Morton Klass’s book is perhaps the most important analysis of the Indian caste system to come out of western scholarship in the last thirty years. It comes at an opportune time – when the economic and social crisis of Indian society has reached the point where caste divisions among the labouring masses have become a major weapon of the ruling classes and ‘atrocities against Harijans’ have leaped into the front pages of all daily papers. From brutal landlord attacks and gun battles in the more feudal areas of Bihar (Belchi, Pipra) to kulak-engineered mass campaigns in the more capitalist areas (Kanjhawala), from mass pogroms against dalits in western India’s ‘land of saints’ (Marathwada) to the land of Gandhi where riots have recently broken out over the issue of reservations, no part of India is immune from the poison of casteism. Klass, as an academic anthropologist safely ensconced in the comfort of an American University, is perhaps little concerned about such events. But the fact is that they have forced the Indian Left which is deeply entrenched among the masses affected by caste divisions and caste oppression—to rethink the issue.
Major theoreticians of almost every left party are publishing pamphlets on the issue of caste; new political trends are emerging from dalit and socialist backgrounds that talk of ‘combining caste and class struggle’; and there has even been a communist organization, the Satyashodhak Communist Party, formed around this issue.
This indicates that the Indian revolutionary movement is reaching a point where it cannot go forward without confronting the problem of ‘caste. Though there have been brilliant historical works by such scholars as D.D. Kosambi and Deviprasad Chattopadhyay, there has been almost no theoretical analysis of caste by Marxists that does not dismiss the phenomenon as super-structural maya. The previous tendency among Marxists has been to reject all talk of caste as simply a western academic conspiracy, to see it only as a ‘weapon’ of the ruling class without analyzing what the objective basis is that makes it possible to use that weapon, to describe it simply as a survival of feudalism which is relevant today, and to argue that organizing on common economic issues will be sufficient to bring a bout class unity and that caste will (like women’s oppression) more or less automatically disappear with the achievement of revolution. All this is now proving inadequate to deal with the crisis. Thus, historical materialists have failed to deal with caste and, on the other side the major academic theories of caste advanced till now have been from idealistic, voluntaristic and even racist viewpoints. The result has been a yawning gap in the scientific understanding of Indian society. Klass’s book is a major step towards filling that gap because he takes a basically materialistic (if ‘eclectic’, as he says) approach in his analysis.
The object of Klass’s efforts is not to describe the ‘essence’ or ‘functioning’ of caste but to analyse its basic characteristics and describe its origin and development as a concrete historical phenomenon, i.e., the ‘South Asian Social System.’ To being with he succeeds fairly well in showing that almost all current theories of caste, including those which avoid the issue of origins, in effect end by leaving the field to the ‘racial theory’ propounded by British administrators in the 19th century. This theory is broadly that caste originated when invading light-skinned ‘Aryans’ conquered native darker-skinned ‘Dravidians’, and restrictions on intermarriage and an interlinked hierarchy developed as a means of subordinating and integrating the conquered population. This theory was quickly picked up first by the brahmin elite who used it as a kind of modern justification of their social superiority (ultimate high-caste Indians were as good as Europeans and the origins and core of Indian culture were in the Vedas), and then by social and cultural radicals from the non-Brahmin movements who turned it upside down to argue that native ‘Non-Aryan’ and ‘Dravidian’ peoples had a superior, equalitarian society before they were conquered by the high caste aliens. Today it is the most widely believed explanation of caste in India, and of all Indian scholar politicians perhaps only the ex-untouchable leader Dr. B.R. Ambedkar was an exception in rejecting the theory.
At the academic level, as Klass notes, some version of the racial conquest theory was accepted even by such scholars as Max Weber, while the more sophisticated anthropological theories of today give it support because of their idealism. That is, because theoreticians such as Louis Dumont, McKim Marriott, Stephen Barnett etc. identify the caste system with the ideology that backs it up, and almost inevitably link this ideology in turn to the high brahmanical forms of sanskritic Hinduism which ultimately finds its roots in the Vedas and so in the religion and social system arising out of (if not completely based on) the ‘Indo-European’ conquest.
