I. The Concept of CasteGerald D. Berreman
II. The Indian Caste SystemAdrian C. Mayer
The term “caste” has been widely used to describe ranked groups within rigid systems of social stratification and especially those which constitute the society of Hindu India. Debate over whether castes are found outside of India has intensified with increased knowledge and understanding of the Indian caste system. Among social scientists, and especially among those who have worked in India, there are basically two views: (1) that the caste system is to be defined in terms of its Hindu attributes and rationale and, therefore, is unique to India or at least to south Asia; (2) that the caste system is to be defined in terms of structural features which are found not only in Hindu India but in a number of other societies as well. Those who hold the latter view find caste groups in such widely scattered areas as the Arabian Peninsula, Polynesia, north Africa, east Africa, Guatemala, Japan, aboriginal North America, and the contemporary United States.
Either of these positions is tenable; which is preferable depends upon one’s interests and purposes. The caste system of India is unique—in the religious ritual which explains it, in its complexity, and in the degree to which the constituent groups are cohesive and self-regulating. Indeed, there are significant differences in caste as it is exhibited within Hindu India from place to place and even from caste to caste within the same locale. However, caste can also be defined in terms which give the concept cross-cultural applicability, thus making possible certain generalizations. This is not to claim that all caste systems are identical but simply to assert that there are recurrent patterns of social organization, commonly termed caste, which exhibit significant similarities along with their differences. These similarities are the basis for analytic comparisons, leading to greater understanding of human social organization and human behavior. For this purpose, similar social facts must be categorized together despite differences which, while not denied, are not critical to the generalization; thus, insights may be gained into the conditions which give rise to and perpetuate caste systems and into their social, cultural, and psychological concomitants.
A. L. Kroeber defined caste as “an endogamous and hereditary subdivision of an ethnic unit occupying a position of superior or inferior rank or social esteem in comparison with other such subdivisions” (1930, p. 254). Thirty-five years later, and with many times as much research literature available on India and on social stratification, this definition has not been significantly improved upon, although there has been greatly increased understanding both of the Indian caste system and of other systems of stratification.
Kroeber’s definition, like other comparative definitions, describes caste systems as systems of social stratification, examples of ranked aggregates of people, that are unusually rigid, birth-ascribed, and permit of no individual mobility. Concepts and hypotheses derived from such comparative definitions stem primarily from the study of stratification, especially of stratification in the United States.
The application of concepts of stratification to the Indian caste system and to other caste systems as well is not wrong, but it fails to convey adequately the nature of these systems. This failing is not found in studies which define caste in a specific cultural framework; but these studies, in turn, do not say anything very relevant about man and society outside of the particular culture and the specific groups which have been described.
A more comprehensive approach, which makes for cross-cultural comparability without sacrificing cultural content, can be derived from three bodies of descriptive and analytical literature: studies of stratification, studies of cultural pluralism, and studies of social interaction. Caste systems are indeed rigid systems of social stratification, but they are also systems of sociocultural pluralism, and both of these facts can best be understood in terms of distinctive patterns of social interaction. In this way, it is possible to analyze a broader range of caste systems, castelike systems, and related phenomena than is possible with concepts derived only from stratification theory. In addition, processes of change in caste systems and changes to or from systems of caste organization can be analyzed, and the continuum from noncaste to caste organization can be described in terms of these defining dimensions.
A caste system, then, can be said to occur when a society is composed of birth-ascribed, hierarchically ordered, and culturally distinct groups (castes). The hierarchy entails differential evaluation, differential rewards, and differential association.
A society. The groups constituting a caste system are differentiated, interacting, and interdependent parts of a larger society. Often, and perhaps universally, they are economically interdependent and/or occupationally specialized. Their members view themselves and are viewed by others as relatively homogeneous elements in a system of differentially ranked component parts rather than independent and mutually unranked self-contained systems. In a caste system, everyone belongs to a caste and no one belongs to more than one caste.
Composed of groups. Each rank in the hierarchy of a caste system is occupied by socially distinct aggregates of people who recognize that they constitute discrete, bounded, and ranked entities. The size and degree of corporateness of such groups vary widely. The members usually share a group name; they interact with one another in characteristic ways; and there are identifiable symbols of group membership, ranging from skin color to cultural features such as language, occupation, dress, or place of residence. Only members of the group are one’s peers. Where group affiliation is relevant, individual attributes are irrelevant.
Birth ascription. Membership in castes is determined by birth. An individual is assigned his lifelong and unalterable status according to his parentage—status which he shares with others of similar birth who are, therefore, assigned to the same group (caste). A common means of guaranteeing this status is by prescribing endogamous marriage in the caste and ascribing to the child the caste affiliation of its parents. But this method, often cited as a defining characteristic of caste, is by no means universal. Even in India, caste, like kin-group affiliation, sometimes is assigned unilineally or according to other, more complex rules based on birth.
Hierarchy. That a caste system is a hierarchy implies that it is a system of differential evaluation, differential power and rewards, and differential association; in short, a system of institutionalized inequality.
Castes are ranked ultimately in terms of the shared “intrinsic worth” that is ascribed by birth to the individuals who constitute them. This criterion of rank may be defined and expressed in many different idioms, such as purity (as in India), honor (as in Swat), or genetically determined capabilities (as is putatively the case in the United States), but always those who are high regard themselves as more worthy than those who are low. Those who are low seem universally to question, if not the criteria of rank, then the judgment which relegates them to the low end of the hierarchy.
Caste systems rank people by birth-ascribed group membership rather than by individual attributes. Class systems, by contrast, define the rank of their members according to their individual attributes and behaviors. In a caste system, one displays the attributes of his caste because he is a member of it. In a class system, one is a member of his class because he displays its attributes. Individual mobility is by definition impossible in a caste system and possible (although in some systems statistically unlikely) in a class system.
Ranking is accompanied by differential power and other rewards contingent upon caste membership: access to goods, services, and other valued things. The ability to influence the behavior of others, the source of one’s livelihood, the kind and amount of food, shelter, and medical care, of education, justice, esteem, and pleasure—all these things which an individual will receive during his life, and the very length of life itself, are determined in large measure by caste status.
A caste hierarchy is to a large extent an interactional hierarchy. Social interaction is inherently symbolic, that is, it has meaning. Rank is expressed and validated in interaction between persons. It is manifest in patterns of interpersonal behavior and in patterns of association. Who may be one’s friend, wife, neighbor, master, servant, client, or competitor, is largely a matter of caste. Everyone is a superior, a peer, or an inferior, depending upon caste. Only within the caste is status equality found. Between castes any kind of interaction which defies or jeopardizes the rules of hierarchy is taboo, even when such behavior does not directly challenge the official bases of the rank system. Thus, there is always a more or less elaborate etiquette of intercaste relations which is stringently enforced from within and above.
It is becoming increasingly evident that castes cannot be defined adequately without reference to interaction patterns. An interactional definition of a caste system might be: a system of birth-ascribed groups each of which constitutes for its members the maximum limit of status-equal interaction and between all of which interaction is consistently hierarchical. A caste might then be defined as a network of status-equal interactions in a society characterized by a network of hierarchical interaction between birth-ascribed groups. The interactions range from informal social encounters to marriage and include a wide variety of networks, such as those based on occupation, economics, politics, ritual, and friendship.
Underlying hierarchical interaction between castes is the existence of what has been termed “status summation” (Barth 1960, pp. 144 ff.). The multiple roles played by individual members of a caste are equivalent in the status they confer. Thus, a person of high ritual status tends also to be of high economic, political, and social status. These statuses tend to coalesce, and people are thus enabled as well as enjoined to interact with members of other castes in an unambiguous, consistent, and hierarchical manner. Part of the dynamics of caste organization is the tendency for status incongruities, when they occur, to be rectified.
Because intensive and status-equal interaction is limited to the caste, a common and distinctive caste culture is assured. This is a consequence of the density and quality of communication within the group, for culture is learned, shared, and transmitted. More is inevitably held in common between those intimately communicating (that is, between caste members) than between such people and outsiders; and, because of shared culture, communication is easier and hence more intense within the caste than outside it.
Castes are discrete social and cultural entities. Caste hierarchies are discontinuous. This is a key factor in understanding the dynamics of caste systems as contrasted, for example, with class systems wherein the hierarchies are continuous. Caste systems are maintained by defining and maintaining boundaries between castes. They are threatened when boundaries are compromised. Even when interaction between castes is maximal and cultural differences are minimal, the ideal of mutual isolation and distinctiveness is maintained and is advertised among those who value the system. Similarly, even when mobility within, or subversion of, the system is rampant, a myth of stability is stolidly maintained among those who benefit from the system.
Caste and other social organizations
Caste systems resemble plural societies in the cultural distinctiveness and in the dissensus found among castes on many key values and attitudes, for cultural plurality obtains when “two or more different cultural traditions characterize the population of a given society” (Smith 1965, p. 14). This pinpoints an important difference between caste and noncaste systems of stratification. Smith (ibid., p. xi) has noted that “it is perfectly clear that in any social system based on intense cleavages and discontinuity between differentiated segments, the community of values or social relations between these sections will be correspondingly low. This is precisely the structural condition of the plural society.” It is also conspicuously the condition of caste systems.
