INTRODUCTION

The Nakarattars of Tamil Nadu

This book examines a vital component in the South Indian economy, the merchant-banking caste: a corporate organization of men and families that has been crucial to processes of capital accumulation, distribution, and investment. My focus of study is the well-known Nattukottai Chettiar caste or, as they call themselves, the Nakarattars. They represent the major banking caste of South India during the period from 1870 to 1930. I present an analysis of the commercial organization and activities of the caste during the period of their preeminence. I also demonstrate the precolonial roots of Nakarattar commercial practices and commercially oriented social institutions extending back to the beginning of the seventeenth century. I do not argue that the caste was unchanged during this period. I do argue, however, that many of the activities and institutions of merchant-banking castes in the colonial era had a deep historical dimension.

My study is based, in part, on field research carried out in the state of Tamil Nadu, South India, from October 1979 to November 1981. My primary residence and base of operations was the city of Madurai, located close to the Nakarattar homeland of Chettinad and one of the many sites of substantial Nakarattar philanthropy and investment. I did not initially plan to focus my research on the Nakarattar caste. In fact, when I first arrived in Madurai, I planned to study the interaction between elite Indian merchants and political authority in this traditional, inland, urban center. Two findings emerged during the course of investigation that fundamentally altered my research topic. In the first place, it soon became apparent that-contrary to the claims of most publications on Indian commerce-elite merchants and businessmen in Madurai had relied in important ways on their participation in corporate caste activities for the conduct of commerce. Secondly, it became evident that the most powerful members of Madurai's mercantile elite were Nakarattars and that the caste organization in which they participated operated beyond the local limits of the city, beyond even the limits of South India, and instead operated throughout the macroregion of British Southeast Asia. The implication for my study was clear: to understand the activities of elite Nakarattar businessmen in Madurai, it was necessary to consider the activities of both elite and nonelite Nakarattars and to consider their interaction over the entire sphere of their coordinated economic activities.

The Nakarattar caste numbered perhaps ten thousand in 1896, forty thousand in 1920, and, by 1980-81, when I carried out my field research, approximately one hundred thousand people.^[1] Their lifestyle combined qualities common to settled agriculturalists, urban industrialists, and itinerant merchants. Accordingly, I did not have the option of confining my study of Nakarattars to a single village, or even to a single urban center. But I grew to know a core of Nakarattar families living in Madurai and Madras. And during the last four months of my field research I toured extensively through the villages of Chettinad, staying with Nakarattar families in their homes or at pilgrim rest houses connected with Nakarattar temples.

Since my focus is historical, I rely primarily on oral and documentary evidence rather than firsthand observation. The documentary evidence available to me included both Tamil and European materials. The former comprise temple inscriptions carved in stone, palm-leaf manuscripts, caste journals, family genealogies, and account books from Nakarattar banking houses. Among documents I label as "European," I include documents such as English-language newspapers written by Indians and documents from different branches of the Indian colonial government as well as documents written specifically by or for Europeans. The oral evidence consists of interviews with Nakarattar informants who described what they knew of their caste and family histories, and the details and rationale for a variety of Nakarattar customs. A few key informants proved especially informative and cooperative. They provided a disproportionate amount of the information reported in the book. But I have checked the data supplied by individuals for its internal consistency, for its agreement with data supplied by other knowledgeable persons, and for its agreement with relevant documentary sources.

Historical Overview

Records of Nakarattar activities are scanty until the middle of the nineteenth century, when they finally blossom under the fertile hands of colonial bureaucrats. Specific information about banking practice comes only in the second and third decades of the twentieth century, in the form of account books and reports from government inquiry committees (although these sources are supplemented by the memories of retired Nakarattar bankers). Interestingly, it is information about the connection between business and religion that recurs most reliably, not only for the period from 1600 to 1930, but also for the period extending back to the early Chola empire (A.D. 900-1200, cf. Abraham 1988; Hall 1980; Spencer 1968; Stein 1980). Although it is possible to trace many Nakarattar commercial practices back to the Chola period, the caste itself does not appear in the historical record until the seventeenth century,^[2] when they were involved primarily in small-scale, itinerant salt-trading activities in the interior regions of Tamil-speaking South India. By the eighteenth century, some individuals had extended their business operations as far south as the pearl, rice, cloth, and arrack trade of Ceylon; others, as far north as the rice and wheat trade in Calcutta. As in the case of other mercantile groups, trade was inseparable from money lending and other credit-extending operations.

