Share |



There have been disturbing hints that common occupa­tion is characteristic of many who are not in the same caste, and that sometimes men of the same caste follow different occupations. On the other hand, some of those called by the same label (‘Brah­man’, ‘Bania, etc.) in fact belong to different marriage groups. We find caste continually referred to as ‘corporation’ without hav­ing this term defined for us. We have been told that castes may, in addition, sometimes be characterized by ‘rules of commensality’ and by ritual exclusivity. Finally, it would appear that these castes are arranged in some kind of order, some system of ranking or stratification—but exactly what this signifies has not been made clear. In this chapter, we propose to explore these and related issues.


Some of our problems, derive from differences in perception, assumption, and interpretation on the part of those attempting to convey a description of the system. The meanings and applicability of terms are not universally agreed upon in an­thropology, let alone in the social sciences in general. We are con­strained to wander in a universe without common measurements, accepted boundaries, and familiar features. We are all like the blind scholars in the parable dubiously attributed to India, running our hands over whatever part of the elephant we can reach, and, in our individual darkness desperately but inadequately attempting to comprehend the whole.[i]


But, to be fair, some of the difficulty in this case may be attributed to the complexity and seeming intransigence of the mat­ter at hand. The caste system is not a simple proposition, nor is it easily described and encompassed. Brilliant writers have pondered it, and if they have come up with different and even mutually exclusive perceptions, they have probably not been all wrong; my own suspicion is that many of them were more than a little right.


Let us seek our baseline. Obviously, the term “caste system” cannot be applied to any homogeneous society, made up of persons who are, essentially, undifferentiated equals—persons who have equal access to occupations, spouses, and various forms of activity and association. Following this line of reasoning, Max Weber saw “caste” as “social rank”.  In other words, it reflects the occurrence of a “class system” exhibiting “closed status groups.


Most scholars of South Asian society would agree, with the main thrust of Weber's delineation: the caste system re­flects a class-structured society, in which the units are accorded sharply distinct degrees of “social honor” and in which member­ship in the units is by birth. Bougle, it seems to me, has said much the same thing, but he has sharpened our focus so that we may distinguish the Indian caste system from other, seemingly similar, class structures.

The spirit of caste unites ... three tendencies, repulsion, hierarchy and hereditary specialization, and all three must be borne in mind if one wishes to give a complete definition of the caste system. We shall say that a society is subject to this system if it is divided into a large number of mutually opposed groups which are hereditarily specialized and hierarchi­cally arranged—if, on principle, it tolerates neither the parvenu, nor mis­cegenation, nor a change of profession.[ii]


These perceptions of Weber and Bougle give us our starting point: we can agree that the South Asian system is a complex, stratified one that is different from other complex, stratified societies because of the nature of the component units. And this brings us swiftly to the many questions we must address if we are to gain a better understanding of what caste is, and how it came about. Let us begin with two issues about which, unfortunately, there has always been disagreement: (a) What exactly are the ‘units’ or ‘groups’ which, taken together, comprise the system? (b) What is the ‘sys­tem’ and how does it function over time?


We have already noted that Indian society has traditionally been divided—by orthodox Hinduism—into four groups, the var­nas: Brahman, Kshatriya, Vaishya, and Shudra. And we have also noted that most observers of the contemporary South Asian scene agree that, whatever the origin and ancient roles of these varnas, in recent (or historical) times the varnas have not really been ‘castes’. That is, they themselves are not endogamous bodies, nor are they occupationally distinct in any way; they lack organization, coher­ence, and even mutually acceptable definition. Hutton puts it this way:

As a matter of fact none of the four terms for varna now represents anything but groups of castes. All Brahmans do not intermarry, but there are many endogamous Brahman castes.... the whole of the Kshatriya varna is claimed by Brahmans to have been extirpated by Parasurama, but if so it has been replaced by manufactured Kshatriyas, and in any case Kshatriya rank is claimed by many whose title is one of function or of creation rather than of inheritance. Numbers of Sudra castes have taken to wearing the insignia of the twice-born, and some of them gain acceptance, after a time, as doing so legitimately.[iii]


There are, in short, many jatis in South Asia claiming to be of a particular varna, and the claim may be denied or accepted by other jatis in turn claiming membership in the varna. Since, how­ever, there is never intermarriage between members of two differ­ent jatis (whether of the same varna or not) there is no way to effectively oppose the claim of another jati to membership in one's varna. Thus, as Hutton points out, some in India may argue that there are no “true” Kshatriya jatis any more, or no “true” Vaishya, but such assertions in no way prevent jatis from claiming member­ship in either varna. The caste system seems to function perfectly well in spite of all this. Indeed, it appears to function despite the fact that there are regions in India without Kshatriyas, or without Vaishyas—while in almost any part of the subcontinent there are jatis to be found which have never been assigned to any of the four varnas!