Klass’s arguments against the racial or Aryan conquest theory are fairly simple. Caste is stronger in the south and in areas which have seen the least invasions; there is evidence of many Indian cultural traits among pre-Aryan peoples including the Harappan civilization; and there is no real evidence at all of any process of massive invasion and conquest by a racially distinct and light-skinned people. In the process he makes a basic statement of his materialism:
Just as in the case of religion, the Vedic system—which here means the classic varna system remains the justificatory and explanatory shell. The caste system is clearly not the classic varna system, even though Hindus believe that castes have derived (or degenerated) from these varnas. Whatever the justificatory framework for believers, we can say—as in the case of the religious system – that the actuality or content of the socio-economic system has an ancestry different from that of the framework.
What then, is caste? Klass’s approach is outlined in a chapter on ‘The Units of the Caste System.’ Sub-castes, or as he calls them, ‘marriage-circles’, are the fundamental units into which every member of the society is born. They exist as ‘corporate groups’ within which a person must marry; normally they are also the group within which other close social relations (e.g., inter-dining) are carried on; they have certain rules of behaviour (from marriage rules to preferred occupations to particular standards of ‘cleanliness’ and relations of social distance with other sub-castes) and enforce these by sanctions including expulsion; and they have a general rank within the socio-economic hierarchy of their area. These sub-castes (known as potjatis, biradiris, etc) are grouped into jatis which generally have an occupational name and are the ‘caste’ of a person in the sense that he/she is known to the wider society by a jati name and not by a sub-caste name; and in turn each jati claims a particular ranking within the all-Indian varna system.
The point here is that nearly all social scientists will agree on these basic characteristics of the caste system. But the difference between Klass and the dominant school of anthropology today is that the anthropological idealists define caste in terms of an assumed ‘essence’ that lies behind these characteristics and provides a logical framework for the system of related sub-castes and jatis. For example, for Louis Dumont and his followers caste is the system of rules of ‘purity and pollution’; for McKim Marriott and Ronald Inden and others caste is a correlated set of rules that define ‘code and substance.’ In contrast, for Klass caste is simply the concrete system of interrelated and coexisting marriage circles, and this is a ‘system’ not in a functionalist sense but because the entire society is composed of such groups so that expulsion from one means expulsion from the society—no person can be ‘adopted’ into a marriage-circle in which entrance is only through birth. The ideological justification or ‘framework’ for this system is a separate and ultimately secondary factor for Klass.
Klass goes on to describe the economic ‘functioning of the traditional caste system, which according to him is basically one way of carrying on agricultural production. Most jatis have occupational names because economic relations are the most important aspect of the relations between the castes. There is no exchange of women and very limited social relations between castes; what they do mainly is to provide economic services (and secondarily cultural and political services) for one another. At the village level this is organized in terms of what has been called the jajmani system in the broader sense in which it includes not only artisan services but also the services of sub-castes who provide labour on the land as tenants or as mainly untouchable field slaves. Such services are not primarily exchanged for money or goods rather they are provided throughout the year according to the ‘duties’ ascribed to the caste and in return those who do not directly control agricultural production receive certain traditional shares in the produce as well as other socio-economic perquisites (e.g., rights to share in particular ways in religious festivals, social functions and the like). This system—and this again is in contrast to a functionalist approach—which does not operate automatically but rather is always under the control of a ‘managerial caste’ or ‘allocative center’, that is the sub-caste or marriage-circle of those who basically control the village land.
What then can we say about the origins of the system? Here Klass turns to a famous thesis of Levi-Strauss that has not yet been integrated into a theory of caste, that is, the comparison’ of ‘caste’ based and ‘clan’—based systems put forward in an essay on ‘The Bear and the Barber’ (the ‘Bear’ represents the totemic names of clans; the ‘Barber’ represent the occupational names of castes). Levi-Strauss argues that a caste society and a clan society have similar structures/at a very general level that can be seen as logical transformations of one another. Castes are endogamous—that is, they do not give daughters outside and have primarily economic relations with one another. Clans are, in contrast, exogamous—that is they exchange daughters externally but are generally economically self-sufficient and have no necessary economic relationship with one another defined in terms of a division of labour. Castes have mainly an economic relationship with one another and their required marriage relations are internal; clans have required marriage relations with one another and their economic relations are primarily ‘internal’ .