By the nineteenth and early twentieth centuries, especially after the opening of the Suez Canal, Nakarattars were the major sources of finance for myriad agrarian transactions between Burma, Ceylon, Malaya, and the Madras Presidency. They dominated the role of mercantile intermediary between foreign British rulers and local populations by monopolizing important components of the credit, banking, and agrarian systems of Southeast Asia, and by remitting huge amounts of capital from Southeast Asia back to their South Indian homeland for industrial investment and large-scale philanthropy. During the twentieth century, the Nakarattar business environment was altered in crucial ways by the development of nationalistic movements in Southeast Asian countries, by the general growth of legislation restricting indigenous forms of banking, and by the increase in industrial opportunities within India for non-British businessmen. The consequences were significant. The caste organization of the Nakarattars began to unravel in the face of multigovernmental interference with traditional banking practices. Elite members of the Nakarattar caste began a gradual transfer and "freezing" of investment capital from mercantile to industrial ventures. Nonelite Nakarattars-perhaps 80 to 90 percent of the caste-were forced to scramble for new employment opportunities, often working as employees in government and business offices, although many of these are owned or managed by Nakarattars.

Being a Nakarattar in Colonial India

To be a Nakarattar in the period from 1870 to 1930 was to belong to one of the great "country fort" houses (nattukottais ) of Chettinad and to worship one's family's deity (kula teyvam ) in one of Chettinad's seventy-six villages. As young Nakarattar boys grew up (I have very little information about young girls), their families trained them in the ways of business and, when they were old enough, apprenticed them to family firms with agency houses located in far-flung business stations throughout South and Southeast Asia. After a three-year tour of duty, the boys would return to Chettinad, to their families, to their neighbors, and to their temples. There, they would rest briefly, perhaps no longer than the three-month summer, during which most Nakarattar marriages and temple ceremonies occur. Then they would return to work in the outside world as young men. On his second tour of duty, a young man might return to the employ of the same firm for which he had worked before, perhaps even working as the principal agent (mudali ) in its agency house. Alternatively, he might take over as resident proprietor or partner in an agency house for his own family firm. Eventually, a Nakarattar businessman would retire to Chetti-nad, where he would direct the activities of agents for his own family firm and join the round of ceremonies with the women, children, and other retirees who resided permanently in Chettinad.

Long before this, when family and financial circumstances permitted, a young man's joint family (kutumpam, valavu ) would take advantage of his periodic return to Chettinad and arrange a marriage with an auspicious bride from another family. The ceremony initiated an elaborate pattern of gift giving and potential business cooperation within a kindred formed by patrilineal relatives (pankalis ) of the husband and wife. This pattern of gifting and cooperation continued until all children born to the couple were established in alliance-forming marriages of their own. Alternatively, it might be continued beyond the marriage of offspring if two joint families of (what anthropologists call) "cross-relatives" within the kindred chose to renew the alliance begun with the original marriage. In either case, candidate families for marriage alliance were chosen from within the same territorial division (pirivu, vattakai ) of Chettinad. Both families would register the marriage at their respective clan temples (nakara-k-kovils ).

Many Nakarattars loaned money to agriculturalists and artisans within the Madras Presidency, especially within their Chettinad homeland and within the rice-growing regions and coastal ports of Tirunelveli and Than-javur. But business in the Presidency was risky and paid a poor return on financial investment. As a consequence, the majority of Nakarattar business dealings took place overseas, in Southeast Asia. There, supported by the British colonial administration, Nakarattar business firms supplied credit to all who required it, at interest rates that were exorbitant by today's standards but generally far lower (in some cases, by an order of magnitude) than those supplied by their competitors.

