just like the English." It is also true that some of the most prominent families counted Christians among their kin, as the nineteenth-century lawsuits, my genealogies, and public knowledge of family histories bear witness. Descendants from among these families continue to be among today's most prominent big-men, nonetheless.

The hereditary family priest serving the Muthialpet Beeri Chettiars once remarked to me that the meaning of "gumbu" was a group doing a single type of business. Although an exaggeration, the families of each gumbu did tend to specialize in particular businesses—a matter of fathers teaching sons their line of work, Bala told me. For example, the Poonamallee and Salem gumbuhal were known as dealers in iron and steel and as wholesalers and retailers of betel nut. But they were also coconut, provision, and rice merchants in Kothawal Chavadi, the city's wholesale produce market, and wholesale and retail textile merchants. Kasukaarar gumbu families were known as close associates of the British and, as I have noted, were distinguished as cashiers and bullion dealers, although they too counted families engaged in a variety of businesses. Similarly, the leitmotif of "Tobacco" (Pangaiyalai) gumbu families was tobacco, while nonferrous metals were the specialty of the "Tin-sheet" gumbu , the Tagararkaarar Chetti gumbu . Again, families belonging to both these gumbuhal pursued other enterprises as well.

Another reason for the preservation of the Beeri Chettiar gumbuhal and the Muthialpet-Park Town community was that under British colonial rule, Madras City was the capital of Madras Presidency, and George Town was located at the center of business, finance, education, and government. Situated at the Town's southern edge, the High Court of the Presidency was literally just across the road from Muthialpet (see map 2), which was a short walk from the Secretariat in Fort St. George. Next to the court was the Law College. Across from it and also in the Town were Christian College (founded 1837), and later Pachaiyappa's College, although subsequently both moved to less crowded quarters.^[4] These three colleges played important roles in the late-nineteenth- and early-twentieth-century creation of a Western educated Indian elite in the Presidency.

The listing of a few names illustrates the growing presence in the twentieth century of Beeri Chettiar big-men among this educated elite. The purohit^[5] of the Muthialpet Beeri Chettiars says that in pre-Independence Madras, Soma Chettiar (Kasukaarar gumbu ), who was head trustee of the Mallegeeswarar temple on Thambu Chetti Street (see map 2), was also a president of Pachaiyappa's College as well as a "director" of Binny and Co., one of the major English companies. The priest says Soma Chettiar was a descendant of the Thambu Chetti family. His grandson still lives in Muthialpet today, although he is an ordinary person without claims to big-man status. Soma Chetti's sister's husband was the hereditary trustee (suroodiriyamtaar ) of the Tiruvotriyur temple to the north of George Town. His sister's son is the head trustee and suroodiriyamtaar today. Another family, well known prior to Independence, was that of

O. (for Ottakadu gumbu ) Tanikacalam Chettiar, descendant of Lingi Chetti. O. Tanikacalam was for a time a small-causes court judge, a Madras Legislative Council member, and president of the Madras City Corporation. One of his brothers, O. Kandasamy Chettiar, was a student of Dr. William Miller, the highly respected founding principal of Christian College. O. Kandasamy Chettiar himself became a well-known teacher in the college. His daughter, O. Jeevarathanam Ammal, B.Litt., was the first non-Brahman woman graduate in the Town. Other members of the family were advocates, doctors, engineers, military officers, and bank officers. V. Chakra Chettiar was yet another prominent graduate. He was the brother of Venkal Sundara Ramanujam Chettiar, a trustee of the Kandasami temple from 1930 to 1935. According to the purohit , V. Chakra Chettiar was an advocate, a well-known labor leader in Madras, and for a time its mayor. He was also a convert to Christianity. In addition to the Kandasami trusteeship, his brother was the president of the Beeri Chettiar caste madam and founder of the Muthialpet Benefit Fund, a charity serving the community's poor. In 1910, Natesa Chettiar, the grandson of Ramasami Chettiar—the man whose outcasting led to the nineteenth-century lawsuits—became the first Poonamallee gumbu college graduate. His gumbu was and still is the dominant business gumbu among the Beeri Chettiars in Park Town. To this day his house is known to old residents as "B. A. House" ("B. A. Viidu"). Later Natesa was a trustee of the Kandasami temple between 1931 and 1936, and an adjunct professor of physics at Pachaiyappa's College. What is apparent from this listing of prominent Beeri Chettiar names is that George Town was a center of new opportunities, and, consequently, big-man institutional leadership, achievement of Western-style education, and eminence in enterprise had all become intertwined.

The big British and Indian banks were also located in George Town, and Beeri Chettiar leaders were prominently associated with these. The elder sister of one of my retired informants married P. Cittaraman, who became the head cashier of the Mercantile Bank (now the Hong Kong Bank). His father before him had also been a cashier in the bank, while his father's brother was the cashier of Binny and Co. Marambuttasamy Chettiar, the grandfather of another informant, was an agent of the M.C.T. Bank, an Indian bank run by Nattakottai Chettiars on Coral Merchant Street. Nattakottai Chettiars were the premier Indian bankers of south India (see, e.g., Rudner 1989). Beeri Chettiars, especially those belonging to the Kasukaarar gumbu , were prominent cashiers in other banks, offices, and important British businesses as well: the Imperial

Bank (now the State Bank of India), the National Bank of India (now Grindlay's Bank), government offices (the tahsildar office and the Collectorate),^[6] and in Wilson and Co., Shaw Wallace, Best and Co., Parry and Co., Burma Shell, and Binny. Positions were often passed from father to son, because, as my informants explained, families had established traditions of "trust and heredity" with these companies. What they meant by this was that so long as a cashier served his company honorably, his son could succeed him. As we have seen in chapter 4, this reflected employment practices dating back to the early days of British East India Company rule.