At this general level, Klass agrees with the contrast, but adds two points. The first is that on the one hand castes may contain ‘clans’ (exogamous groups normally exist within sub-castes) and that in turn the tribes within which clans exist do not normally exist in isolation in a geographical universe but generally as relatively small units within broad regions of many tribes which have some minimal interaction. His second point is that while it is true that each system can be seen as a logical transformation of the other, historically the transformation could only be one way. It makes sense to see a clan society evolving into a caste society but not vice Versa. The reason is that clan or tribal societies are associated with a hunting-and-gathering or horticultural economy, while caste is linked to a developed agricultural economy generating a substantial surplus.
Klass then goes on to ask, in terms of general anthropological theory, what happens to a clan-type society when new forms of agricultural production begin, when a surplus emerges and economic stratification develops? He refers to anthropologists who described two types of clan or tribal societies: those based on clans which are equalitarian in the sense that all who claim descent from the original or assumed founder or ancestor are socially equal as members, and those based on inequalitarian clans or lineages in which there is stratification within the lineage based on nearness of relationship to the ancestor or to other group members. In lineages there is differential membership, often a more ‘aristocratic’ section exists and at points the relationship becomes so distant that a person is no longer defined as a lineage member. The consensus of most anthropologists making such a distinction has been that inequalitarian lineage-based societies have found it relatively easy to adjust to the rise of a surplus and economic inequality. They simply provide a framework within which an aristocratic class and an inferior labouring class can more or le.ss gradually .develop. But equalitarian clan-based tribal societies cannot make this type of adjustmeent and so have been assumed to be a kind of evolutionary ‘dead end.’
It is at this point that Klass adds the final piece to his puzzle. We cannot assume that anything is a ‘dead end’, and in fact equalitarian clans can adjust to the rise of inequality—by evolving into castes! Specifically what he argues is that in pre-agricultural India the whole subcontinent was an area primarily inhabited by tribal groups based on an equalitarian clan system. As wheat and rice agriculture began to develop thousands of years ago, culminating in the Harappan civilization, a process of widespread social innovation also took place, in which all over the subcontinent castes began to emerge. The process was one in which certain tribal groups could gain control in an agriculturally more productive area—these became the dominant landholding or ‘managerial’ castes—while others moving in would claim a part of the new surplus being produced in exchange, for providing certain kinds of specialist services, as barbers, carpenters, priests, or (if they were economically very weak) as simple dependent field labourers. Once the process of caste formation became established, invaders could fit into it as well, and this is essentially what happened to those Indo-European tribes who actually did wander into India. And the Vedic and later Hindu religious and philosophical systems developed as an ‘explanatory framework’ for an already existing social system—a framework that certainly influenced the development of that system and perhaps was even crucial to its survival, but which was not equivalent to the system itself.
Klass puts forward his argument as a hypothesis, a theory to research and prove or disprove. The strength of his case rests at present on its strong logical plausibility in contrast with the implausibility and bankruptcy of racial conquest or mere occupational theories of caste origins. Scholars tend to ignore the racial theories, and radicals have been polemicizing for some time against the current idealistic theories of caste. But without anything concrete to put in the place of racism and idealism, the result has been to leave a vacuum, a non-analysis at the theoretical level, of Indian caste and its relationship to an exploitative mode of production. It is this non-analysis that is now proving an obstacle in the context of the struggles of the Indian masses and the problems they are facing of caste divisions and caste oppression, and it is here that Klass’s approach offers a way forward.
Nevertheless, this is only a way forward and not a complete theory. There are some points I would like to raise here concerning the deficiencies in Klass’s approach. They are interlinked, and they are perhaps all related to the fact that a theory of ‘caste’ is not equivalent (as Klass’s title implies) to a theory of ‘The South Asian Social System’ —and ultimately to the fact that Klass can be called at most a materialist, not a historical materialist!