This is not to say that Nakarattar business dealings were undertaken as acts of selfless charity, although Nakarattars in hearings by the colonial government did occasionally suggest a rather lofty motivation (just as bankers do today). Certainly their clients were under no such misapprehension (Adas 1974). Under the colonial government, Southeast Asian agriculturalists lacked the legal rights and the local organization that made money lending in Madras so risky. And, in an economic environment in which agricultural return on borrowed money was marginal, they often lost the land that they had improved as much by the labor of their hands as by inputs of Nakarattar capital. Nevertheless, it was Nakarattar bankers who provided the financial wherewithal for many Southeast Asians to make a living and who ultimately made possible the tea and coffee plantations in Ceylon, the rich rice frontier of lower Burma, and the tin and rubber industries of Malaya.

Within the Madras Presidency, a few of the wealthiest Nakarattars involved themselves with the largest peasants in the Indian countryside: the zamindars ("landlords") descended from Tamil chieftains who, during the last two hundred years, had lost a millenial struggle for local autonomy to the emerging British colonial regime. Nakarattars regarded these "little kings" with a very British attitude of scornful sympathy for irresponsible natives who wasted money on useless ceremonial expenditures. Nakarattars often interceded on their behalf, lending money to cover revenue arrears in return for a permanent lease on the most productive villages in a zamindar's domain. These villages might, with help from the goddess, show a profit in the gradually improving commodities market of Madras in the latter part of the nineteenth century. In other cases, borrowers were unable to pay their debt, and Nakarattar lenders gained clear title to land. In the process, some Nakarattars even gained ceremonial rank in colonial society as they became zamindars themselves, or received other titles such as Rao Bahadur or Raja Sir.

The major sources of Nakarattar pride and honor, however, were their families, their business acumen, and their devotion to their gods: three indissociable components of Nakarattar life. Nakarattars gave generously to Siva temples in their ur (residential village) and in their nakara-k-kovil (clan temple). They contributed to temples for Siva, Aiyyanar, and the goddesses for the villages in their leased land. And they made contributions for the construction of Murugan temples with the other Nakarattar bankers in their nakaravituti (Nakarattar "rest houses" or choultries) wherever they did business. As recently as 1980, a wealthy Nakarattar textile mill owner reported to me that he proposed to give one hundred thousand rupees to each of the twenty-three major Saivite temples in India. He fully expected Siva to bless his family and his business in order to make this possible.

Nakarattar Studies

A central theme of this book is that it is impossible to understand Nakarattar business practices without understanding Nakarattar kinship. The focus on kinship, however, poses an evidentiary problem since kinship constitutes the least-documented domain of historical Nakarattar practices. Accordingly, my description of Nakarattar kinship during the colonial period is based primarily on personal observation of contemporary Nakarattar rituals in Chettinad and on field interviews with Nakarattars about their family histories and about the histories of their clan and village temples. Despite the absence of concrete historical documentation, I believe the picture I present of Nakarattar kinship is reasonably accurate as far back as the beginning of the seventeenth century. Certainly, documents from that early period indicate that the basic units of Nakarattar kin organization were all in place (Rudner 1987), including its hearthholds (pullis ), joint families (valavus ), lineages (kuttikkira pankalis ), residential villages (urs ), clans or temple groups (nakara-k-kovils ), affinal kindreds, and territorially bounded "microcastes" (vattakais ).

Aspects of Nakarattar history and customs have been addressed in a variety of publications.^[3] Of these, Thurston (1909), Krishnan (1959), Tun Wai (1962), and authors of various papers contained in reports of the Madras, Burma, and Ceylon Banking Enquiry Committees (MPBEC 1930, BPBEC 1930, CBEC 1934) constitute the basic sources of information for most of the more recent work. Thurston provides a standard account of Nakarattar social organization and custom. Krishnan, Tun Wai, and the banking enquiry committee papers present useful information on the operation of Nakarattar financial instruments and banking practices. It was my impression that, due to reliance on these sources of data, most scholars know Nakarattars primarily in their capacity as money lenders. Their involvement in and contributions to Asian industrialization were almost unrecognized. To remedy this bias, a great deal of my field investigation was devoted to tracing Nakarattar involvement in South and Southeast Asian industry by consulting native documents collected by one of my principal informants, the late Nakarattar writer and publicist Somalay (S. M. Lakshemanan Chettiar). After I had carried out considerable research, however, I became aware of Raman Mahadevan's (1976) thesis on precisely this topic, which changed the direction of my own research.