“Varna”, Hutton suggests, is a term best translated as “group of castes”, and we can leave it at that. It may be of some ritual, social, and occasionally even economic, significance if one's jati is accepted as belonging to a particular varna, or to none of them. The varna itself, however, has no unity, no formal organization, no leadership, and—most important—no control over its members. The claim in India to varna membership might be likened to the claim in England to be of “Norman blood.” Such a claim might generate support from the genealogist, doubt from the historian, and amusement from the geneticist; it might affect one's socioeco­nomic standing in certain circumstances, but it is of little real use to the analyst of the present social system.[iv]


Increasingly, therefore, the attention of those who would un­derstand the caste system has been directed to the social category called ‘jat’ or ‘jati’ over most of South Asia—and this is the category usually referred to nowadays as ‘caste.’ Is the jati, then, the unit we are seeking: endogamous, organized in some fashion, with control over its members? The answer, unhappily, is no. Adrian Mayer has dealt with the matter most clearly and succinctly. Discussing the ‘caste’ as a category, he observes: in some ways this is nothing but a category composed of subcastes, rather than a group in its own right. For, though the caste is endogamous, the smallest endogamous units are the sub-castes. Again, the caste as a whole has no mechanism for settling disputes, for adjusting the status of mem­bers, and so forth. Only in relations with other castes is the caste a signifi­cant unit. For people of other castes do not, as a rule, regard the caste as the sum of the constituent sub-castes, but see it as an undifferentiated group (Mayer 1970: 5).


Where does this leave us, then, in our search for the effective unit of the system? We learn that the caste (or jati) is endogamous, and is viewed by outsiders—those of other castes, as well as foreign observers—as an “undifferentiated” and, perhaps indivisible body. The same thing, however, might have been said of varna; it, too, is endogamous in the sense that marriages may be observed to take place within it,' and its internal divisions, too, may not always be apparent to outsiders. We have now learned that the jati, like the varna, is without unity, organization, or leadership. Most of all, as Mayer points out, it too has “no mechanism for settling disputes, for adjusting the status of members, and so forth”. Thus we observe that the caste, like the varna, has no control over its membership.[v]


This issue of “control over membership” is a vital one, for—as we have been told by practically everyone—endogamy is the prime distinguishing characteristic of the caste system, and those who violate the rule are punished by expulsion, usually referred to as “out-casteing.” But who enforces this rule of endogamy, who has the power of expulsion? Clearly not the varna—and, as we now see equally clearly, not the caste (the jati). Mayer draws our attention to the• social unit he calls the “subcaste,” defined in the passage quoted above as “the smallest endogamous unit” within a caste.  Mayer was not the first to use the term “subcaste”; he notes that he derives it from G. S. Ghurye. Indeed, he even quotes Ghurye's view that: There is ample reason why, to get a sociologically correct idea of the institution, we should recognize subcastes as real castes (Mayer 1970; quot­ing from Ghurye 1950: 20). That Mayer does not entirely subscribe to Ghurye's conclu­sion is interesting. Apparently he considers “caste” just as real as “subcaste,” though he acknowledges the “penetration” of Ghurye's observation. Our own view is that, given the importance from a structural perspective of what Mayer, Ghurye, and others refer to as “subcaste” (whatever the “reality” of other units), we  wish they had chosen a stronger, more discrete term for the cate­gory. Manifestly, this unit is not sub- anything—it is the unit of endogamy, organization, and political control within the system, for there is none other.[vi]


Unhappily, however, no one term is in use for this social category in South Asia (or even in the literature on South Asian society). For the region of Madhya Pradesh in which he conducted research, Mayer reports the use of the term biradari to represent “subcaste” (1970: 152).