We see, therefore, that individual Beeri Chettiars were influential leaders in their community and often close associates of the British, employed in government, banking, and business, right up to Independence. If it is remembered that personal reputations, relationships of personalized trust, and "connections" with people of influence were important determinants of how things were done and of how relationships were established in business and government, then it is clear that maintaining community ties would have proved valuable in this special urban environment, especially for Beeri Chettiars, who were members of a caste that was not only wealthy and well connected, but also the dominant caste of the Muthialpet and Park Town sections. In short, George Town was and continued to be a place where people of influence lived. If, in towns outlying Madras, people left their localities seeking to achieve influence and wealth, then Madras City was where the most ambitious were likely to go. The George Town Beeri Chettiars were at the center of things in Madras Presidency.

Nonetheless, there were significant changes in the organization of the Beeri Chettiars after the nineteenth-century lawsuits. True, the eighteen gumbuhal that composed the Beeri Chettiar caste in George Town continued to divide the caste into leader-centered constituencies. But these leaders lacked the authority of headmen and, Bala tells me, were now called perundanakaarar (syn. of periyadanakaarar ) rather than ejamaanan , headmen. Further, the Periyagramam, which had been composed of headmen and had given the caste its administrative unity, was now defunct, as was the post of caste guru. The failure to replace the caste guru, the guardian of the caste's ritual identity and behavioral rules, who had gone off in the midst of the nineteenth-century disputes, is also a clear indication that leaders remained divided among themselves and that, internally divided, they were never again able to establish their collective authority.

But the perundanakaarar still had an important role to play, which was similar to that of headmen, even if they no longer administered the caste locality as a unit. As headmen had done before them, these leaders used caste institutions—especially the three big caste temples, the Kandasami, Kachaleeswarar, and Mallegeeswarar temples—as institutions of leadership. Genealogies indicate that some of these preeminent men were former headmen, their descendants, or the offspring of marriage alliances with descendants of headmen. Indeed, as noted, Natesa Chettiar was one such leader, and Bala's father-in-law's father-in-law was Natesa's kinsman, Kali Rattina, the "diamond-giver." In the last years of the nineteenth century, probably most temple trustees and preeminent gumbu leaders were either ex-headmen, kin to them, or their close allies.

Kaattavur Subramaniam Chettiar (hereafter, K.S.) was such a twentieth-century big-man. A wealthy iron and steel merchant and the recognized perundanakaarar of Salem gumbu , he was a man of humble origins who came to Madras to work for a Salem gumbu big-man family and, proving his worth, ended up marrying the boss's daughter. Later, he started the South India Corporation with Raja Annamalai Chettiar, a company that subsequently grew very large under Raja Annamalai's direction. K.S. was the head trustee of the Kandasami temple for three consecutive terms, from 1941 to 1958. In addition, he was trustee of the Mallegeeswarar temple and of the caste madam , and, with four other big-men, served as a trustee and member of the caste funeral shrine and tank (the Attipattam Kulam) building committee. An elderly informant^[7] told me that K.S. was among the last big-men to hold regular gumbu community meetings. He called his meetings at the caste madam , sending meeting notices to Salem gumbu families living in Muthialpet, Park Town, and Chintadripet.

Another big-man and contemporary of K.S. was Maangadu Ellappa Chettiar (Maangadu gumbu ), prominent member of a low-ranking gumbu that counted a number of wealthy families among its members. Although not the preeminent leader of his gumbu , Ellappa worked hard to establish his status as a wealthy and eminent businessman. He built temple cars for both the Mallegeeswarar temple and for the Tiruvotriyur temple located in north Madras. He also built a choultry in his name at the Tiruvotriyur temple, and he built the Mallegeeswarar gopuram . With K.S. he served as a trustee of the caste burial shrine and tank and renovated the burial shrine building. These two men may be taken as examples from a lengthy list of twentieth-century gumbu leaders residing in

Muthialpet-Park Town prior to the 1960s. Their generation was the last before the breakup of the George Town Beeri Chettiar community.

Thus, although headmen and their Periyagramam council had ceased to function, eminent men, including ex-headmen, continued to play big-man roles. Asserting their generosity and altruism, some of them made spectacular endowments to community temples, as my now familiar example of the cup of diamonds that Kali Rattina gave to the Kandasami temple in 1901 illustrates. But where previously the Periyagramam had appointed the trustees of community institutions from among their number, now trustees were elected. Then again, big-men acted as brokers and arbiters of social trust on behalf of their gumbu constituents, just as headmen had done, but without the overarching authority of the Periyagramam. In other words, gumbus persisted in George Town because endogamy did not stand in the way of new paths to success, and because, even without the office of headman, gumbuhal continued as groupings of social identity and prestige, constituting important contexts within which individuals organized action, manipulated connections, achieved reputations, and managed personalized trust outside the law courts.