Among other publications concerning Nakarattars, Adas (1974) and Siegelman (1962) make use of information contained in reports of the Burma Provincial Banking Enquiry Committee (BPBEC) and various Burmese settlement reports to describe the role of the Nakarattars in Burma. Christopher Baker's 1984 study draws on the Madras Provincial Banking Enquiry Committee (MPBEC) report (1930) and other useful publications, such as Krishnan's (1959) study of indigenous Tamil banking, and offers an insightful and contextualized appreciation of Nakarattar commercial practices in the financial markets of inland Madras.^[4] Other writers have concentrated on legal, political, or ritual activities of members of the community. Weersooria (1973) points out that Nakarattar litigation was responsible for significant developments in Ceylonese legal precedent and focuses his study primarily on this topic. Breckenridge (1976) and Price (1979) make use of information contained in legal briefs to explore the interaction of specific Nakarattar individuals with South Indian temples and kingdoms. Moreno (1981, 1984; Marriott and Moreno 1990) makes use of observations and interviews obtained during recent anthropological field work to formulate a cultural account of a Nakarattar pilgrimage.

To date, however, there has been no systematic effort to synthesize the various kinds of information emphasized in the work of these Nakarattar scholars. Moreover, none of these scholars really address the unique blend of Nakarattar caste organization, religious custom, and business practice or the implications of this blend for understanding Indian culture and society. As a result, their findings about Nakarattars have had little effect on South Asian scholarship. In fact, outside the circle of anthropologists and historians mentioned above, most South Asian scholars know of the Nakarattars primarily through acquaintance with Thurston (1909 V:249-271).

Nakarattars and the Anthropological Study of Caste

The present study of Nakarattars, then, explores two aspects of South Indian commerce and society. On one hand, I examine the way a specific caste acted as an institution of banking and trade. On the other hand, I examine the way a specific financial institution functioned as a caste. My findings have implications for issues presently at the forefront of research in Asian history and anthropology, including the nature of noncapitalist economic formations and protoindustrialization, the impact of colonial rule on indigenous commercial systems, and variety and change in India's caste and ethnic groups. These considerations shape the organization of this book, guiding it from an initial survey of the study of commerce in Indian society to a critical revision of standard anthropological and historical conceptions of caste.

My approach to the last topic deserves some preliminary comment. To begin with, I have deliberately undertaken an analysis of the social organization of the Nakarattar caste as well as the business practices of individual Nakarattars. In anthropology, such a focus has become increasingly unfashionable during the last twenty-five years, along with studies of social organization in general. I return to it, not out of a fetishistic love for the anthropology of days gone by, but because the focus on collective institutions holds potential for bringing together scientists and humanists, who, in various incarnations, have polarized anthropological discourse since its inception as a formal discipline. The "scientists" of the field have generally been interested in large-scale, diachronic, "macro" phenomena; the "humanists," in small-scale, synchronic, "micro" phenomena. There are, of course, exceptions. One need only think of work by such humanistic stalwarts as Dumont, Geertz, or Tambiah to recognize that the scientists have not yet acquired a monopoly on macro perspectives. Nor are psychological and cognitive anthropologists-not to mention economic anthropologists employing microeconomic theory in non-Western societies-willing to cede microanalysis to the humanists. Despite these qualifications, it seems to me that anthropology has flown outward in two different directions from its older focus on institutions. In doing so, it has lost track of institutional structures that mediate between individuals and their historical environments. Economic anthropologists and political economists have moved onward and upward in their analyses; interpretive and postmodern anthropologists have sunk downward and inward. One reason-though certainly not the only one-that the two groups don't talk to each other is that they are no longer interested in understanding the same kinds of phenomena. Those parts of the present study that focus on institutional analysis are offered in the belief that they provide a neutral ground for cooperative action.