Stratification theory presupposes, explicitly or implicitly, that among people who constitute a rank continuum there is a wide consensus on the criteria and expressions of rank in a society. It also implies a wide consensus on values, attitudes, motives, and goals. The concept of pluralism implies “a discontinuous status order lacking any foundation in a system of common interests and values while its component sections are genuine status continua, distinguished by their differing systems of value, action, and social relations” (ibid., p. 83). The plural society is held together by power rather than by consensus. Institutional distinctiveness and independence (except in certain spheres, notably, the economic and administrative ones) are also key features of pluralism.
Clearly, India falls short of being a plural society by these criteria of the ideal type, and probably all caste systems do. In fact, if the definition is taken literally, few societies fit it. But caste systems, and specifically that of India, have many plural features. Several contemporary social anthropologists have made the point, for example, that while castes in India share much by way of cultural traditions, values, attitudes, and goals, they are culturally distinct in each of these spheres as a result of their histories and associations. The nature of their shared culture is different, and its extent is considerably less than in societies which are culturally and socially continuous. Similarly, castes share some institutions with other castes, but each has important distinctive institutions as well.
We can suggest that all caste systems are characterized by plural features similar to those found in the Indian instance and that all caste systems are held together in large measure by considerations of relative power among castes—power expressed physically, economically, politically, and socially. India and the United States are typical in that the caste system of each functions as a result of sanctions in the hands of the dominant group(s) and is readily upset if the balance of power, as perceived by those in the system, changes (Berreman 1960). The dominant caste(s) exercises the power which maintains the status quo just as does the dominant group in a plural society. There is invariably an official rationale which indicates that the system functions by mutual consent—by consensus. Malfunctioning or change is likely to be attributed, therefore, to alien intervention. Actually, it is more often a result of changed power relations between groups, with consequent attempts by some of those groups to realize formerly suppressed aspirations.
Social cohesion need not rest entirely on common motives and values. It can, and more commonly does, rest on the articulation of divergent motives and values. Certainly this is true of caste systems. Consensus is not lacking between castes— they could not function within a society if it were —but it is a distinctly limited consensus. People largely agree on the facts of the behavioral and interactional hierarchy, on the membership of particular castes, and on their publicly accorded status. They agree on the hierarchical meaning ascribed in the society to particular attributes and behaviors. Where they disagree is on subjective and ideological matters: on the legitimacy of the hierarchy or, more commonly, on the place their group has been accorded in it; on the legitimacy of the criteria of ranking; and on the legitimacy of the requirements and rewards of rank. Such disagreements are often obscured by power relations and sanctions, threatened or applied.
In caste systems, as in all plural systems, highly differentiated groups get along despite widely differing subjective definitions of the situation because they agree on the objective facts of what is happening and what is likely to happen: on who has the power; and how, under what circumstances, and for what purposes it is likely to be exercised. They cease to get along when this crucial agreement changes or is challenged.
From a functional perspective, caste systems can be seen as facilitating cultural and socioeconomic differentiation. They provide mechanisms for incorporating distinct groups into a society and maintaining them as groups without incorporating their members into extant groups. Thus, they function to maintain cultural, occupational, and economic differences; to inhibit mobility; to maintain power relations; and to protect the status quo. Such systems provide blueprints for the articulation of sociocultural diversity and the protection of power and privilege.
Caste systems combine the principles of stratification and pluralism. A caste system resembles a plural society whose discrete sections are all ranked vertically. A plural society resembles a caste system wherein the groups (except the dominant one) are unranked relative to one another. In both instances there is a dominant group whose sanctions assure persistence of the system by articulating its component parts. In the village caste system of India this is the dominant caste or castes; in plural societies it is often the colonial power.
By utilizing the criteria of ranking and cultural distinctiveness (both expressed in patterns of interaction), caste systems as ideal types can be distinguished from three other important systems of organizing people in society: pluralism, class systems, and homogeneous systems wherein there are no culturally distinct or ranked aggregates (see Table 1).
|Culturally distinct groups||Caste system||Plural society|
|Cultural continuum||Class or status system||Homogeneous society|
Also, by utilizing the criteria of ranking and mode of recruitment, castes can be distinguished from other social aggregates, namely, consanguineal kin groups, classes, and community or residential groups. These are distinctive but not mutually exclusive social aggregates. All four, for example, occur in Indian society, even on the village level. Thus, a caste is generally also an extensive kin group; most castes are territorially delimited; and class distinctions commonly occur within castes as they do within kin groups and communities. Indian villagers can be defined, and can define themselves in many instances, by caste, kin, class, and community criteria (see Table 2).
|Birth-ascribed membership||Caste||Consanguineal kin group|
|Acquired membership||Class||Community (locality)|
Ranking and mode of recruitment appear to be especially important variables in the social psychology of groups and individuals, accounting in large measure for the widespread similarities in attitudes, values, beliefs, and behaviors discovered in societies sharing similar expression of these variables, for example, caste societies.
The definitions and discussion above should make it clear that “caste” and “caste system” are concepts describing one end of a continuum. Some instances fit the definition better than others. That is, some are more fully castelike than others, while some may be castelike in only certain respects. There is no mechanical rule by which all instances of social organization can be unequivocally and definitively judged as either caste or noncaste, no matter how perfectly documented or how well analyzed. In fact, by utilization of the concepts of stratification, pluralism, and interaction, this should be clearer, and the reasons should be more accurately specifiable, than if only the first of these were used.
Special types of caste organization
Each of two particular and peculiar types of caste organization have often been assumed by commentators to be either the only, or the characteristic, kind of caste organization. These are (1) caste in India and (2) pariah caste status. For the first, a culturally and regionally specific phenomenon, a term indigenous to that culture area of reference may profitably be used. Jati is a term widely applied in the Sanskritic languages of south Asia to those groups defined in English as caste or, sometimes, subcaste. South Asian caste is certainly unique and deserving of special study, but that fact should neither preclude nor inhibit its comparison with analogous systems in other societies.
Pariah status or “untouchability” is another special variety of caste organization; in this case structurally rather than regionally or culturally specific. It refers to the intrinsically polluted, stigmatized, denigrated, and excluded caste status found in many societies. It has been reported not only in south Asia but also in Japan, Korea, Tibet, preEuropean Africa, contemporary South Africa, and the United States. Pariah castes and the relations between pariah and nonpariah castes are recurrent and interesting phenomena within the category of caste organization. But care must be exercised not to attribute the characteristics of these phenomena to all caste systems or to assume that they are diagnostic of caste organization. The general or universal characteristics of pariahs and their relations with nonpariahs may be quite different from—or at least may be only part of the picture when compared with—the characteristics and relations of castes not so widely separated in the hierarchy, such as the many intermediate castes and castes close to one another in the hierarchy in India. Pariah–elite relations highlight many of the general characteristics of caste systems, but they neither define such systems nor exhaust their range, and it is erroneous to assume even by implication that they do.
Perhaps dual systems, such as that of Negro–white relations in the southern United States and of eta–noneta relations in contemporary Japan, are sufficiently different from multiple caste systems, such as that in India, as to deserve to be treated as two major subtypes of caste organization. It may be, for example, that abolition of the system is the characteristic goal of ambitious low-status groups in dual caste systems (where they have nothing to lose but their inferiority), while upward mobility within the system is the aim of low-status groups in multiple caste societies (where, in the event of abolition, equality with superiors would be bought with the loss of superiority to inferiors).
An advantage of the comparative approach to the study of caste systems is that variety as well as consistency in their characteristics can be defined and studied and the consequences can be analyzed. The general concept of caste must not blind us to distinctive types of caste systems, just as it must not lead us to overlook or deny unique attributes of particular systems.
Concomitants of caste systems
The utility of a comparative concept of caste lies in the light it can throw on the sociocultural and psychological concomitants of a definable, recurrent type of social structure in the context of general social processes, universal human nature, and variant cultural and historical environments. That is, in addition to being a type of structure, a caste system is a peculiar pattern of human relationships and a peculiar state of mind. Some of the concomitants of caste organization can be suggested briefly (they have been reviewed in greater detail in Berreman 1966).
There appear to be common and distinctive patterns of life in caste systems which make it possible to speak of “cultures of caste organization.” The imposition of birth-ascribed and unalterable membership in ranked, mutually isolated, but interacting groups with conspicuously different life experiences, life chances, and public esteem seems to have common psychological, attitudinal, and behavioral consequences wherever it occurs. Many of the common elements of caste cultures are doubtless characteristic of any sharply stratified society; others may be characteristic of any plural society; still others of any society with important inherited statuses. Caste cultures may be expected to exhibit a characteristic combination of elements from these three sources.