We lack sufficient information about the variety of ways mar­riage-circles may be organized in different parts of the subconti­nent. Just as the name differs from region to region (and perhaps, in some cases, from caste to caste), so too do the conditions of marriage-circle organization. Mayer, for example, refers to what he calls “kindred of recognition”. This is the population within which marriages are made and/or kin links can be traced through mutual kin. These two features tend to go together, for people will not make marriages with families about which they know nothing, and the information runs along kinship channels. The kindred of recognition forms a large and rather amorphous body, being much the same for the members of a single village's subcaste group (for they are usually inter-related),.... The kindred of recognition is, in the instances we have recorded, a defecto endogamous body, for it contains enough people to satisfy the search for partners. ..(1970: 4). Mayer then goes on to contrast this “de facto endogamous body” with one described by E. Kathleen Gough for Kumbapettai in Tanjore: where the definite number of ,eighteen villages forms an apparently dejure endogamous unit, and this may be a fundamental difference between types of local structure.[vii]


Thus we observe two structurally different types of marriage-circles. One, Mayer's defacto type, constitutes a network radiating outward from any given component family. The second, Mayer's dejure type, is composed of a specified number of families residing in a specified set of villages. Many interesting questions could and should be pursued: What is the distribution of these types in geo­graphic, social, and economic terms? What other types, if any, remain as yet unreported? Such questions must resolutely be ignored here; what is im­portant for us is that marriage-circles are found throughout South Asian society; in principle, every family on the subcontinent belongs to one. Further, the marriage-circle—whether of the de facto, the de jure, or whatever, variety—is everywhere in principle the endogamous body with the power to control the behavior of the membership.[viii]


Now this is a very big statement. How can we be certain that marriage-circles of every region and station have this power when, as we have just seen, jatis and varnas lack it? For one thing, we can infer it: the system is characterized by a rule of endogamy and so somewhere within the system there must be a unit capable of enforcing such a rule—and these are the ‘smallest endogamous units’. Further, the field anthropologist observes the unit in action as soon as he or she reaches the South Asian countryside. An Indian marriage-circle reflects both territo­riality and kinship, and to ignore one of these organizational princi­ples is to distort one's understanding, of the structure. A marriage-circle is, on the one hand, a kin group: Mayer's “kindred of recog­nition” or even, in Nur Yalman's words, an “endogamous kindred”.   It contains all those families with which a given villager considers himself related (by blood, marriage, or whatever), and it contains, in principle, all the families with which marital arrange­ments may be made by members of that villager's family. Follow­ing Yalman, we may say that this body includes all those too closely related to be acceptable as marriage partners, all those to whom one is already related by marriage, and all those sufficiently distantly related to be eligible for marriage. Beyond this body, it should be noted, are not only (and obviously) total strangers who are therefore not acceptable as marriage partners, but also some who are in fact perceived as kinsmen. These are the members of distant marriage-circles of the same caste, who are assumed to derive from the same origin, but who are too distant in space and too different in behavior to be acceptable in marriage.[ix]


And, on the other hand, the de jure marriage-circle is a territorially circumscribed body. It is composed of all the families in a given geographic area who belong to a specific jati: all the Brahmans, say, or all the Barbers, or all the Sweepers, in a de­marcated area. Such an area has specified boundaries—north, south, east, and west—and it contains a particular number of villages. Let us suppose, for example, that we refer to the total geographic area enclosed between River A and River B from east to west, and from a railroad track in the north to an escarpment in the south. Let us imagine that there are, in all, some seventy-five villages within this area, but that Brahmans of a particular jati are found in only twenty of these, and that the Brahmans average some ten families in each of the villages in which they are found. The marriage-circle of the Brahman jati in this area would therefore consist of some two hundred families residing in some twenty villages scattered over the area. Note that in this area there are also villages with no Brahmans, and perhaps even some villages containing Brahmans of another jati, who have their own marriage-circle and who do not associate—at least in terms of marriage exchange—with the alien Brahmans whose marriage-circle intersects with their own. Mayer derived his example of de jure sub-caste from a paper written by Kathleen Gough. In that work (1956), Gough refers to a marriage-circle of Smartha Brahmans of Tanjore living in some eighteen villages (see also Gough 1962). In this study, we have noted a marriage-circle of Kanauj Brahmans. The de facto marriage-circle described by Mayer appears to constitute a network, but a territorially-based one, for it is made up of all the families of one's jati, in a circle around one's own, within a manageable distance, in terms of travel and of monitoring.[x]


The geographical distribution of a marriage-circle of any particular caste represented in a given village need not coincide with that of any other caste of that village. We find that every marriage-circle is in principle independent of all others, both those of the same jati and those of different jatis. Each marriage-circle has its own independent leadership; each sets up, and enforces, its own rules and regulations.