But despite the persistence of gumbuhal and the importance of the Muthialpet-Park Town caste domain, the Beeri Chettiar's nineteenth-century lawsuits do pinpoint the period of transition from headman rule and corporate community to what might be described as a more informal period of big-man influence, when the caste was no longer corporate, but still retained a strong community identity. Not surprisingly, this transition also marks the period when the integrity of the Town Beeri Chettiar community began to unravel. And unravel it most certainly did. Today, under Bala's headship, the Kandasami temple complex and its endowments are the only institutional aggregate left that still provides a base for castewide big-man leadership. And as we have seen, only a fraction of the Beeri Chettiar families that used to live in George Town still do.

What happened that would explain this near-collapse of what was once one of Madras City's most powerful caste communities by the end of the 1950s? The answer is that with few exceptions gumbu leadership simply failed to reproduce itself. This was not for the lack of sons, but because for several reasons the locale had lost its importance as the locus of interests and influence among caste members. In part, this was a result of Independence, which removed the British, with whom Beeri Chettiars had for so long maintained symbiotic ties, from a major role in the area. And in part, it was a consequence of the declining role

business played in the lives of caste members. My genealogies show that the sons and daughters of prominent families were attending school and, in most cases, university, but sons were not returning to run the family businesses or to take over roles in George Town-based enterprises. Instead, they were taking jobs in government service or in professions which offered secure employment but lacked significant centers of influence in George Town. As a result, families no longer felt compelled to live in their old neighborhoods, and in increasing numbers, when brothers partitioned, they sold their ancestral homes in the Town and, each taking his share, moved separately to what each considered more desirable, less crowded, less expensive residential areas of the city. The role of the Town's remaining big-men was diminished by this process and their hold over their constituencies weakened. Elder Beeri Chettiars explained to me that educated caste members often felt little compulsion to subordinate themselves to gumbu leadership. Why should they? Their successes depended on education and were no longer embedded within the caste community. The community's decline was a result, therefore, of the changing manner in which Beeri Chettiars made a living. While business was potentially far more lucrative than bureaucratic employment, it was also much riskier.

In 1939, emerging from the world Depression, secure employment was new and novel, and jobs that Beeri Chettiars would consider not very desirable today seemed wonderful opportunities. When K. Sundaram (b. 1921), a retired General Post and Telegraph officer and scion of a once influential family, told me his own story, he laughed good-naturedly at what he clearly saw as the irony of his own career choice. Sundaram comes from a prominent Kasukaarar family that counts caste leaders among its members and once had close ties with the British. I have already mentioned his maternal grandfather, O. Tanikacalam (d. 1929), who was a lawyer and judge of the Madras Small-Causes Court, as well as a Justice Party^[8] leader. Sundaram is an only son, but reflecting the family's prominence, one of his sisters is married to an advocate who practices before the High Court, while another is the widow of a former appointed trustee of the Kandasami temple, P. Seetharama Chettiar, M.A. (trustee 1964-1969; 1969-1977), who was cashier of the Mercantile Bank prior to his death.

Sundaram told me that he graduated in 1939 with a B.Sc. degree in chemistry. After graduation, without giving it much thought, he applied to medical school and was accepted. In those days, he says, it was relatively easy to gain admission. Now, of course, medicine and engineering

are regarded by most south Indians as among the most desirable professions one can pursue. But in 1939 that was not the case. He discussed his options with his family and with them decided that instead of pursuing medicine he should take a job with the postal service, which he did in 1940. He says being an only son married to a woman who was an only child factored into the family decision. His father and in-laws wanted him to lead a "quiet life," one that would keep him near.

Sundaram told me with a chuckle that, given the prestige and earnings associated with medicine today, it is hard to imagine that the postal service could have actually appeared the better choice. But in those days many still considered medicine an unclean profession, and government service was desirable and secure work. He says his starting salary was Rs. 40 per month. In 1940, you could get a meal for 20 p., a small fraction of a rupee. "Now you cannot get an egg shell." When he retired in 1977 after thirty-eight years of service, he had risen to assistant director of the General Post and Telegraph Office in Madras. A good position, but one that has no constituency within the Beeri Chettiar community.

Retired and receiving a pension of Rs. 850 per month when I interviewed him, Sundaram is an ordinary man, well liked, but without the wealth and connections that he could call on to make himself an important leader among Beeri Chettiars. Yet, for all his ordinariness, in a neighborhood emptied of Beeri Chettiars, he holds a number of positions that once would have marked him as an important and influential big-man: he is a trustee and treasurer of the caste madam located in the Kachaleeswarar agraharam off Armenian Street and the locality representative to two other madams in which the Beeri Chettiars have interests, one in Tiruvaalangadu, some twenty-plus miles from Madras, and the other in Nerinjipet near Bhavani to the south (see map 1). In addition, from 1962 until 1985, he was first the joint secretary and then general secretary of the Tamil Nadu Ayira Vaishya Sangam, the statewide association with which Beeri Chettiars are affiliated. And he is a lieutenant and ally of Bala's. Nonetheless, despite his offices, Sundaram is a minor leader with limited influence. None of the institutions in which he holds offices attracts much interest from George Town Beeri Chettiars today and none of his offices gives him control over significant assets. In other words, his offices are "hollow crowns"—markers of status that lack substance.

Sundaram's story points up the relative absence of preeminent big-men in George Town today. The big-man form of organization had been well adapted to the high levels of personalized trust that facilitated business,

but for university-educated professionals and government employees big-man leadership is unnecessarily restrictive and personal. Elder residents in George Town explained to me that by the 1950s big-men were finding that the community was no longer willing to give them the respect they needed to carry out their leadership roles. There were still community giants in the 1920s, 1930s, and even 1940s, but by the late 1950s, the link between claims to eminence and strong community constituencies had weakened to the point where conflict between contenders was rampant.