The consequences of caste systems surely include many of the responses, reactions, and mechanisms described in the literature on Negro-white relations in the United States, on touchable-untouchable relations in India, and on relations between rulers and ruled in colonial contexts. The similarities in these relationships is remarkable in view of the widely differing cultural environments in which they occur. Although members of any particular caste do not characteristically adopt a view of themselves entirely consistent with the view held by others in their society, they cannot avoid being influenced by those views and the associated behaviors. Common adjustments to low-caste status include avoidance, apathy, withdrawal, and over-compliance, as well as mobility, escape, and resistance, both passive and active. Self-justification, self-deception, rationalization, and fantasy serve both low and high castes. Those near the top of the hierarchy exhibit feelings of superiority relative to those beneath them. Prejudice, as it has been described and analyzed in American race relations, is a manifestation of this. Other attitudes and associated behaviors can be suggested by such terms as paternalism, noblesse oblige, condescension, segregation, discrimination, and exploitation.
The most striking examples of similarities between caste systems are found in social relations— in patterns of interaction. If an interactional definition of caste is used, these relations may be regarded as defining features; if a structural definition is used, they are concomitants. To maintain the sharp boundaries, the hierarchical ranking, and the power relations among castes, there are numerous rules restricting interaction between them. Most commonly these take the form of restrictions on marriage, on sex relations, on living together, on eating together, on sitting together, and on a variety of other forms of interaction symbolic of social equality. What constitutes status-equal interaction varies from society to society. What does not vary is the fact that some kinds of interaction and some kinds of behavior are defined as appropriate only between those who are equal in status (that is, within the caste), while others are defined as appropriate for superiors vis-à-vis inferiors and vice versa (that is, between castes). Each caste system, dominated as it is by high castes, has controls of variable effectiveness to assure that the rules are followed and the system is not jeopardized.
A closely related area is that of the processes of sociocultural change. Within caste systems there is constant mobility striving. It is generally sought through “status emulation” as groups attempt to imitate their social superiors (for example, “Sanskritization” in India). Another means to mobility is the adoption of reference groups wherein traditional considerations of status are irrelevant and new criteria are operative (for example, “Westernization” in India). Mobility striving, while intrinsic in caste systems, is a constant threat to the status quo. It is suppressed whenever possible, but the process of suppression is difficult and never completely effective. Caste systems are characterized not by consensus but by conformity. They are maintained not by agreement but by sanctions. It takes much physical and psychic energy to maintain an inherently unstable, conflictive situation in a semblance of working order. The dominant high-status groups must suppress mobility striving among others; rules restricting social interaction must be enforced; the purity and integrity of the group must be maintained; a myth of stability must be supported in the face of overt disconfirming evidence. On the part of low-status people, self-respect must be maintained despite constant denigration; resentment must be suppressed or carefully channeled. The social costs of such systems may be seen in in-group and intergroup conflict and violence. They might also be discovered in individual psychological disturbances. Certainly they are found in exploitative economic relations; in the waste of human resources resulting from discriminatory selection of occupational and other specialists; and in the distorted and inefficient distribution of goods, services, and opportunities. The social costs become manifest when the traditional hierarchy of power and privilege is confronted by democracy and equalitarianism and when pluralism is confronted by the homogenizing influence of mass media, public education, and the like. The resulting changes may take the form of group or individual mobility, of implementation of equalitarian practices, of elimination of traditional prerogatives, of wider dissemination of power and access to valued things in the society. Such changes are likely to be traumatic in their achievement. The fact that they are sought by some in the society and bitterly resisted by others belies the notion that caste systems are intrinsically consensus-based, equilibrium-maintained, personally satisfying, and conflictless.
Caste organization may seem to be an anachronism in the modern world. If so, it is at least an extremely important and quite widespread one whose human implications are becoming fully apparent as a result of the fact that such organization is being vigorously challenged and vigorously defended in many parts of the world. Unless it is understood, the problems it evokes are unlikely to be resolved. Recognition and understanding of it in its varied manifestations thus take on acutely practical significance in addition to theoretical importance.
Gerald D. Berreman
Bailey, F. G. 1959 For a Sociology of India? Contributions to Indian Sociology 3:88–101.
Bailey, F. G. 1963 Closed Social Stratification in India. Archives européennes de sociologie 4:107–124.
Barth, Fredrik 1960 The System of Social Stratification in Swat, North Pakistan. Pages 113–146 in Edmund R. Leach (editor), Aspects of Caste in South India, Ceylon and North-west Pakistan. Cambridge Papers in Social Anthropology, No. 2. Cambridge Univ. Press.
Berreman, Gerald D. 1960 Caste in India and the United States. American Journal of Sociology 66:120–127. → See the subsequent exchange of letters with O. C. Cox in the American Journal of Sociology (1961) 66:510−512, and the exchange with L. Dumont in Contributions to Indian Sociology (1962) 6:122–125.
Berreman, Gerald D. 1965 The Study of Caste Ranking in India. Southwestern Journal of Anthropology 21:115–129. → Comprises a critical review of the literature.
Berreman, Gerald D. 1966 Caste in Cross-cultural Perspective. Pages 275–304 in George DeVos and H. Wagatsuma (editors), Japan’s Invisible Race. Berkeley: Univ. of California Press. → See also the other chapters that describe and analyze pariah status in Japan.
Cox, Oliver C. 1945 Race and Caste: A Distinction. American Journal of Sociology 50:360–368.
Cox, Oliver C. (1948) 1959 Caste, Class & Race: A Study in Social Dynamics. New York: Monthly Review Press.
Dollard, John (1937) 1957 Caste and Class in a Southern Town. 3d ed. Garden City, N.Y.: Doubleday.
Dumont, Louis (1960) 1961 Caste, Racism and “Stratification”: Reflections of a Social Anthropologist. Contributions to Indian Sociology 5:20–43. → An English version of a paper first published in French in Cahiers internationaux de sociologie.
Gould, Harold A. 1960 Castes, Outcastes, and the Sociology of Stratification. International Journal of Comparative Sociology 1:220–238.
Kroeber, A. L. 1930 Caste. Volume 3, pages 254–257 in Encyclopaedia of the Social Sciences. New York: Macmillan.
Leach, Edmund R. (editor) 1960 Aspects of Caste in South India, Ceylon and North-west Pakistan. Cambridge Papers in Social Anthropology, No. 2. Cambridge Univ. Press.
Marriott, McKim 1959 Interactional and Attributional Theories of Caste Ranking. Man in India 39:92–107.
Nadel, Siegfried F. 1954 Caste and Government in Primitive Society. Journal of the Anthropological Society of Bombay New Series 8:9–22.
Passin, Herbert 1955 Untouchability in the Far East. Monumenta nipponica 11, no. 3:27–47.
Pohlman, Edward W. 1952 Semantic Aspects of the Controversy Over Negro–White Caste in the United States. Social Forces 30:416–419.
Smith, Michael G. 1965 The Plural Society in the British West Indies. Berkeley: Univ. of California Press.
Weber, Max (1916–1917) 1946 India: The Brahman and the Castes. Pages 396–415 in From Max Weber: Essays in Sociology. Translated and edited by Hans H. Gerth and C. Wright Mills. New York: Oxford Univ. Press.
Weber, Max (1921) 1946 Class, Status, Party. Pages 180–195 in From Max Weber: Essays in Sociology. Translated and edited by Hans H. Gerth and C. Wright Mills. New York: Oxford Univ. Press. → First published as Part 3, Chapter 4 of Wirtschaft und Gesellschaft.
The term “caste” is used to designate each unit in the hierarchically arranged organic systems of closed groups to be found on the Indian subcontinent. Besides this, it has been applied to the classical division of Hindu society and to systems of ranked and closed populations found outside India.
The Indian caste system presents an extraordinarily complex social phenomenon. The great size and spatial extension of the population concerned and the close interlocking of religious and secular features have produced a luxuriance of local variation from which it is difficult to draw consistent features and give them their precise emphasis. Social scientists have been interested in the caste system as a type of social stratification. But for a long time their knowledge rested on the largely descriptive accounts of writers interested in providing general ethnographies, on official documents, such as censuses, and on the Hindu literary classics. It is only in recent years that detailed studies of local caste hierarchies have been made. Hence, the relation of the Indian caste system to other systems of stratification and social grouping has by no means been agreed upon, and it is debatable whether castes exist outside India and its immediate neighbors.
Units of very different scale have been denoted by the word “caste,” as well as by vernacular terms, of which jati is the most common. Such units include reference categories extending throughout India, hereditary occupational units (such as potters, barbers, tanners), and the endogamous units within the occupational units. The terms “varna” “caste,” and “subcaste” will be used to distinguish these three broad levels, with a further distinction, where necessary, between the total population subsumed under the term and the operational local group.
Caste and “varna.” The earliest written mention of division in Indian society refers to the distinction between the autochthonous Dasa and the immigrant Arya populations. Later texts specify a threefold and then a fourfold division of society into Brahmana (priestly), Rajanya or Kshatriya (warrior–ruler), Vaishya (merchant), and Shudra (servant) varna, with the population outside this scheme being subsequently categorized as Untouchable. The varna formed a hierarchy marked by differing material and spiritual privileges. Little is known about the internal structure of the varna, but individuals could change their varna membership.
The varna now form categories that have preserved their traditional hierarchical positions within Hindu society. They are not organized groups, although some attempts have recently been made to unite the people of a varna for political purposes. Because there is no formal mechanism for dealing with claims to varna membership, the varna provide at the local level a means for groups (although not for individuals) to gain a traditionbased validation of an achieved status. Varna membership enables people from distant parts of India to assess each other’s approximate local status. Varna are socially significant, therefore, as a charter for hierarchical status and as a unifying feature of the system.