The first issue is the relationship between what Klass calls the ‘marriage circle’ (sub-caste) and the jati, which is what most usually translate as the ‘caste.’ According to Klass, it is only the sub-caste that is real and functions as the actual unit of the system because only it has existence as a concrete corporate unit with mechanisms for enforcing behaviour. The jati, in contrast, is simply a cluster of sub-castes with no such corporate existence of its own, no means of enforcing approved behaviour, and is merely the name by which people are known to other sub-scastes members locally. But on this point Klass is wrong. That is, in traditional (pre-British) caste-feudal society, there were actually two types of mechanisms for enforcing caste-prescribed behaviour. One was indeed the sub-caste—but the other was the feudal state, including all of its representatives from the raja or maharaja at the top down to the village ‘rulers’ (dominant caste, headmen etcj. The state was particularly concerned to enforce jati behaviour, i.e., to see that members of each jati followed their prescribed caste duties, did not infringe the general social rules of relationships with other jatis, did not seek to emulate higher jatis, etc. In addition the rajas also often served as courts of appeal for decisions (e.g. decisions of expulsion) made by sub-caste councils.
Because of this, the jati did in fact have a stronger social reality that Klass admits. We may say that the sub-caste was indeed the basic unit of the caste system, but the jati or caste was the basic unit of the whole society, that is of its social division of labour and this was enforced and maintained by the two mechanisms of the sub-caste as a corporate unit and of the caste-feudal state. Now of course the situation has changed: since British rule and then the formation of the postcolonial state, the state mechanisms of enforcing caste have been revolutionized, and the jati now indeed exists only as a cluster of sub-castes.
And this brings us to the issue of the modern transformation of the caste system. Here controversies have been raging: is ‘caste’ dying away? Is the ‘caste system’ being replaced by something else, by ‘classes’ or by something like ‘ethnic groups’?
Answers to this question seem to depend largely upon one’s theories about caste. As far as concretely describing what is happening, there is fairly broad agreement. Caste rules of behaviour are no longer enforced so strictly as before and in many cases except for marriage rules are widely broken; there is now considerable economic differentiation within every caste and sub-caste, so that each now has members of all classes from worker and agricultural labourer to at least middle-class employee and often rich farmer and bourgeoisie. At the same time there is a broad correlation, i.e., low castes are poorer on the whole, upper castes have monopolized the higher class positions. Along with this differentiation castes seem to be emerging as broader groups that compete with one another in the political and economic arenas, and their middle class and bourgeois leaders use their caste membership and caste appeals against one another and against the unity of the proletarianized members of various castes.
All of this is very untraditional caste behaviour. Does it mean that the system is fundamentally changing? For most Marxist scholars and activists, the attitude up .to now has been to say yes—that caste will be overcome and replaced by class as capitalism prevails over feudalism, or if ‘caste’ remains it is simply a sign that feudalism or feudal relations remain. This now seems to be erroneous: for capitalist relations are spreading and new classes are clearly developing, even in the rural areas, and yet some of these most capitalistically-developed areas (northwest India, Maharashtra, Gujarat) have seen the worst caste riots. In some sense, something we must call ‘caste’ clearly continues to exist.
For the schools of anthropological idealists, the answer to the question has also been yes. For if caste is defined in terms of an ideal essence or certain ‘rules of behaviour’, then a change in this essence or these rules mean a fundamental change in the caste system. Thus for example, Stephen Barnett who holds to the ‘code and substance’ theory, argues that the ‘code’ aspect of caste is vanishing (that is, rules of behaviour are no longer fundamental to defining a caste) and only the ‘substance’ aspect remains so that now castes are defined in terms of blood and inherited membership. For him this means that castes are fundamentally changing into something different, and he suggests that they are becoming like ‘ethnic groups’.