This was especially true with regard to temple trusteeships, the institutions that gave leaders communitywide recognition. Unlike Sundaram's offices, temple trusteeships were not hollow crowns. Education, occupational diversification, and the dispersal of Beeri Chettiar families meant that too few men had constituencies large enough or cohesive enough to enable them to claim the preeminence needed to defeat or discourage contenders in competition for control of the caste's denominational temples. Seen from the perspective of leaders, the fabric of community had become thin. Looked at the other way around, Beeri Chettiars no longer felt a compelling need to support caste leaders, since big-men now played a much less important role in their lives. When, in the late 1950s, heated squabbling over control of caste temples erupted among Beeri Chettiar residents, the Hindu Religious and Charitable Endowments (HRCE) Department, a state agency, stepped in to fill the leadership vacuum, first taking direct charge of the management of the Mallegeeswarar and Kachaleeswarar^[9] temples.

By the early 1960s, trouble was also brewing over management of the Kandasami temple. Symptomatically, in 1962, the man whose symbolic role it was to call temple elections, the man who would have been the head of caste in the nineteenth century, the head trustee of the caste madam , resigned his madam position. I suspect he did so because, although there were vacancies on the Kandasami temple board that should have resulted in a call for elections, the two temple trustees remaining in office, P. A. Ragava Chetty and K. Venugopal Chetty, instead entered into an agreement with the HRCE Department wherein the department was given the right to appoint the temple's board, with the stipulation that appointees should be drawn only from the Beeri Chettiar community. Why they made this agreement is not completely clear today, but it appears that they may have done so as a bid to maintain Beeri Chettiar involvement in the management of the temple. Because of temple leadership disputes, the HRCE Department had been at the time actively

considering taking over management of the temple. The result of the deal they made was that for a time it looked as if the Beeri Chettiars, while they would remain temple trustees, nonetheless would lose direct control of the last remaining important institution of big-man leadership in George Town. With this loss, the crucial social roles that Beeri Chettiar big-men had for so long played as the organizers of their caste corporate community within the city seemed at an end.

Bala Tries to Rebuild the Community: 1964-1087

In keeping with the deal, the HRCE Department selected P. A. Ragava Chetty in 1964 as a returning member of the first appointed board of the Kandasami temple. Of K. Venugopal Chetty, however, we hear no more. P. Seetharama Chettiar, Sundaram's brother-in-law, was among the four other appointed members.

But big-men dislike HRCE Department control because bureaucratic constraints inhibit the personalized and entrepreneurial nature of their patronage. By giving the department control over the appointment of trustees, the Beeri Chettiar caste had in effect lost its right to select its own leading men, including who among them was to be preeminent, the head trustee. Even more important, HRCE control greatly curtailed the ability of big-men to compete for temple offices and resources. Since the HRCE Department was a state agency, its control of appointments also meant that party politics would play an important role in the future selection of trustees. The HRCE takeover marked the low point in Beeri Chettiar community integration. It looked as if the community had lost its last institutional context in which meaningful eminent civic individuality could be achieved and expressed.

Three years later, in 1967, several residents rallied to regain control of the temple. Sundaram was a key figure among them. On behalf of the community, he and three like-minded Beeri Chettiars^[10] filed a petition with the HRCE Department requesting that the right to conduct temple elections be returned to the caste, arguing that the agreement that had been made was without the consent of the Beeri Chettiar community and that the two trustees purported to have made the deal, in fact, never signed the agreement. Sundaram told me that only the lawyers of the two men had signed. He insisted that because of that fact the agreement should not be binding on the caste community.

In due course, the HRCE Department ruled that the issue of control could not be decided by a writ petition, but would have to be settled by

a civil suit. Accordingly, in 1976, Sundaram and his colleagues filed suit and requested the court to allow the caste again to hold temple elections until such time as the case could be decided. This request was granted, and elections were called in 1978, the first since 1959. As a result of those elections, Bala was elected to his first term as head trustee.

Like the generation of leaders who immediately preceded him, Bala straddles the worlds of education and enterprise. He is both an engineer employed in a government undertaking, All India Radio, and, succeeding his father, a successful businessman, the proprietor of a retail iron and steel enterprise. He is a man of enormous energy, intelligence, and accomplishment with extensive ties within the Beeri Chettiar caste community in Madras City. In part, as mentioned before, these ties stemmed from his links to preeminent men of the past. In particular, he counts among his wife's antecedents a number of prominent community leaders including KaliRattina, the diamond-giver, his wife's grandfather. Although he is himself a member of the Salem gumbu , Bala, who married in 1959, belongs to the first generation of Beeri Chettiars among whom inter-gumbu marriages were common. His wife's gumbu , like that of Kali Rattina, is the top-ranked Poonamallee cluster. Bala also counts among his affines Natesa Chettiar, whom I have described above, a link that ties him to the progressives of the nineteenth-century lawsuits. Through these ties, Bala is related to a number of former Kandasami temple trustees, both in the distant and more recent past, and he traces kinship ties to some of his closest allies as well as to some of his most dedicated enemies.