Caste and subcaste. Members of Indian caste society normally become members of one of its constituent units at birth. These units are local populations within which there is an effective implementation of the rules of caste behavior. Sometimes formal boundaries designate units that contain populations of between five and one hundred villages. But in other cases the size of the unit depends on the composition of informal gatherings that exert social control; since these are mainly of kin, such units are to some extent defined by the area and by the pattern of marriage ties of a family or local group. Units generally form part of a wider population bearing the same name and, at least in theory, following the same customs and allowing the intermarriage of members. Easier means of communication now make it possible for these larger units to regulate the behavior of members, at least in major matters. Nevertheless, variations often exist between the smaller local units, and minor differences can sometimes be observed even between nearby villages. The smaller, locally effective units can be called “subcaste groups,” and the larger, named populations can be called “sub-castes.”
The subcaste group (which, of course, may compose the whole of a small or well-organized subcaste), besides controlling internal relations, regulates the behavior of members toward people of other subcaste groups in the area. Some of these may be subcaste groups of the same caste; others belong to different castes. Relations between two subcaste groups of the same caste may be close and may include intermarriage if the two sub-castes are moving toward an amalgamation. In such transitional cases only the different name may distinguish the subcastes. More often, however, their relations are marked by endogamy or hypergamy and by many of the separations that characterize relations between the castes to which the subcastes belong.
Most castes contain a varying number of sub-castes; in addition, a few are unitary and others have one or more intermediate sections composed of several subcastes. Caste membership seldom involves the regulation of internal relationships, for these are taken care of by the subcastes. Nevertheless, caste membership may influence behavior between people of different castes. The subcastes of a caste bear a common caste name, and they are not usually differentiated by outsiders, who therefore act in the same way toward them all, seeing them as parts of a single unit. This unit has been called a “caste-cluster” by scholars who wish to call the subcaste the “caste” and so stress the fact that sub-castes are the basic units of action, which may have evolved independently and not from the fission of a larger unit (Karvé 1961). No single terminology is at present in use; other scholars designate the subcaste group by the vernacular word jati, a term the people themselves apply to all levels of the caste system.
Relations between members of different castes occur mainly at the local level, and whole caste populations seldom take concerted action. The same distinction can therefore be made between the caste group and the caste as has been drawn between the subcaste group and the subcaste. Until recently, analyses of the caste system have rested upon studies of relations between caste groups.
The caste hierarchy. The relations between caste groups (and, mutatis mutandis, between subcaste groups) can be characterized as hierarchical. They are based on evaluations of differences into which both religious and more mundane considerations enter. The belief in a differential innate purity of each caste is foremost in this evaluation. Purity is ascribed to caste members at birth and rewards them for the quality of their actions in their previous life. Thus, a belief in metempsychosis serves as an incentive for an individual to better his future rebirth. One way for him to achieve this is by adhering to the rules of society, in particular to those of his caste group as enforced by his subcaste group’s council (panchayat).
Foremost among such rules are those that safeguard the level of caste group purity. Its pollution results from contact either with lower caste groups or with objects that are themselves impure. Contact with the former is avoided by restrictions on intermarriage and sexual relations with caste groups of less purity, sometimes on personal touch and approach, and on commensal relations, such as eating, smoking, and drinking. Impure materials include things having to do with the dead and with bodily emissions; hence, the practitioners of occupations connected with them (such as barbering and leatherworking) have a low status. This is especially important because castes are linked to hereditary occupations, whose degree of purity affects the members whether or not they are practicing these occupations. Members who become polluted involuntarily (through contact with the dead at funerals, through menstruation, etc.) undergo purificatory rites of varying complexity and, until these are completed, must stay apart from their caste group fellows. Those who pollute themselves by knowingly breaking caste group rules (by eating or having sexual relations with lower caste groups or by carrying on demeaning occupations) must also be purified and are barred from social intercourse until purification has taken place. The manner of their purification is decided by the subcaste group council, which also may determine the size of any fine that accompanies purification. There is a ceremonial meal at which a formal resumption of commensal relations readmits those who have been polluted to the subcaste group (and, in the eyes of outsiders, to the caste group). If they refuse to reform their behavior, they are excommunicated. An individual may then be recognized as a member of the lower caste group with which he has associated himself; a seceding local group will tend to form a separate subcaste group.
It is easy to see considerable variation of behavior at each end of the hierarchy and to recognize hierarchical differences. At the top the Brahman stands in an acknowledged position of purity and fulfills priestly duties, and at the bottom are caste groups that are traditionally associated with such occupations as scavenging and tanning and whose diet may contain beef (meat is more polluting than vegetables and beef is more polluting than other meats). But not all ranking is so easily distinguishable. Many caste groups have agriculture as their traditional occupation and follow similar dietary and other customs. Yet these groups may also claim to be higher than one another. In such cases, other factors may assist society to rank them. Land ownership and the wealth it brings, education, and hereditary positions of authority are attributes that may influence the assessment of rank and at the same time make it possible for others to distinguish ranks when dealing with two caste groups claiming equal status. Thus, the different economic and political positions of the same caste in different areas may account for variations in its hierarchical status.
These features are also often the mainsprings of mobility in the system. Although an individual cannot improve his caste affiliation, a caste group (or, more often, a subcaste group) can try to move upward by copying the behavior of either the ritually preeminent Brahman or of the castes that are economically and politically powerful. Thus, if the model group considers marriage to be a binding sacrament, an aspiring subcaste group will ban the remarriage of its widows. Again, many politically dominant castes do not object to alcohol, and their imitators also will not do so. To help it rise, a subcaste group may change its name and lay claim to a long-lost membership in a higher varna. This helps it sever relations with other subcaste groups in the caste in order to be recognized by outsiders as a separate, higher caste group. The impetus for such changes is frequently given by recently acquired wealth or political influence. However, success can only be gained slowly, for the traditional associations of the caste remain. Its members may no longer perform the traditional work, but their name will announce their former connection with it, and even after a change of name they must overcome tenaciously held memories. Modern conditions have made it possible for subcaste groups that have acquired economic and political power to disregard hierarchical status by leaning on the formally casteless institutions of the state. But a validation of power is still often sought within the caste system.
Theories of caste. A number of theories about caste devote themselves to explaining its origin. These include the hypothesis that the system was created by the Brahmans for their own benefit (Sherring 1872–1881, vol. 3, p. 231) and the classical view (Manu, chapter 10) that castes developed from unions between members of different varna. It has also been suggested that castes were formed on “a community of function” through common occupation in a division of labor (Nesfield 1885, p. 88). An alternate theory claims that the underlying principle was a physical antipathy of Arya for Dasa, resulting in an endogamy that produced measurable physical distinctions, so that one could almost say for at least certain regions of India that “a man’s social status varies in inverse ratio to the width of his nose” (Risley 1891, vol. 1, p. xxxiv). Hocart ( 1950, p. 68) suggests that the functions and concomitant purity of participants in court rituals became hereditary, and when this organization later spread to meet the ritual requirements of the rest of the population a ritually ranked hierarchy was created. Such theories tend to center upon a single factor as basic to the origin of the caste system. Others are less sweeping, such as that which maintains that the system arose from Aryan institutions that were adapted to the conditions found in India (Senart  1930, p. 213). Hutton ( 1964, p. 164) is even less ambitious, giving only a list of 15 factors whose concatenation (perhaps with others) contributed to the emergence of the system, and Weber ( 1958, pp. 130–131) suggests that the institution could have been produced only by the convergence of several major factors.
Whatever the validity of such theories, it is clear that many of the features they stress are basic to caste. These include the emphasis on hereditary occupation, with its resulting functional interdependence and lack of competition, and the opposition of purity and impurity. Such features are also prominent in studies of present-day caste. These studies deal with caste groups as units linked to one another in a system, rather than with individual caste groups, and are based on analyses of the qualities of these connections. Many years ago Bougié (1908) distinguished the basic principles of hierarchy, hereditary specialization, and repulsion (better expressed as separation); these are still considered to be primary, with intercaste links characterized as organic and as containing some degree of summation of roles. It is, however, possible to emphasize as a fundamental feature of caste either the opposition of purity and impurity or an economic and political aspect that may then be expressed in ritual terms. The latter approach stresses that caste groups that do not follow the precepts of pure behavior may nevertheless have high rank. Meat-eating and liquor-drinking groups may, in some local hierarchies, find themselves placed above vegetarian groups. In such cases rank may depend less on the pure or impure attributes of the caste group’s behavior than on the dominance it has in its economic and political interaction with other caste groups. On the other hand, it can be maintained that to stress the political and economic aspects at the expense of other elements of caste behavior is to ignore the way in which caste is regarded by the people themselves. Rather, notions of purity and impurity lie behind a hierarchic principle that is distinct from, although not unconnected to, the economic and political power structure and that gives the caste system its nature.
The different weight scholars give to these approaches has a bearing on the extent to which caste is viewed as a general social phenomenon existing outside of India. For a definition in political and economic terms is more easily applied to hierarchical situations in other societies, whereas that resting on a particular form of purity-impurity opposition is more likely to be restricted to the Hindu situation.