Max Weber has provided us with the term verband, usually translated as “corporate group” and defined, according to Weber, as: “A social relationship which is either closed or limits the admis­sion of outsiders by rules” and in which the “order is enforced by the action of specific individuals”.  From what we have seen, it is clear that neither varna nor jati can be termed a verband. It is equally clear that the Indian marriage-circle is just such a verband. As a matter of fact it is a special kind of verband. To use Weber's terminology again, it is an autocephalous verband—one with its own leadership and internal control, admitting of no other, or higher, level of authority.


“'Corporation” (or “corporate' group”) has of course been defined differently by other writers. Morton Fried, for example, approaches ‘corporate group’ in terms of ownership of common property, and as he seeks to unravel the evolution of complex political systems he explores the development of the division of functions (such as ownership, management, interest, etc.) in a corporate group. We have been influenced by Fried's work, as will be demonstrated in a later chapter. One problem here, as one might suspect, is that there is the validity of both Weber's and Fried's approach to ‘corporate group.’ One can use one in analyzing the structure of caste and the other in exploring the emergence of the system. Unhappily for us, neither approach to “corporate group” makes any provision for the other. What, then, does one mean by 'corporate group”? Depending on the problem, and for the purposes of this essay, one can mean either Weber's or Fried's definition—and in the end one can mean both, properly synthesized. Thus if the caste system is characterized by a rule of endogamy, active and enforce­able, then it is the verband—the dejure or de facto marriage-circle ­that is the body, and the only body, with the power to enforce the rule.


It is, of course, necessary to bear in mind that there is much regional variation in the internal regulations of marriage-circles. Cer­tain marriage-circles, for example, are composed of exogamous sub­units, or patrisibs, usually referred to as gotras. In such cases, the rules require not only that all marriages be contracted within the marriage-circle, but, in addition, that no marriages be contracted between families of the same gotra. The variations are many. Mar­riage-circles found in the same village, but representing different jatis, may have different rules about internal exogamy: some may exhibit gotras, some may have other types of exogamous subunits, some may forbid only the marriage of close kin. Some marriage-circles extend the boundary of endogamy to include one or two contiguous and closely related marriage-circles of the jati; some per­mit hypergamous unions with women of other jatis in special cases, and so on. The variations, interesting and important though they are, must not obscure the larger fact: the marriage-circle is ubiqui­tous in the South Asian countryside, and it is everywhere and at every level an autocephalous verband, enforcing its own rules and controlling the behavior of its members.[xi]



In the South Asian countryside, marriage and death are invaria­bly viewed as life crises. The appropriate observation of these events requires the participation of substantial numbers of the affected family's marriage-circle. In other words, the head of the household in which the marriage of a girl, or the death of an adult person, is to be solemnized, must invite to that event the representatives of as many households of his marriage-circle as he possibly can.


Again, let us not ignore or forget variation: some marriage-circles place greater emphasis on marriage than on death as the event to be marked, some invite guests to solemnize other life crises (such as birth), and so on. In some cases, entire households are invited, and in some cases only household heads. Wealthy families can invite and feed many more guests than can poor ones (whatever the group's range of wealth), and there are innumerable other permutations and variations.


Thus, everywhere, in every marriage-circle, there is a charac­teristic event bringing together a substantial number of household heads. These events occur with a sufficient frequency within the marriage-circle to permit the body to function as a verband, cap­able of controlling the behavior of its members. Should it become impossible to have such gatherings with sufficient frequency, atten­dance, and/or representation, then the marriage-circle will divide into smaller units which have the necessary capacity.[xii]


In every village, then, of every region of traditional seden­tary, agricultural South Asia, every household is normally a constituent member of such a marriage-circle. All households are in principle equal to each other; at least, each household is entitled to one vote, through its head or representative, at those occasions when a vote is taken. Such occasions occur (a) when a substantial number of household representatives are gathered together for the observance of some member's life crisis, and (b) when the leader­ship of the marriage-circle converts the gathering at the event into an official assembly of the marriage-circle for the purpose of delib­erating some matter of concern to the body. Further, in an emer­gency, the leadership may call a special gathering of household heads.[xiii]