But Bala's success in the 1978 elections was by no means merely a matter of connections. Already by that year, he had earned a reputation of his own, one which he had achieved as a community leader among the Park Town Beeri Chettiars. Key to this reputation was the "arts" association that he had founded in 1969 to promote music, drama, and literature, the Kandan Arts Academy. Bala says he founded this association with the aim of bringing the Beeri Chettiar caste community together again, uniting them through patronage. Today the Academy has four hundred members, drawn mostly from the Park Town Beeri Chettiar business community. It annually sponsors two major public concerts, held each year on a platform erected in the street in front of the Kandasami temple under a huge thatch pandal (roof) constructed for the occasion. The concerts are grand events that bring well-known music stars to perform, and they are extremely popular. Spectators pack the street. During the concerts, Bala plays a prominent public role and is seen as its organizing patron.

Leaders like to sponsor dramatic events to attract public interest. In 1974, the arts academy gave two cash prizes of Rs. 51,000 to artists it judged up-and-coming, one to the music director S. B. Subbaya Naidu, the other to the poet S. D. Sundaram. Artists like these two men are low-income people, Bala says, so the prizes gave them a boost. The prizes were raised by the academy's performances and were a gesture of grand patronage that attracted a great deal of public attention. Bala told me in 1986 that he hoped to make similar awards in the near future. Through the academy, therefore, Bala successfully organized the Park Town Beeri Chettiar merchant community as sponsors of what have proved to be very popular annual events. Doing so, he earned a reputation as a hard-working altruistic community leader as well as the backing among his fellow merchants that enabled him to win his first trusteeship election.

In 1978, the extent of Bala's ties within the caste community was reflected in the 1,062 votes he garnered. The election was a close race in which he won the head trusteeship by a mere 28 votes. The nearest contender was his ally, K. Kesavalu Chettiar, who had served as an appointed trustee since 1964. Twenty-two candidates had contested the election, competing for 2,667 voters, each of whom voted for five candidates, five being the number of board members to be elected. Bala's detractors put this 1978 election in perspective by noting that less than half of the 7,500 eligible voters voted. But the electorate had to travel from all over the city to the Kandasami Spring Hall next to the temple where ballots were cast. This is more than a small inconvenience in a city the size of Madras, and indicates that Bala and his runner-up did well to rally the supporters they did.

Big-man politics accentuate personal rivalries and are typically contentious. It is hardly surprising, therefore, that Bala was to have enemies even among his fellow elected board members. As it turned out, his key rival was the younger brother of P. A. Ragava Chetty, a man whom I shall call Kaliraja. Recall that Ragava Chetty was one of the men who originally made the deal with the HRCE Department and was then reap-pointed by the department to the temple board. After the 1978 election, as Bala's successes as a leader began to mount, Kaliraja was to do his best to unseat Bala through a steady barrage of charges of wrongdoing, including a lawsuit accusing him and his temple board allies of malfeasance in the administration of temple affairs (Mines and Gourishankar 1990). Kaliraja's aim was to succeed Bala as head trustee. Failing that, he hoped to weaken Bala's influence by supporting increased HRCE Department involvement in temple management, including as a last resort

a department takeover of the temple. If Kaliraja could not be head trustee and thereby enhance his own civic individuality, then he wanted to destroy the temple as an institution of leadership for others.

In the context of these conflicts, the case to return temple elections to the caste community finally came to trial in 1986 and, despite Kaliraja's efforts, was decided in favor of the Beeri Chettiar community in 1987. During the trial, Sundaram and his associates were de facto representatives of Bala's factional interests, while Kaliraja backed HRCE control. As events transpired, Bala's allies were able to discredit Kaliraja in court and so disarm his charges against Bala while simultaneously undercutting Kaliraja's claim that HRCE control was needed. During the course of the trial, Sundaram told me that no one was taking Kaliraja's charges seriously, including the HRCE Department. If they had, he said, the department would have intervened in temple management.

The turning point in the trial seems to have come when Sundaram was able to impugn Kaliraja's integrity by gleefully countercharging that Kaliraja was apparently not even a Beeri Chettiar and so should not be involved in Kandasami temple affairs. To Kaliraja's great embarrassment, Sundaram documented in court that Kaliraja had fraudulently registered his son as the member of another caste in order to lay claim to special educational benefits. The court burst into laughter at this piece of evidence. Following that court session, to this enthnographer's eye, Bala and his friends could only be described as in the highest of spirits. They were absolutely tickled with the way their enemy had been trapped by his own indiscretion. Some months later the trial concluded, and in September 1987, for the first time in nearly a decade, new elections for the temple board were held. Bala was again elected the head trustee of the temple board, this time in what amounted to a landslide victory. His contentious opponent, Kaliraja, failed to be reelected.^[11]

Since his first election, Bala has dreamed of using his status as the head trustee of the Kandasami temple to revitalize the sense of community among Beeri Chettiars. His motivations for doing so are a mix of community spiritedness and his self-interested desire to strengthen his own leadership, which is necessarily caste-based. His motivations, therefore, are somewhat paradoxical, just as they are also inseparably intertwined. In this regard at least, Bala is probably much like his predecessors. His success as a leader depends on displays of community spiritedness, which earn him a reputation and civic stature. These are attributes that in turn distinguish him as unique, as an individual. But Bala is also aware that conditions and leadership roles have changed since the times of his pre-

decessors. He is moved by a nostalgia for the glory days of the old caste neighborhood, which he would like to recreate. But the conditions that in those days gave authority and power to caste headmen are gone, and the caste neighborhood has dispersed. Bala can build a reputation as a civic leader, and as such, he has influence, but he has little authority over his caste fellows. He can influence, but outside the context of the temple, he cannot command.