Present and future research. One focus of present research has been on the problem of caste rank. This arises where the criteria of rank conflict with one another. Such a conflict may lead the caste groups involved into a relation of equal separation rather than one of differential rank, but others in the community may weight the various factors so as to produce a generally acknowledged rank order. The problem consists in finding out what these weightings are. Attempts have been made to do this through the analysis of different interactional contexts of ranking, especially the commensal (Mayer 1956), and, recently, efforts have been made to assess a collective hierarchy through the statistical analysis of village opinion about ranking rather than through observed behavior (Freed 1963). At the same time, there is study of the factors underlying the number of ranks (the elaboration of the hierarchy) in different places (Marriott 1960). It is hoped that this work will lead to a greater knowledge about the determinants of rank and provide data against which hierarchical change can be measured.
Related research is concerned with the scales of power and prestige underlying caste hierarchies. Studies have been made of the system of socioeconomic services (jajmani) to evaluate its place in intercaste group relations. Here, artisans and other specialists render economic and social services to clients, for which those having land pay in kind at each harvest. Such ties between the families of client and specialist may continue for generations and may be treated as heritable property by the specialist. The system has been viewed by some as providing for the interdependence and social cohesion of the local group (Wiser 1936); others have stressed that greater power lies with the landholder, who is assured constantly available services (Beidelman 1959). The jajmani system is related to the caste hierarchy because many of its occupations are the monopolies of specific castes and because the major landholders usually belong to higher caste groups. Such caste groups are dominant politically as well as economically, and this has led to the exploration of the connections between economic and political power, through analysis of local political systems and discussion of the extent to which class and party interests link people of different castes and internally differentiate caste and subcaste groups. Such studies enlarge our knowledge of the patterns of summation of roles in intercaste group relations. The relation of caste to class and party is especially important for the study of caste in towns, where there may be a low summation of roles and a weak caste organization, part of whose place may have been taken by other associations (for example, trade unions).
Yet another series of problems arises from the study of the internal constitution of caste units. The extent of the effective subcaste group has by no means been fully ascertained. Where there are formal councils with authority over the subcaste populations of definite numbers of settlements, the boundaries of social control are easy to define. But it is difficult to delimit subcaste groups where control is exercised at ad hoc meetings of kin. Caste and kinship have, until recently, been largely considered as separate social fields. It is now clear that the two are closely linked; some work has explored this connection (for example, Dumont 1957; Mayer 1960; Orenstein 1965), but it has yet to be fully analyzed.
A last set of problems stems from the role of caste in independent India (see, for example, Srinivas 1962). New fields for caste distinctions and rivalries have been provided by the introduction of universal franchise, semiautonomous community development committees, and legislation against untouchability and by the augmentation of benefits for castes classed as backward. The lower castes and others not traditionally dominant may now have the chance to gain local and regional political power. At the same time, occupational specialization is tending to diminish, with castes being economically linked more through competition than through functional interdependence. There are many newly emerging social questions to be studied, such as the degree to which these changes have affected patterns of rank, the extent to which the caste system is still a factor of social organization, and the extent to which the new situation, added to a greatly increased ease of communication, has stimulated larger caste groupings in the form of regional and national caste associations. Some study has been made of caste associations (Rudolph & Rudolph 1960) and of the role of caste in local, regional, and national politics (Brass 1965; Harrison 1960, chapter 4), but much more needs to be done before a detailed assessment of caste’s new role can be made and, indeed, before it can be decided whether these associations should be classified as caste bodies.It has been suggested that they are interest groups based on caste membership rather than manifestations of caste, since they form competing units in a segmentary system rather than organically interdependent parts (Bailey 1963, p. 123). Social scientists may therefore be faced with the problem either of redefining caste to include situations where caste-named groups act in qualitatively new ways or of adopting a new term for these groups.
Each of these sets of problems is devoted to providing a deeper understanding of caste in specific situations. Questions of technique arise where these situations are spatially extended (as in the study of the boundaries of the subcaste group) or where they involve single-interest roles rather than those with a high degree of summation. Therefore, like their colleagues working in other “complex” societies, students of caste in India are concerned with the problem of adapting their methods and concepts to new situations. Further problems exist for those who wish to make general statements about the place of caste in Indian civilization, notably the problem that concerns the usefulness of traditional literary and historical sources not only for the question of the origin of caste but also for the ways in which they can enlarge the temporal and spatial framework of sociological inquiry.
Caste outside India. The term “caste” has been applied to social strata in a number of societies outside India. There is disagreement, however, over whether it should be used in this way or retained to define the pan-Indian phenomenon. Supporting the latter view are those who see caste as a form of structural organization specific to India and, at most, castelike situations in Pakistan, Nepal, and Ceylon (and possibly overseas Indian communities). Others stress the difference between what can be called de facto and de jure stratification (Nadel 1954) and maintain that a definition of caste must take into account the values of the society and the meaning given to the term by the people themselves. According to this view, caste systems are based on de jure stratification, which is qualitatively different from stratification in societies in which equality is the ideal relationship (Dumont 1960). Caste systems are therefore to be found in relatively few societies outside India, and even then with little elaboration. On the other side, the view is put forward that institutions should be defined so as to facilitate comparison. For instance, Berreman (1960) shows that a limited definition of caste as a hierarchy of closed divisions can lead to a comparison between India and the southern United States. However, the need to limit the definition of caste for comparative purposes suggests that Indian caste has features only partially duplicated in other societies. A comparative approach to stratification, although sociologically valuable, should thus be distinguished from studies of the operation of Indian caste and of its relation to its pan-Indian context.
Adrian C. Mayer
Bailey, Frederick G. (1957) 1958 Caste and the Economic Frontier: A Village in Highland Orissa. Manchester Univ. Press.; New York: Humanities.
Bailey, Frederick G. 1963 Closed Social Stratification in India. European Journal of Sociology 4:107–124.
Beidelman, Thomas O. 1959 A Comparative Analysis of the Jajmani System. Monographs of the Association for Asian Studies, No. 8. Locust Valley, N.Y.: Augustin.
Berreman, Gerald D. 1960 Caste in India and the United States. American Journal of Sociology 66:120–127.
Blunt, Edward A. H. 1931 The Caste System of Northern India: With Special Reference to the United Provinces of Agra and Oudh. Oxford Univ. Press.
BouglÉ, Celestin (1908) 1935 Essais sur le régime des castes. 3d ed. Paris: Alcan.
Caste: A Trend Report and Bibliography. 1959 Current Sociology 8, No. 3:135–183.
Dumont, Louis 1957 Une sous-caste de I’Inde du Sud: Organisation sociale et religion des Pramalai Kallar. Paris: Mouton.
Dumont, Louis (1960) 1961 Caste, Racism and “Stratification”: Reflections of a Social Anthropologist. Contributions to Indian Sociology 5:20–43. → First published in French in Volume 29 of Cahiers internationaux de sociologie.
Freed, Stanley A. 1963 An Objective Method for Determining the Collective Caste Hierarchy of an Indian Village. American Anthropologist New Series 65:879–891.
Ghurye, Govind S. (1932) 1961 Caste, Class and Occupation. 4th rev. ed. Bombay: Popular Book Depot; New York: Heinman. → First published as Caste and Race in India.
Harrison, Selig S. 1960 India: The Most Dangerous Decades. Princeton Univ. Press.
Hocart, Arthur M. (1938) 1950 Caste: A Comparative Study. London: Methuen. → First published in French as Les castes.
Hutton, John H. (1946) 1964 Caste in India: Its Nature, Function and Origins. 4th ed. Oxford Univ. Press.
KarvÉ, Irawati 1961 Hindu Society: An Interpretation. Poona (India): Deccan College.
Leach, Edmund R. (editor) 1960 Aspects of Caste in South India, Ceylon and North-west Pakistan. Cambridge Papers in Social Anthropology, No. 2. Cambridge Univ. Press.
Marriott, McKim 1959 Interactional and Attributional Theories of Caste Ranking. Man in India 39:92–107.
Marriott, McKim 1960 Caste Ranking and Community Structure in Five Regions of India and Pakistan. Deccan College Monograph Series, No. 23. Poona (India): Deccan College.
Mayer, Adrian C. 1956 Some Hierarchical Aspects of Caste. Southwestern Journal of Anthropology 12:117–144.
Mayer, Adrian C. 1960 Caste and Kinship in Central India: A Village and Its Region. London: Routledge; Berkeley and Los Angeles: Univ. of California Press.
Nadel, Siegfried F. 1954 Caste and Government in Primitive Society. Journal of the Anthropological Society of Bombay New Series 8:9–22.
Nesfield, John C. 1885 A Brief View of the Caste System of the North-western Provinces & Oudh. Allahabad (India): Government Press.
Orenstein, Henry 1965 Gaon: Conflict and Cohesion in an Indian Village. Princeton Univ. Press.
Risley, Herbert H. 1891 The Tribes and Castes of Bengal: Ethnographic Glossary. 2 vols. Calcutta: Bengal Secretariat Press.