But for Klass, caste is not fundamentally changing. His argument is simple:
There is no reason why the system cannot continue to function adequately even when members of a caste engage in different occupations and vary in degrees of individual prosperity …. Endogamy is crucial, however … The rule of endogamy is necessary for the maintenance of the total caste system. In other words, the issue is not ‘purity of descent’ but maintenance of distinct boundaries over time between the units. Without these boundaries, the particular structure lacks any mechanism for enforcing all the rules of the system. Everything rests on this keystone of distinct and bounded units and therefore upon endogamy. Were it ever to become possible for individual households—in large numbers and with ease—to continue over time outside the bounds of the marriage-circle, the system would undergo massive structural revision …. None of the many changes and stresses South Asia has experience in the last century, however has significantly altered the rule of endogamy. Occupations change, rules of diet and association have been drastically revised, even the ideological underpinnings of the system have been challenged and in some cases swept away. But still, an expelled household of any marriage-circle, of any region (in the South Asian countryside) finds it almost impossible to obtain spouses for the children. And the system continues, remarkably unimpaired.
Here I think Klass is basically correct – but incomplete. The marriage-circle (sub-caste) and its mechanisms for enforcing membership and behaviour remains and thus one core of the caste system remains. And because it remains, the caste-sub-caste membership of any individual has an important determining affect on his/her position in the entire society (including class, political system etc), and at the same time the survival of caste in this sense continues to provide a material basis for the retrogressive, hierarchical ideology of caste superiority-inferiority, purity-pollution etc. that is not simply equivalent to racial ideologies elsewhere. But at the same time, the fundamental changes that have taken place in Indian society mean that this caste system now has a different relationship to the entire mode of production—so that at points caste does seem to be functioning much as ‘ethnic groups’ do elsewhere.
The final point has to do with the relation between ‘class’ and ‘caste’, an issue now much debated in India. The traditional Marxist position has simply been to say that ‘it’s really class struggle, it only has the form of caste struggle.’ As against this, trends are now developing to say that ‘class struggle’ and ‘caste struggle’ exist as parallel phenomena, both important, which must be interlinked. This view puts the working class in leadership of an ‘economically’ defined class struggle and dalits and low-castes in leadership of a ‘socially’ defined casts struggle, but does not really show their interrelationship. One version of this even claims that prior to British rule there were no ‘classes’ in India, only ‘castes’ and the struggle between castes.
Klass himself has little to say on this. issue; in fact he does not subject the concept of ‘class’ to any analysis at all. At points he does seem inclined to say that ‘caste’ was really a substitute for ‘class’ as a way of organizing a stratified society (hence the title of the book!). But this rests on an implicit definition of class as an ‘open’ (not birth defined) economic grouping, a definition drawn from bourgeois sociologists. If in contrast, we take a Marxist definition of class as defined by the relations of production, we must come to different conclusions. That is, wherever there is an exploitative mode of production, wherever the surplus is ‘pumped out’ of basic producers under the control and to the benefit of non producing exploiters, classes must exist. Wherever there is surplus and exploitation, classes exist in some form—the question is which form? Here it can be said (very briefly) that in Indian feudal society classes were fundamentally shaped through the caste system (so we call it a ‘caste-feudal society’). Whereas today with the development of capitalist relations and the separation of economic, political and social spheres that is characteristic of capitalism, ‘classes’ and ‘castes’ are coming to exist as separate, but linked phenomenon.
Here important work needs still to be done in developing our analysis of Indian society. The important conclusion, though, for the present is that the caste system still exists in India; it is not withering away and will not easily vanish, and though it is no longer the fundamental defining feature of Indian society it still plays a central role. Exploiters still use caste appeals to divide the masses and there are still sections (dalits and other low castes) who are particularly oppressed because of caste. Hence there is a need for a conscious and central fight against caste oppression itself, a fight in which dalits and their organizations can and must play a leading role. But there is a basic transformation in the relationship of caste to the mode of production, new and revolutionary multi-caste proletarianized classes have come into existence (as well as multi-class exploiting classes)—and a fight on a caste basis alone, or by dalits alone, cannot be sufficient to destroy caste. This can be done only by the masses taking up this fight as a crucial part of the revolutionary transformation of society.
Gail Omvedt is Sociologist and Political Activist.