Today in Madras, few Beeri Chettiars would say that there are any recognized leaders of caste. Not even Bala gets that recognition. Yet, if Bala could, he would have the galactic polity that he controls be the organization of the Beeri Chettiars in Madras. In fact, built as that polity is around the Kandasami temple, the caste electorate is his constituency, and so, in point of fact, his polity is an important part of what constitutes the organization of the Beeri Chettiars in the city today. But for most, that organization and leadership only lightly touches their lives.

Bala sees himself as a leader struggling to pull together a disintegrating community. He would like to reinvolve community members—to create the conditions where caste members would have a renewed stake in their relations with each other and would again turn to the big-men of the caste for leadership. He would also like to bring all the institutional interests and assets of the caste under a single administrative structure, which he envisions himself as heading. But there are few Beeri Chettiars who share Bala's vision or who see real benefits in the community organization of old. They no longer need gumbu headmen and are glad to be free of the constraints that the old system imposed on them. Nearly every Beeri Chettiar in Madras knows Bala. Many have even voted for him. But they do not think of him as someone who has power over their lives, nor do they envision him ever having such power.

Among his caste fellows, Bala is a perundanakaarar , a big-gift-giver, an influential big-man, and the preeminent trustee of an important Beeri Chettiar caste temple, but that is it. To be effective he must appeal to followers as an altruistic benefactor, but there is a halo of suspicion that always surrounds public altruism when it is associated with achieved charismatic preeminence and control of public resources. After all, the more a leader benefits his followers, the more important his own individuality becomes. Who is he helping more—himself or his constituents? In Bala's case, the suspicion is mild, but it is there. Tamils think it naive to presume that any public figure's actions are moved purely by altruism. One elder informant, a retired municipal politician and something of a self-styled skeptic, summed up his feelings about Bala's leadership

this way: He did not know what went on behind the scenes at the Kandasami temple (meaning that, for all he knew, Bala might be mismanaging temple funds, as his enemies charge), but as head trustee of the temple, Bala had done good work and the community had benefited. This was high praise from a man who for philosophical and political reasons had not entered a temple for over thirty years.

Today, the central institution of Bala's galactic polity is, therefore, the Kandasami temple. And it is in his role as head trustee of the temple that Bala enacts most broadly his role as patron, struggling to recreate his community. The Kandasastri Festival (Kandasastri Kotti Arccanai), a six-day period ending the day after Deepavali, the fall "Festival of Lights," may serve as a brief reminder of the breadth of appeal that the temple has among Beeri Chettiars and so its importance to Bala. During this festival, Bala says, about twenty-two thousand families visit the temple to make an offering and to sit and be greeted by Bala, the head trustee. For Bala, this is exposure at the grassroots constituency-building level. Moreover, it is exposure in an important ritual context that depicts him as the preeminent big-man within his community. The worshipers get the prestige of being seen with him, and he the opportunity to provide them the honor of his presence. It is in public contexts such as this that Bala makes himself most visible as a leader and builds his reputation as a civic individual. Bala's desire, therefore, is to achieve as much exposure of this sort as possible. This means attracting crowds to temple events.

A chief aim of Bala's is to increase the popularity of the Kandasami temple—to make the Kandasami temple the Beeri Chettiar temple—the most popular of all the temples controlled by the Beeri Chettiars in Madras City. To do so, Bala pursues two primary strategies. His first is to maximize the spectacular in temple events. Temple trustees have always used grand display and ritual innovation to attract crowds and dramatize their patronage. His second strategy is to reinvigorate the sense of community among the caste, as he says, "to get the community together." For this purpose, Bala uses the institutional resources that he controls to create benefits for as many people as possible. Again, his aim is constituency building.

A popular name for Park Town, the location of the Kandasami temple, is Kandakottam, "Kandan's Place," reflecting the landmark importance of the temple to the area. Since 1980, Bala has invested over two crore rupees (a crore equals Rs. 10,000,000) of Kandasami temple funds in projects designed to enhance the prestige of the temple and benefit the Beeri Chettiars in the city. During each of his terms in office, Bala has

completed a temple project with high spectacle value. His first, completed in 1984, was to build the gold-plated temple car. Covered with seven kilograms of gold and costing twenty lakhs^[12] (Rs. 2,000,000), it is a rarity in south India. The gold temple car has added significantly to the fame of the temple and to Bala's reputation as head trustee. Today, the splendor of this magnificent car is a major feature of temple processions in which it is used. Who would want to miss seeing it? His second project, carried out in 1989, was to orchestrate and conduct the ritual purification and renewal of the temple, a kumbabisheekam ceremony, an elaborate and expensive consecration ritual of major importance, the performance of which seems to be one of the dream goals of most temple head trustees. As part of this latter event, Bala published a commemorative volume, which documents the spectacle and describes the history of the temple and the George Town Beeri Chettiars. The volume includes numerous photographs portraying Bala himself as a principal figure. It also includes congratulatory advertisements from businesses and community institutions, some of which are Bala's, and a large collection of color photos depicting the wealth of the temple, the crowns and jewelry of the Kandasami idol, and the many golden processional vehicles that Beeri Chettiars have built for the temple. In effect, the volume is a statement that the George Town Beeri Chettiars are a wealthy and influential community and Bala is a central figure within it.