Rudolph, Lloyd I., and Rudolph, Susanne H. 1960 The Political Role of India’s Caste Associations. Pacific Affairs 33:5–22.
Senart, Émile (1896) 1930 Caste in India: The Facts and the System. Translated by E. Denison Ross. London: Methuen. → First published in French.
Sherring, Matthew A. 1872–1881 Hindu Tribes and Castes. 3 vols. Calcutta: Thacker; London: Trubner.
Srinivas, Mysore N. 1962 Caste in Modern India, and Other Essays. New York: Asia Pub. House.
Stevenson, H. N. C. 1954 Status Evaluation in the Hindu Caste System. Journal of the Royal Anthropological Institute of Great Britain and Ireland 84:45–65.
Weber, Max (1921) 1958 The Religion of India: The Sociology of Hinduism and Buddhism. Translated and edited by Hans H. Gerth and Don Martindale. Glencoe, III.: Free Press. → First published in German as Hinduismus und Buddhismus in Volume 2 of Gesammelte Aufsütze zur Religionssoziologie.
Wiser, William H. 1936 The Hindu Jajmani System: A Socio-economic System Interrelating Members of a Hindu Village in Services. Lucknow (India) Pub. House.
Nearly all societies have had some form of social stratification, whether ascriptive or achieved, based on race, class, religion, ethnicity, language, education, or occupation. The Hindu ascriptive caste system in India is perhaps the most complex and rigid. It is based on birth, which determines one’s occupation (especially in contemporary rural India), and is maintained by endogamy, commensality, rituals, dietary practices, and norms of purity and pollution. The English term caste is derived from the Portuguese word casta, which refers to lineage, breed, or race.
The Hindu caste system is interpreted in two ways. First is the standard varna description of the caste system as a fourfold division of Hindu society. Its origins are noted in the Rig Veda, one of the sacred texts of Hinduism, which dates back some 3,000 years. The term varna, which literally means color, does not reference the actual racial features of those who fall into the four varnas, as many scholars have established. Brahmins constitute the sacerdotal order, which has priestly duties including the interpretation of numerous complex religious texts in Sanskrit, a language that traditionally only they mastered. Below them are the Kshatriyas, the caste of warriors and rulers. They are followed by the Vaisya, who typically engage in trade and commerce. The bottom is the Shudra, or peasant and laborer caste, a large and diverse group that comprises artisans ranging from goldsmiths to washermen and peasants who may own sizeable tracts of land. Outside the fourfold system are the “untouchables,” now commonly referred to as Dalits (“the oppressed”), who perform the most menial tasks. Normally, women in the top three varnas do not pursue the hereditary occupations, whereas Shudra and Dalit caste women do. The vast majority of India, which is rural, is caste-based as far as inheritance of occupations is concerned.
The hierarchy among various castes is further based on the notions of ritual purity and pollution. The higher the caste, the greater the purity of the group, while lower caste status is associated with pollution. Moreover, the nature of the occupation that one is born into also confers purity or pollution. For example those who dig graves, clean latrines, sweep streets, or work with leather are more polluting than those in “clean” occupations such as trade or priesthood. Those usually engaged in polluting occupations must maintain a prescribed distance from the castes deemed pure to avoid polluting them through contact. For this reason, it is not uncommon for lower caste groups to live in segregated colonies on the outskirts of Indian villages even today.
M. N. Srinivas (1962) and André Béteille (1996), among others, consider the varna description simplistic because it does not represent the empirical reality of caste in either ancient or modern India. Their interpretation of the caste system as a constellation of more than 3,500 jatis with internal and regional variations has gained validity among scholars. Whenever ordinary Indians refer to caste, they are referring to jati, not to varna. In the words of Béteille, “whereas varna refers primarily to order and classification, the primary reference of jati is to birth and the social identity ascribed by birth” (p. 22). In most sociological analyses (and here) the term caste is used to represent its jati dimension.
Although the caste system has eroded to some extent, it still has a hold in contemporary Indian society. One factor in this has been the Indian constitution, which empowers the state to make special provisions for the advancement of low-caste citizens, including the more than 160 million Dalits (who are listed in a schedule attached to the constitution and thus called “Scheduled Castes”), the nearly 50 million tribals (also listed in a schedule and hence called the “Scheduled Tribes”), and the 500 million “Other Backward Classes.” Under these provisions 15 percent of government jobs and university places have been reserved for members of the Scheduled Castes, 7 percent for Scheduled Tribes, and 27 percent for the Other Backward Classes. Initially, this affirmative action was to remain in place only until these marginalized castes caught up with the more privileged upper castes, but this has not yet happened, and there are strong pressures not only to extend indefinitely these policies but also to include caste groups that traditionally belong to Shudra status. The constitution also provides for the reservation of electoral seats for Scheduled Castes in the parliament of India and all the state legislatures. Similar rules govern elections for village and district councils. Although these measures are necessary to create a level playing field for historically deprived caste groups, they also go against another constitutional objective—the elimination of discrimination based on caste.
The affirmative action measures have empowered Dalits to some extent. In 1997, a Dalit, K. R. Narayanan, became the president of India, and by 2001 more than 13 percent of senior bureaucrats in the government of India were Dalits (Gupta 2001, p. 13). However, such gains are overshadowed by the stubborn continuation of inequalities. Dalits and Scheduled Tribes, particularly women, are at the bottom of the economic ladder (Deshpande 2002). Dalits in rural India are still forced into indentured farm labor. In some parts of the country there is still strong opposition to them owning land and sharing public facilities such as temples and wells.
A second factor that has enabled the caste system to flourish is its function as a “vote bank.” More often than not, elections are fought not so much over political ideologies and programs but on the caste identity of the contestants. Virtually all castes have well-organized and well-funded associations that mobilize voters for their caste’s candidates. Although such mobilization enhances political awareness and participation by various castes, at the same time it also undermines efforts to create a casteless society.
A third factor is that at the individual level, caste identity is still hereditary. Unlike Christianity and Islam, Hinduism does not approve proselytization, and it has no ecclesiastical order. Effectively, then, caste functions as the church of Hinduism, operating with centuries-old customs, norms, and values. Given these conditions, there are no recognized means by which, for example, a Dalit can move up the ritual hierarchy, or a Brahmin move down. However, following India’s independence in 1947 and adoption of a constitution that stipulated creation of a secular society, various mechanisms have evolved that have enabled members of lower castes, as a group, to claim superior social status when they emulate the customs, rituals, and way of life of upper castes (Srinivas 1962; Shah 2005). This process is called Sanskritization. For example, an individual (or a group) belonging to Shudra jati may become vegetarian, worship the gods that upper castes worship, and even recite Sanskritic hymns as part of its regular prayers, thus claiming status mobility (but not mobility in the ritual hierarchy). Sanskritization is an informal and voluntary process that does not involve participantsʾ merging their identities with the caste whose way of life they imitate, nor will the higher jati welcome them to its fold just because they adopted their ways. Sanskritization is most effective when it occurs at the group level. However, the basic nature of ascription continues. For example, an African American can earn high status in terms of his accomplishments, but his ascriptive status remains unchanged—he is not white. Likewise, an untouchable in India can rise to the position of president of India, but he is still characterized in the media as first untouchable to become president: his caste identity precedes his accomplishment.
A fourth factor that facilitates the continuity of caste is endogamy. The vast majority of marriages in India are still arranged by elders who ensure that their children marry from their own caste. However, a recent report in a south Indian newspaper titled “An Arranged Love Marriage” refers to the flexibility that is emerging in arranged marriages (Deccan Herald ). Such marriages are becoming common among the urban middle class when future partners who belong to same caste (and perhaps, class too) meet at work and go on “dates.” When they find they are mutually suitable, the couple seek the consent of their parents who are more than willing to bless the union since it liberates them from the hassle of dowry negotiations, etc. And yet, this is still an urban phenomenon occurring only among a minority of those who are in the marriage market.
Some aspects of the traditional caste system are changing, especially in urban areas. First, the inheritance of occupations by birth is no longer common except among the Dalits, especially in rural areas. More urban Dalits have succeeded in availing themselves of affirmative action measures. Furthermore, those of lower castes who live in urban areas have greater access to higher education, which makes them more competitive in the job market, especially in the private sector, where reservation policies are not applicable. At the time of this writing, the government of India has proposed to increase quotas for Old Backward Classes in all centrally funded institutions such as the Indian Institutes of Management, Indian Institutes of Technology, and others. This move has resulted in a public debate about the continuing centrality of caste in admission policies which in the end works against value of merit (for details see India Today ).
Second, although restrictions based on purity and pollution continue to shape social distance and interactions between high and low castes in villages of India, where roughly 70 percent of Indiaʾs population resides, they are becoming increasingly hard to observe or enforce in large towns and cities. Although it is easy to identify a Dalit in a small village, it is not that easy to identify a Brahmin or a Dalit among, for example, public-transit passengers in a large city. As Indian society becomes increasingly modern, the norms of purity and pollution that are central to the traditional caste system are weakening.