Although my study of the Nakarattars requires that standard conceptions of caste be drastically qualified, even a cursory glance through the pages of this essay reveals how much I owe to scholars who have preceded me. The first is Louis Dumont. But it is Dumont the ethnographer and author of Une Sous-Caste de l'Inde du Sud (1957a) rather than Dumont the grand theorist and author of Homo Hierarchicus (1980 [1970]) who has been most influential. The brilliance of Dumont's (1957a, 1957b) work on the Pramalai Kallars established the importance of status hierarchy and affinity or marriage alliance as twin principles complementing the principle of descent in South Indian kinship relations. Although in his most recent (1983) work Dumont alters the way in which he presents his theory-speaking of values held by South Indian actors rather than of principles ordering anthropological models-he has not altered his view that descent, hierarchy, and affinity are all crucial for understanding South Indian kinship, whether one is a South Indian or an anthropologist. Nor does Dumont's shift in formulation allow an anthropologist working in this area the luxury of ignoring his insights.

My quarrel with Dumont is that he tends to see the general Indian landscape through Kallar'ed glasses. It seems to me, as it has to others, that Dumont has closely observed forms of descent, marriage alliance, and hierarchy, as exhibited by the Kallars and other dominant agrarian castes, but that he overgeneralizes this pattern to all South Indian castes. Many caste groups do not structure their lives around an elaborate formation of descent groups, even if they do exhibit marriage alliance and hierarchy. Other castes do not exhibit perpetual marriage alliance even if they make use of Dravidian kin terms to classify their caste mates. Still other castes exhibit structural equality rather than hierarchy among their constituent segments, yet they do not fail to exhibit a holistic, many-place ordering of all their constituents that is irreducible to the single-place predication of individuals.^[5] In other words, it seems to me that Dumont's generalizations about all castes actually represent caste-specific values for variables that range over many more possibilities. Some of these values are apparently instanced by the Pramalai Kallars. But other castes exemplify different values, including the polar opposites of Kallar values.

McKim Marriott is the second anthropologist whose influence is prominent in my argument. Whereas Dumont is a French structuralist (even when appealing to values rather than structures), Marriott is an influential theorist among American cultural anthropologists. Perhaps more importantly, whereas Dumont's theory emphasizes unifying themes that run through wildly differing caste groups, Marriott focuses precisely on the dimensions along which they differ. Without necessarily subscribing to their most elaborated and still evolving formulations (cf. Marriott 1990), I have found some of Marriott's most general concepts to be useful in developing a model of South Indian kinship that encompasses caste groups ranging from Nakarattars and Kaikkolars to Kallars, Maravars, Vellalars, and Brahmans. In particular, I have found it useful to modify Marriott's (1976) preliminary formulations about Hindu gross and subtle substances and about maximal and minimal transactional strategies as part of a theory of caste as symbolic capital that complements my treatment of caste as an institution. Nakarattars maximized transactions of subtle substances, especially money, hundis (bills of exchange), and credit. They minimized transactions of gross substances, especially wives and the "seed" of their descent lines.^[6]

Brenda Beck's (1972) analysis of left-hand and right-hand castes in Kongunad, Tamil Nadu, links together the structuralism of Dumont and the systematic study of variation of Marriott. Like Dumont, however, she tends to overgeneralize from the data of her field study. It is not the case that all regions of Tamil Nadu are like Kongunad: they do not at present, nor did they in the past, exhibit a bifurcate social structure split between mercantile and agrarian castes. However, like Dumont and Marriott, Beck explores the structural dimensions of caste organization in ways that define a research agenda for my own study. This is particularly true with respect to her focus on caste histories, marriage alliance, and the linkage between descent, territorial control, and cult membership.

A fourth anthropological contribution to my own understanding of caste traces its lineage to A. M. Hocart's (1950) theory of caste as an institution based not on kinship but on kingship. More recent ethnohistorical studies by Appadurai and Breckenridge (1976; Appadurai 1981; Breckenridge 1976), Dirks (1988), and Price (1979) extend Hocart's interpretation by showing how South Indian kingship was articulated with South Indian temples and religious endowments. Together, all three institutions provided an arena for the interaction of worship and politics in South Indian society.