Together these two projects, the gold temple car and the kumbabisheekam , have kept Bala and the Kandasami temple in the eye of the public both because they are spectacular and because they are major projects that build on anticipation and require a great deal of community involvement. They have enhanced the reputation of both temple and leader. These projects have also been community building, increasing the pride Beeri Chettiars feel in their temple, which they equate with their collective identity.

Using the popularity of the temple as a resource, Bala is also trying to revive the tradition of giving first honors to the heads of the old leading families of the various gumbuhal . In this instance, first honors consists of sending a temple employee to the houses of these men to present them personally with invitations and announcements concerning temple functions. It also involves publicly offering them special respect when temple processions pass in front of their houses. Bala feels that too many of the old prominent families who used to handle gumbu affairs have become "just average" people and so no longer claim eminent status nor feel that they are in a position to help others. Nonetheless, he reasons that it is these men who could bring the community together again. If

they are again given respect within the community, then ordinary people may once more turn to them for help in their affairs; for example, in settling a dispute, arranging a marriage, or finding a job. Make these men eminent within the caste, reasons Bala, and they will form points of influence around which the community can coalesce.

But even Bala is skeptical about whether this can be done. Too many families, both influential and ordinary, are moving out of George Town. High real estate costs and a desire for more open surroundings are frequently cited incentives. Bala uses his own family as an example. He says his ancestral home is worth ten lakhs. Since he is one of five brothers and his father is deceased, when his mother dies, he and his brothers will sell the house. Each will take his two lakhs, Bala says, and go his own direction. Real estate is too expensive in Park Town, and his brothers already live elsewhere. It used to be, Bala says, that all of his neighbors were Beeri Chettiars. Now most have sold their homes or let them out, some to businesses. It is getting so now there is very little neighborhood left.

But, although sometimes discouraged, Bala has not given up. The temple resources that are at his disposal as head trustee have been particularly important to his community-building efforts. Using temple funds, Bala has founded a number of satellite big-man institutions designed to benefit the caste community and extend its collective reputation. Using donations collected at the temple, for example, he has founded and built two schools and is planning the development of a polytechnical college. One of the schools is a high school, open to all communities, which enrolls one thousand students. It is located in Periswakam to the west of George Town on land valued at ten lakhs. The land was donated by Bala's deceased affine, P. Venugopal Chettiar. The other is located in Park Town itself. Called "Hindu High School," it instructs children from nursery school through high school and is housed in a new two-story building, which Bala says cost seven lakhs. In conjunction with Kandan Arts, Bala has also started a small music school, which at present trains students in vina and singing.

Located to the front of the temple's kingly tower (Raja Gopuram), the main entrance into the temple, Bala has also founded a free library, featuring magazines, dailies, and more than four thousand books concerned with literature, religion, and how to start small industries. The latter category of books reflects an ideal of Bala's. He believes that all Beeri Chettiars would benefit if caste members shared a community interest in helping their poorer caste fellows. An aim of the library is to provide a source of information for the average Beeri Chettiar seeking ways to come up in life. Bala points to the Komati Chettiars and Jains as examples of communities that have successfully pursued self-help strategies in Madras.

Since the 1950s, Jains have been particularly successful in their competition with Beeri Chettiar merchants in George Town, and Bala feels they are out-competing and gradually displacing Beeri Chettiars in part because they lend money to their caste fellows who want to start businesses. The Beeri Chettiars should do the same, Bala feels.

Bala has also founded two free medical dispensaries, one on Thambu Chetti Street, the Muttu Kumaraswami Devasthanam [Kandasami temple] Chamundi Iswari Ammal Free Medical Dispensary, named after the woman who donated her house for the dispensary. The second dispensary, the Kandakottam Free Medical Dispensary, is located opposite the temple on Rasappa Chetti Street in Park Town. Within the temple itself, Bala also manages the rental of the temple's sixty-plus properties and sells at concession rates prasadam (savories "blessed" by the god) and prayer-offering (arccanai ) packets for the benefit of worshipers. And—a sign of the times—another temple concession sells religious videos.

Bala is also the managing trustee of the Vasantha Mandabam, the Spring Hall, which abuts the temple, and of a marriage mandabam on Mint Street, which is owned by the temple and rented for wedding parties and other important family occasions, such as the seventh-month ceremony celebrated on behalf of a woman by her parents during her first pregnancy. The Spring Hall is a well-endowed institution that serves as a community meeting hall as well as an institutional extension of the temple. I have mentioned, for example, how each spring the idol of Kandasami is taken to the hall to "rest." The hall is used by Bala for Beeri Chettiar Association meetings, for community meetings, for example, to select caste representatives for the Nerinjipet madam , and as the voting place for temple elections. The hall also funds a charity, the Vasantha Mandabam Arakattalai, which each month gives Rs. 500 to poor Beeri Chettiars to assist them in starting a business, finding work, or meeting marriage expenses.

Bala has also founded several other associations, independent of the temple, with the aim of serving his community and strengthening grass-roots support for his leadership. I have already mentioned Kandan Arts. In 1984, he organized the Park Town Benefit Fund, a financial institution offering secured and unsecured loans to middle-class Beeri Chettiars, as well as a deposit scheme designed to assist parents saving to meet dowry expenses. The fund was created to provide a means by which Beeri Chettiars could help one another. Bala is the fund's "honorary advisor."