Finally, traditionally caste-based dietary practices are eroding, especially in urban areas, due in part to the way in which nonvegetarian meals are packaged, especially in Western fast-food outlets. Many upper-caste Hindus may not cook meat in their homes, but they have few qualms about breaking dietary taboos in McDonaldʾs, Pizza Hut, or similar fast-food outlets, which are becoming fashionable dining places for urban upwardly mobile Indians.
Many of the religions that entered India during the last five to ten centuries targeted poor and marginalized castes for conversion, but becoming a Christian or Muslim did not accord converts a status free of caste. Instead, for the vast majority, their caste identities stayed with them, and their children and grandchildren have been unable to shed them. At the same time, the converts have been denied by many legal jurisdictions the constitutional benefits of Scheduled Caste and Scheduled Tribe status. Recent attempts by those who converted to Christianity from the lowest castes of Hinduism (Scheduled Castes) to claim Scheduled Caste status and thus eligibility for constitutionally guaranteed benefits have not been successful. The same failure has greeted attempts by individuals in some regions of the country whose status is in proximity to Scheduled Castes, such as dhobis (washermen) and chamars (leather workers). In fact, conversion of a Dalit, who is Hindu by definition, to Christianity has not been advantageous to many. It might have given them a sense of hope and status but it also deprived them of some privileges that their old status as oppressed Hindus, entitled by the constitution to certain privileges, afforded.
Features of caste continue even in those religions such as Sikhism that emerged in protest against the rigidity of Hindu rites and rituals and, more importantly, against the ascriptive caste system. Over time, social divisions resembling caste hierarchy became part of Sikh society as Sikhs strongly protected and promoted their identities as Jat, Mazabi, and Ramgarhia Sikhs, with claims to superiority over each other and norms of endogamy.
Buddhism arose around the sixth century, partly in protest against the Hindu caste system. Although its founder was a Kshatriya, the most likely candidates for conversion have come from lower castes. While Buddhism extended its influence beyond the shores of India, it was not a great success in India until the 1950s, when Dr. Babasaheb Ambedkar, a Dalit who gave independent India its constitution, encouraged fellow Dalits to convert to Buddhism; millions did, and some still continue to do so. However, conversion to Buddhism (just as in the case of Christianity) did not amount to renunciation of oneʾs caste identity. Although changing oneʾs religion may be an act of protest against the caste system, Buddhist converts are reluctant to renounce their caste identity because they still want to obtain the benefits that the Dalit status entitles them to under the Indian constitution—a situation which is contradictory because converts to Christianity are not entitled to similar privileges.
In the late nineteenth and early twentieth centuries Britain, the colonial ruler of India, encouraged Indians to migrate to its other colonies in the Caribbean and Africa as indentured laborers in its bid to maintain its economic success. Most of the Indians who chose to migrate were members of lower castes who saw migration as an opportunity for upward mobility. Although the first generation of immigrants tended to retain their caste identities, particularly in matters of marriage and religious rites (Schwartz 1967), subsequent generations did not, because of the assimilative nature of economic, political, and juridical forces (Motwani, Gosine, and Barot-Motwani 1993; Gosine and Narine 1999). Caste cannot be easily transplanted to an environment where Hinduism is not the operative religion.
Systems of stratification comparable to the Indian caste system have been identified in other parts of the world. For example, in Nigeria the relations between the Ibo and Osu groups are similar to those of upper and lower castes. In Somalia a social group called Midgam or Madibhan suffers from all the impediments that Dalits experience: impurity, pollution, and social distance. The Burakumin of Japan have been compared with the Dalits, as they have faced similar restrictions. These restrictions were outlawed in 1871, but, as in the case of Dalits, discrimination continues, especially in matters of employment and marriage (Henshall 1999). These dichotomous divisions, however, do not come close to the intricate caste system. At best, they compare two opposite ends of caste system with another system similar to it, ignoring the middle, wherein lies the heart of caste system.
Even though there have been stout rejections of the claim that caste can be equated with race (see, among others, Gupta 2001) purely on the grounds of universal practices of discrimination based on ascription, scholars such as Gerald Berreman (1960; 1972) have attempted to compare American blacks to untouchable castes in India. However, the black-white dichotomous system in the United States differs from the fourfold caste system in India in that it is ordained not by religious considerations, but by economic and social ones (Cox 1948).
Nearly all societies are stratified in one way or another, and some groups will always be relegated to the margins. However, the Indian caste system is unique because of its complexity, its religious foundation, its hereditary occupational system, and its norms of endogamy. More important, caste has served to energize Indian polity because it has been a primary means of motivating and mobilizing citizens to take part in electoral politics. Perhaps that has been a positive aspect of caste in Indian society, but the time may come to look for other means of motivating the electorate in India.
SEE ALSO Caste, Anthropology of; Hierarchy; Inequality, Political; Segregation; Stratification
Berreman, Gerald. 1960. Caste in India and the United States. American Journal of Sociology 66 (2): 120–127.
Berreman, Gerald. 1972. Race, Caste, and Other Invidious Distinctions in Social Stratification. Race 13 (4): 385–414.
Béteille, André. 1996. Varna and Jati. Sociological Bulletin 45 (1): 15–28.
Cox, Oliver C. 1948. Caste, Class, and Race. New York: Monthly Review Press.
Deccan Herald. 2006. An Arranged Love Marriage. June 26.
Deshpande, Ashwini. 2002. Assets versus Autonomy? The Changing Face of the Gender-Caste Overlap in India. Feminist Economist 8 (2): 19–35.
Gosine, Mahin, and Dhanpaul Narine, eds. 1999. Sojourners to Settlers: Indian Migrants in the Caribbean and the Americas. Windsor, NJ: Windsor Press.
Gupta, Dipankar. 2001. Caste, Race, Politics. http://www.india-seminar.com.
Henshall, Kenneth G. 1999. Dimensions of Japanese Society: Gender, Margins, and Mainstream. New York: Palgrave Macmillan.
India Today. 2006. Casting for Votes. May 15.
Motwani, Jagat K., Mahin Gosine, and Jyoti Barot-Motwani, eds. 1993. Global Indian Diaspora: Yesterday, Today, and Tomorrow. New York: Global Organization of People of Indian Origin.
Schwartz, Barton. 1967. Caste in Overseas Indian Communities. Chicago: Chandler Publishing.
Shah, A. M. 2005. Sanskritization Revisited. Sociological Bulletin 54 (2): 31–39.
Srinivas, M. N. 1962. Varna and Caste. In Caste in Modern India and Other Essays, 63–69. Bombay: Asia Publishing House.
Srinivas, M. N. 1966. Sanskritization. In Social Change in Modern India, 1–45. Berkeley: University of California Press.
G. N. Ramu
The word caste probably comes from the Portuguese word casta, meaning "species" or "breed" in relation to botany and animal husbandry. It was first applied by the Portuguese to describe the predominant organizing principle of Indian society. The word subsumes two kinds of categorization. One categorization is religious, represented by the word varna, which means "color" in Sanskrit. According to the varna principle, Hindus are divided into four caste groups, together with a fifth group, the untouchables, that exists outside the caste system. The other categorization is by what is called jati, the endogamous, (that is, in-marrying) birth grouping that determines a person's social position and duties and primary nonfamilial allegiances. There are thousands of jatis, often with highly contested (and changing) rankings by varna–and even within a particular varna.
The origins of caste are debated but probably include a mixture of scriptural injunctions, ancient ideas about racial exclusivity, long-term occupational heredity, and colonial categorizations and impositions that converted local endogamous units into pan-Indian groupings.
Caste-like categorizations also exist in some other societies, although without the fine grades of classification found in India. The best-known example outside South Asia is that of the Burakumin ("village people"–social outcasts, known also by the more pejorative term Eta) in Japan, a group of about 2.5 million that has faced persistent barriers to social, economic, and marital integration into mainstream Japanese society based on their ancestry.
Institutionalization of Caste
So tenacious is the hold of caste (and the concepts of social ranking and exclusivity associated with it) that it is a continuing feature of the Indian diaspora even in the developed countries of the West–witness, for example, the caste details specified in the marriage advertisements in Indian publications in the United States. Caste categories are often applied also to non-Hindus of the Indian subcontinent. Sociologists have recorded the caste-consciousness of Muslim and Christian communities in India, the caste referring to that of the Hindu ancestors of these groups before they converted to Islam or Christianity. Such caste-consciousness restricts social intercourse and deters marriages across these ancestral caste lines.
This institutionalization of the social hierarchy implicit in caste rankings in Hinduism is politically important because of the commitment of the post-independence Indian government to a casteless society and to affirmative action to improve the situation of the lowest castes. Pan-Indian and regional caste loyalties have been exploited by both the upper and lower castes to press economic and political demands, increasing intercaste rivalries. In the process, the lower castes have become better able to organize and resist the authority of the upper castes.
There are also more direct demographic implications of caste endogamy and caste hierarchy. The widespread acceptance of caste rankings has meant that groups lower down in the hierarchy try to raise their status by adopting the practices of the higher castes–a phenomenon the sociologist M. N. Srinivas termed Sanskritization. Sanskritization is not modernization, though the latter usually accompanies the former. Instead, Sanskritization often entails copying the most traditional, oppressive, and insular habits of the upper castes and giving up of many of the social and, especially, gender equalities and freedoms that characterized the lower castes. In many respects caste is little different from class, and in most societies, as observers such as Mary Wollstonecraft, the eighteenth-century English feminist and writer, and others have pointed out, the upper classes have not been known for greater gender sensitivity.