The importance of this perspective lies not merely in its expansion of the anthropological gaze on India but also in its correction of a broadly Western orientalist and colonial bias that viewed indigenous institutions of governance as merely "hollow crowns," devoid of power or authority.^[7] In my study, I focus on the further articulation of kings, temples, and endowments with a variety of commercial institutions (formal and informal) that, like institutions of kingship, have been excluded from the Western image of India. Again, my concern here lies in the characteristic ways in which different castes integrate themselves at any of these points of institutional articulation.

The blind spots in orientalist visions of caste are symptomatic of a pervasive essentialism and totalism common to anthropological treatments of cultures. Although my own understanding of caste rests on the shoulders of my disciplinary elders, my approach reflects the last decade's historio-graphically conditioned discontent with any search for timeless cultural essence, whether of hierarchy and hollow crowns in India; mud, blood, and semen in New Guinea; shame and honor in Mediterranean cultures; or alienated individuals in Western civilization (Appadurai 1986; Bourdieu 1977; Ortner 1984). Certainly, my analysis of the Nakarattars is directed not at distilling the essence of caste or of the caste system but at exploring the structure of variation among different castes located throughout India's social and historical landscapes.

Accordingly, historiographic treatments of caste influence my analysis as much as do the anthropological treatments reviewed above. Indeed, the various anthropological approaches to conceiving of caste provide only one vertex of a triangular discourse in which I have tried to act as mediator. The other vertices consist of a discourse on caste by historians and a reflexive discourse on their own identity by Nakarattars themselves. I rather fear that some anthropologists may find my tale overburdened with historical detail, while some historians will find my history underdocumented and, what's worse, deflected by digressions into the trivia of kinship terminology and ritual. I am sure that my Nakarattar friends will consider that I have barely touched the rich texture of their lives and histories. To all these likely complaints, I must agree. My excuse in steering this preliminary investigation of Nakarattar history between three sets of interests is to provide a platform for future studies of a people whose practices can teach us much about the political economy of South and Southeast Asia.

Outline of the Book

My study is broken into four parts. The first provides a theoretical orientation based primarily on empirical studies-ethnographic and historical-of caste and capitalism, the two focal concepts of this book. Chapter 2 reviews central themes in standard and revisional theories of caste, setting the stage for my reconstruction of a system of continuous linkage between Nakarattar business practice, kinship organization, and religion. My guiding assumption here is that it is possible to examine the internal social organizations of different caste groups, contrast features found within their organizations, and explain these differences as adaptive responses to the occupational specializations of the castes. Chapter 3 discusses the range of theoretical perspectives developed by anthropologists and historians for interpreting Indian commerce and discusses the blindness of these perspectives to major components in Indian commerce, such as castes specialized around the business of banking.

In the second part of this book, I present three chapters exploring the history, practice, and organization of Nakarattar business. Chapter 4 provides a historical context for analyzing Nakarattar social organization in the colonial period. It traces the explosive growth of the Nakarattar Southeast Asian financial empire from early, scattered references to it in nineteenth-century East India Company documents to the prominent role it plays in the banking enquiry committee reports of 1930. Chapter 5 describes the formal organization of financial cooperation between caste members, especially in regard to their activities of deposit banking, exchange banking, collective decision making about interest rates, and accounting. These activities are shown to depend on careful reckoning and bookkeeping of degrees of trust, especially among the Nakarattars themselves. Chapter 6 continues to explore the formal apparatus of Nakarattar finance, shifting attention to the institutions devoted explicitly to business activities. Alternative interpretations of Nakarattar business institutions are evaluated in light of information gathered during my field study.

The third part of this book explores the informal bases of Nakarattar business organization. Three chapters focus on elite philanthropy, marriage alliance, and descent-based cults in Chettinad, respectively. Each chapter places its discussion in the context of relevant anthropological literature concerning the relationship of religion to kinship, on the one hand, and to business practice, on the other.

The conclusion of this book returns to the theoretical issues that were framed in Part 1: the nature of caste and commerce. Specifically, in Chapter 10, I summarize my findings about Nakarattar kinship organization as an adaptive response to financial occupational specialization, explore their implications for standard models of Dravidian kinship, and begin a project of systematic comparison between Nakarattars and nonmercantile forms of caste organization.