In 1985, Bala also organized a new Beeri Chettiar Association (sangam ) with himself as president. Initially, the rationale for the sangam was to certify the caste identity of members who wished to register for

"backward class" benefits. These are government benefits aimed at assisting educationally disadvantaged communities.^[13] This service, however, has failed to induce many Beeri Chettiars to join, and the sangam remains a "hollow crown" at present. Nonetheless, Bala has aspirations for the association. He feels if he can get enough of his caste fellows to join the sangam , he can claim for it the right to represent caste interests generally. His goal is to make the sangam the premier institution over-seeing all organizations and endowments controlled by the caste, including the Kandasami temple. This would provide him with an institution that would administratively unify all caste assets and organizations, including the caste madam , giving him control of resources worth many crores. This kind of integrated institutional control, however, has never before existed, not even in the nineteenth century, when the Periyagramam board, composed of caste headmen, used to meet to administer caste affairs at the madam . In those days, the Periyagramam and madam were the premier institutions of caste.

Today, there is no chance of the madam regaining this premier institutional role. The symbolic head of the Beeri Chettiar's madam normally should be its Brahman guru, a part of whose function it is to instruct his Beeri Chettiar disciples about what for them is an appropriate code for behavior. As we have seen in chapter 4, by the 1870s both the role of caste guru and the enforcement of strict caste codes clashed with urban needs for behavioral flexibility. In 1876, when the Beeri Chettiar's guru left the madam , he was never replaced. Today, persistent anti-Brahman attitudes continue to make the madam unsuitable for the role that Bala has in mind for his sangam . It is likely, therefore, that the caste madam itself will remain a hollow crown. The current head of the madam is Bala's lieutenant, Sundaram. Bala's plan would give his sangam control over madam interests and would subordinate the trustee who oversees them. Sundaram told me privately he doubted the consolidation would ever take place.

Nevertheless, Bala is currently attempting to make sangam member-ship attractive to Beeri Chettiars. He has sent his supporters on a door-to-door campaign championing the benefits of a reunited community and soliciting members. But Beeri Chettiars are not attracted to join abstractions; they join when they calculate that there are real benefits to be gained. In 1985, therefore, in the name of the sangam , Bala also founded a form of financial institution known as a "chit fund," the Kandan Financial Corporation, this for the benefit of wealthier Beeri Chettiar merchants. The hope was that if the caste's perundanakaarars joined the

fund, they might be persuaded to join the sangam as well. If they did, then ordinary men might also see benefits in joining the sangam in order to be able to associate with such influential men. In point of fact, however, the fund is not really under the control of the sangam , although Bala tries to play up connections between the two.

Businessmen join chit funds of this sort to self-finance lump-sum business expenses (see Mines 1972). To begin the fund, Bala brought together twenty founders, each of whom gave between Rs. 20,000 and 25,000 to establish a reserve fund. Each of the Financial Corporation's "chits" is for Rs. 100,000. A chit-fund group consists of forty members, each of whom contributes Rs. 2,500 for forty months. Month by month, members pressed for cash bid to buy the fund, members being eligible to buy once in every period of forty months. At month's end, the month's bid is distributed as earnings among fund members and at the end of the forty-month period, when the chit is completed, members are each given a silver tumbler or other silver utensils. By 1986, the corporation had twenty groups of forty members each. Bala told me that in business, people tend to borrow outside the caste community for fear of spoiling their reputations. The chit fund, which is seen more as a risky form of investment^[14] than a form of borrowing, serves a valuable function among wealthy merchants, providing them with both earnings and a source of unencumbered cash, which they can access quickly with a high bid, if need be.

Bala's abounding personal vigor and determination can be seen in all these efforts described here. His efforts also illustrate how a leader who wishes to expand his civic identity must multiply the contexts in which he attracts and serves different clientele. A community is composed of such multiple contexts. In each new institution that Bala founds, he seeks to attract a constituency of followers, who together will recreate the community that has been lost. But except for the temple and Bala himself there is really very little that connects the different groups that constitute his following. In 1978, when Bala was first elected head trustee of the Kandasami temple, he was, as he is today, a caste leader of a new generation, a generation that lacked a strong connection between caste identity and neighborhood community. Without this connection, Bala has been constantly faced with the challenge of how to pull together a constituency he could serve. His response has been to attempt to revitalize his caste as an integrated community, not as it was—that would be impossible—but as a citywide community, founded on shared interests in the Kandasami temple and the institutional galactic polity that

surrounds it. If successful, the caste he would create would be a community only weakly integrated by a shared need to control social trust; the new community would not be a phoenix, therefore, raised from the ashes of what once was. Although Bala does not fully realize this, and despite the fact that he would still use temples and institutions to attract constituencies, much as his predecessors did, the new community would necessarily be a new kind of organization. But despite his efforts, it seems unlikely that Bala can succeed. The conditions that created tight-knit organization are past. Although he believes he is trying to revitalize his caste fellows' sense of community, what he has done in fact is to put together successfully a coalition of constituencies within his caste without recreating the community for which he nostalgically longs. Grappling with a variety of approaches to recreate community integration, Bala is succeeding as a big-man but failing to recreate the caste community. Although caste neighborhoods still exist, they are disappearing. They no longer reflect the dynamics of city organization, which, after all, reflect the needs and interests of the city's inhabitants. The city is undergoing profound change and the organization of civic leadership is changing with it. Big-man networks are still important, indeed central to city organization, but the caste neighborhoods that leading individuals once created to organize their affairs and regulate social trust are no longer essential.