The marriage-circle assembly may deliberate a variety of matters, but the most serious and ubiquitous problem—one that seems to receive the attentions of almost every marriage-circle at one time or another—is that of the marriage of a member to an outsider. Almost everywhere, the rule is that spouses for children must be chosen from households belonging to the marriage-circle. In those cases where there can be an exception to the rule (i.e., cases of permitted hypergamy, or where certain outsiders are in­vited to join the jati, etc.) it is understood to be just that, a special exception to what everyone knows to be the rule. But every marriage-circle has many rules—and not just about marriage—and all are enforced by the marriage-circle assembly. A miscreant may be ordered to pay a fine, or he may be beaten, or punished in other ways. The most serious punishment, and the one most likely to be held out as a threat to the household that has contracted an unacceptable marriage, is that of “out-casteing” or expulsion from the jati. We have already noted that the jati has no power to expel anyone. Technically, therefore, the marriage-circle has the power to expel the miscreant person or household only from the verband itself, but the members of the latter see themselves as representing the jati, and they may term their action an expulsion from the jati—or even from the entire varna.[xiv]


Whatever else such an expulsion may mean—for there is great variation from region to region and from caste to caste in terms of whether expulsion carries with it deprivation of livelihood, loss of living quarters, etc. it always means, specifically and in­trinsically, that the expelled household is henceforth out of the pool of prospective spouses. Their children are not eligible mates for members of their former marriage-circle, and they are forever barred from seeking mates within the marriage-circle.


It is this eventuality, and not military power or divine sanc­tion or economic deprivation that underlies the authority of the marriage-circle and enables it to function as a verband. Further. the strength of the verband in this regard is a reflection not so much of the internal organization of any verband as it is of the fact that the entire society is composed of such bodies, each with a rule restrict­ing its members from marrying outsiders.


In a society in which substantial numbers of persons are free to contract marriages with whomever they please, such an expul­sion would be an inconvenience, but hardly a catastrophe. If, for example, you are a citizen .of the United States,, or of any other Western nation, and you are excluded from marriage with other members of what has been your ethnic or religious, or whatever, body of association, you are of course perfectly free to join another such body—or to move freely among the many people without such primary allegiances. You may reasonably assume, therefore, that neither you nor any of your children will experience particular difficulty in finding a spouse—at least, not for the sole reason that there are no potential spouses available. What we are discussing in these pages, however, is a society in which every household belongs to an endogamous marriage-circle. If, in such a society, a house­hold is expelled from its marriage-circle, where shall the members seek mates? The answer, in principle and often enough in fact, is nowhere’.


This ubiquitous rule of endogamy, along with its dramatic enforcement through expulsion, has unquestionably captured the imagination and attention of students of South Asian society. It is necessary for us to note again, therefore, that a marriage-circle, and most particularly its assembly, is concerned with many matters other than misalliance and its prevention. In broad terms, the mar­riage-circle assembly (as all such bodies are continuously aware) is concerned with the maintenance of the marriage-circle over time, and the continual protection and enhancement of its membership. One might legitimately inquire, at this point, why it is so desirable that the marriage-circle be preserved, maintained, and enhanced. The question is indeed legitimate, but the answer must be saved for later.[xv]



In any event, the marriage-circle seeks, as a body, to preserve and enhance itself, and it does this, to begin with, by seeing to it that all members observe all the rules of the body. These include: marital rules (such as endogamy—but also, and variably, patrisib exogamy, prevention of widow remarriage, certain kin avoidances within households, and so on), dietary rules (such as those on eating with persons of other castes, accepting food from persons of other castes, acceptable or non-acceptable foods, and so on), occupational rules (activities and occupations permitted and not permitted for members of the marriage-circle), and many others. There are, needless to say, some loopholes, some special exceptions and occasional possibilities. Very low-ranked castes, for example, in, some cases permit expelled individuals of much higher-ranked castes to join their marriage-circles (part Dun). There is no happy solution, however, anywhere in South Asia, and so expulsion from the marriage-circle has always been, and still remains, a formidable threat and therefore the most effective source of control by a marriage-circle over its membership.