Caste is an important marker of demographic outcomes. India's censuses and official surveys no longer collect information on caste as such (all censuses from 1872 to 1941 included some question on caste), but they do separate out the Scheduled Castes (SCs) and Scheduled Tribes (STs). The Scheduled Castes are the former untouchables, and the Scheduled Tribes are the non-Hindu tribal groups that have remained outside the Indian cultural mainstream. The numerous jatis that make up these two groups have been listed for the purposes of affirmative action. Many of the affirmative policies also apply to what are called the "Other Backward Castes" (OBCs), a mixture of the lowest castes in the fourfold varna system and those untouchable groups that have converted to other religions. Together, SCs (about 19% of the population), STs (about 9%), and OBCs (about 32%) account for some 60 percent of the total population of India (a share that was probably fairly stable over the twentieth century). It is difficult to be sure, because these categories are more fluid than they appear.
In spite of affirmative action policies, socioeconomic differences by caste continue to be large. Even using the three broad caste-group categories, there are significant differences in fertility, mortality, and health that are not explained by differences in standard socioeconomic factors such as income and education. For example, the 1998–1999 Indian National Family Health Survey (NFHS) found infant mortality rates (per 1,000 births) of 83 for the SCs, 84 for the STs, 76 for the OBCs, and 62 for the other (upper) castes. The corresponding levels of under-five mortality were 119, 127, 103, and 83. The disadvantage of the SC and ST groups is obvious. Fertility differences are less stark: The total fertility rate in 1998–1999 was 3.15 for the SCs, 3.06 for the STs, 2.83 for the OBCs, and 2.66 for the other castes. For fertility, the regional contrasts in India are much larger.
Caste in India also continues to be a determinant of demographic behavior because it is strongly associated with socioeconomic class, and socioeconomic differentials in fertility and mortality are still marked at this stage of the demographic transition. For instance, literacy levels among ever-married female respondents in the NFHS were 27 percent for the SCs, 21 percent for the STs, 39 percent for the OBCs, and 56 percent for the other castes; corresponding figures for regular exposure to the mass media were 52 percent, 38 percent, 59 percent, and 69 percent. As can be seen in all these indicators, the Scheduled Tribes are the most disadvantaged groups, significantly worse off than even the untouchables. Not only are these tribal groups at the lowest levels of socioeconomic development, they are also the least organized for any kind of concerted political action.
For historical (often to do with the emergence of charismatic leaders), demographic (often to do with their relative numbers), political (often to do with mass mobilization), and cultural (often to do with the position of women) reasons, caste plays more or less salient roles and the relative power of the lower castes differs in different parts of the country. The future of caste as an organizing principle of society is also difficult to predict. With education, urbanization, and all the forces associated with modernization, it could become less relevant in all but the most intimate areas of social relations (that is, marriage) and a less obvious marker of demographic behavior. However, it is also plausible that the upheavals of modernization will mean more caste-based conflict and unrest before this happy homogenization of socioeconomic aspirations and achievements comes about.
Bailey, Frederick George. 1960. Tribe, Caste, and Nation: A Study of Political Activity and Political Change in Highland Orissa. Bombay, India: Oxford University Press.
Bayly, Susan. 1999. Caste, Society and Politics in India from the Eighteenth Century to the Modern Age. Cambridge, Eng.: Cambridge University Press.
Beteille, Andre. 1969. Castes, Old and New: Essays in Social Structure and Social Stratification. Bombay, India: Asia Publishing House.
Dangle, Arjun, ed. 1992. Poisoned Bread: Translations from Modern Marathi Dalit Literature. Hyderabad, India: Orient Longman.
Das, Veena. 1977. Structure and Cognition: Aspects of Hindu Caste and Ritual. Delhi, India: Oxford University Press.
Dirks, Nicholas B. 2001. Castes of Mind: Colonialism and the Making of Modern India. Princeton, NJ: Princeton University Press.
Dumont, Louis. 1970. Homo Hierarchicus: An Essay on the Caste System. Chicago: University of Chicago Press.
Moon, Vasant. 2001. Growing Up Untouchable in India: A Dalit Autobiography. Oxford: Rowman and Littlefield.
Srinivas, M. N. 1962. Caste in Modern India, and Other Essays. New York: Asia Publishing House.
Alaka Malwade Basu
Caste orders the lives of Indian Hindus and has as its basis the fivefold varna division embracing Brahmin, Kshatriya, Vaishya, Shudra, and Untouchable. Within each varna there exist myriad jati, which are small endogamous groups, tied to a defining occupation, based in a village or group of villages, and which provide for the element of mobility within a system where otherwise birth determines social rank.
The varna system provides the system of values, the jati its functional organization and practice. Jatis may seek promotion within the caste hierarchy by adopting the practices of higher varna, which can result in promotion within their varna but not between varna, a process known as sanskritization. It is believed that mobility between varna can only be achieved through rebirth, where the successful practice of the caste code or dharma earns for the individual an increased karma and therefore higher status at rebirth.
The major dividing-line between and within the castes centres around the rules of pollution. These affect commensality (sharing and preparing of food), intermarriage, and any form of social intercourse. Since pollution of food is most likely, the higher varna tend towards vegetarianism, and are also teetotal. For this reason too, meat consumption is gradated, with distinctions being made between mutton, pork, and beef. Spatial segregation is a natural consequence of the jati system, and the segmentation inherent in the system and its attendant rules are overseen by a caste court. The caste system has been able to assimilate non-caste, non-Hindu outsiders very successfully.
Since independence in 1947, the Indian state has attempted to break down caste divisions, although in practice caste retains an important role in the social structure. Some sociologists have attempted (controversially) to extend the term beyond the Indian situation, and to apply it to the analysis of the South African system of apartheid, and even to the system of racial segregation in some parts of the United States during the twentieth century. See also CASTE SCHOOL OF RACE RELATIONS; HINDUISM.
The castes (and sub-castes), numbering many thousands, fit into the divinely originated varna framework, though their origin is later and usually based on secular criteria relating to occupation and area of origin.
Criteria of caste maintenance such as pollution, hereditary occupation, and commensality are necessarily gradually disappearing in public places in urban, industrialized India, but in the countryside (where c.80 per cent of the populace still lives), and in home life, such beliefs and the discriminatory practices related to them still prevail.
Caste has long been the target of reforming groups, both within Hinduism and from the outside. Gurū Nānak and successive Sikh Gurūs declared caste irrelevant to salvation. However, intercaste marriage has always been rare among Sikhs, and at least at that level, caste is far from being eradicated among Sikhs: see BĀLMĪKĪ; BHĀṬṚĀ; JAṬ; KABĪR; KHATRĪ; MAZHABĪ; MISL; RĀMDĀSĪ; RĀMGAṚHĪĀ; RAVI DĀS; RAVIDĀSĪ.
caste / kast/ • n. each of the hereditary classes of Hindu society, distinguished by relative degrees of ritual purity or pollution and of social status. ∎ the system of dividing society into such classes. ∎ any class or group of people who inherit privileges or are perceived as socially distinct: those educated in private schools belong to a privileged caste. ∎ Entomol. (in some social insects) a physically distinct individual with a particular function in the society. ORIGIN: mid 16th cent. (in the general sense ‘race, breed’): from Spanish and Portuguese casta ‘lineage, race, breed,’ feminine of casto ‘pure, unmixed,’ from Latin castus ‘chaste.’
1. In social insects, the existence of more than one functionally different form (polymorphism) within the same sex in the same colony, characterized by morphological features, age, or both. In bees and wasps, the female morphological castes are queens and workers. Within hive bees there are three castes, each performing a special function in the colony: non-reproductive workers (females which are usually sterile but help with provisioning of food for the queen in the hive); drones (reproductive males); and queens (fertile females). In termites, castes are not distinguished by sex: worker and soldier termites may be sterile males or females. There are several different kinds of workers among ants, all of which are sterile females, and ants also have a soldier caste.
2. A system of social classification in humans, in which membership is determined culturally by birth and remains fixed; the group is ranked in a hierarchy of groups in the system.
1. In social insects, the existence of more than one functionally different form (polymorphism) within the same sex in the same colony, characterized by morphological features, age, or both. In bees and wasps the female morphological castes are queens and workers. Within hive bees there are three castes, each performing a special function in the colony: non-reproductive workers (females which are usually sterile but help with the provision of food for the queen in the hive); drones (reproductive males); and queens (fertile females). In termites castes are not distinguished by sex: worker and soldier termites may be sterile males or females. There are several different kinds of workers among ants, all of which are sterile females, and ants also have a soldier caste.
2. A system of social classification in humans, in which membership is determined culturally by birth and remains fixed; the group is ranked in a hierarchy of groups in the system.
Recorded in English from the mid 16th century (in the general sense ‘race, breed’), the word comes from Spanish and Portuguese casta ‘lineage, race, breed’, feminine of casto ‘pure, unmixed’, from Latin castus ‘chaste’.