These rules function to maintain the marriage-circle, not to main­tain “purity of blood.” Thus, hypergamy, hypogamy, caste adoption, and other such “violations” of “purity of blood” may be perfectly acceptable in cases where they pose no threat to the maintenance of the body—however much they may astonish the foreign observer preoccupied with questions of “purity,” or even of “descent:” region to region and from caste to caste. Nevertheless, the rules for each marriage-circle are stated, specific, and known to all the mem­bers of that body. One marriage-circle, for example, will not permit its members to eat meat of any kind, while another permits the eating of mutton and fish, though not of beef and pork. One mar­riage-circle permits its members to engage in any occupation or activity that does not involve the taking of plant or animal life, while a different group permits any activity to its members, barring only that of removing the bodies of dead dogs from the streets. In one marriage-circle a widow may never remarry or even live with another man, in a different one a marital union of sorts is permitted for widows, but only with men who have themselves been married before. We may find a marriage-circle in which a man is punished severely for sexual association with a woman of any other caste, and still another one in which a man may have sexual relations with (though not marry) women of certain specified other castes. Here, a man is punished if he touches—or even, speaks to—his younger brother's wife, there he is punished if he shaves himself or cuts his own toenails.  One might say that the point is not what the rules are, but the fact that there are always rules—meticulously defined and rigor­ously enforced—to distinguish the marriage-circles from all others, even of the same jati.[xvi]


In addition, a marriage-circle can modify, drop, or add any rule—for who is to forbid it? Changes in the rules, therefore, are or can be an agenda item for any marriage-circle assembly. The abandonment of dowry, the acceptance of a new occupation or practice, a change in dietary or marital restriction-all of these, and many more, are referred to in the ethnographic literature. Some­times the modifications in the rules constitute responses to internal stresses—as in the case of abandonment of dowry, say—and some­times they reflect changing economic circumstances or new dimen­sions of interaction with the world beyond the marriage-circle. The experiences of the Jadays of Agra (Lynch 1969) and the Chamars of Senapur (Cohn 1955) are but two classic examples of marriage‑circles enforcing, rather than preventing, change, but they are more than enough by themselves to bury forever the charge, or assumption, that the caste system is ‘rigid’.


Finally, a marriage-circle—when circumstances demand, and to the extent that it can—strives to protect and support its mem­bers in their daily relations with other people. It is this particular aspect of a marriage-circle's behavior that has caused some people to liken caste to ‘guild’.


The marriage-circle must come to the aid of a persecuted member, as in the case of the Senapur Chamars who were beaten by fellow villagers for observing the new Chamar rule to eschew scavenging. If men of the dominant caste of a village refuse to pay the village barber what he feels he is entitled to, he may complain to the leaders of his own marriage-circle. If they feel his complaint is justified, they may decide to help him relocate to another village—and then refuse to permit any other members of the mar­riage-circle to replace him in his former village despite the en­treaties of the lords of that village, who are now faced with the awful prospect of cutting their own hair and nails. It is obvious that we are now moving from unit to system; the two are in any case inextricable. The units, the marriage-circles, control the behavior of their members by threatening expulsion. This threat has teeth because everybody in the society belongs to a marriage-circle—this is a society composed of autocephalous ver­bands—and so the expelled person is not only cut off from his traditional source of spouses, but he will not be able to find another source! Similarly, the members of the dominant caste in a village cannot exceed certain bounds in their treatment of subordinate villagers because each of these belongs to a marriage-circle of his or her own. If the subordinate villager's marriage-circle can exercise control over a necessary occupation, the group can protect the fellow member. In a society in which people may be forbidden by their own marriage-circle even to cut their toenails, there are going to be a lot of “necessary occupations”. Thus, for a marriage-circle to be a meaningful and effective body, it must be part of a system of autocephalous bodies—and the system, we begin to see, has social, economic, and ideological ramifications, some of which will have to be explored further.[xvii]


Having examined the units of the system, we are in a position to say that it is a particular type of stratified society—one in which the individual articulates with the total society as a member of an autocephalous verband (specifically, in this case, a marriage-circle), rather than solely as an individual, or even as a member of a ‘social class’. E. R. Leach makes a similar point, though he does not appear to be aware of the structural implications of marriage-circle, or subcaste, as opposed to caste: a caste does not exist by itself. A caste can only be recognized in contrast to other castes with which its members are closely involved in a network of economic, political and ritual relationships. Furthermore, it is precisely with these intercaste relationships that we are concerned when we discuss caste as a social phenomenon. The caste society as a whole is, in Durk­heim's sense, an organic system with each particular caste and subcaste filling a distinctive functional role. It is a system of labour division from which the element of competition among the workers has been largely excluded. The more conventional sociological analysis which finds an anal­ogy between castes, status groups, and economic classes puts all the stress upon hierarchy and upon the exclusiveness of caste separation. Far more fundamental is the economic interdependence which stems from the pat­terning of the division of labour which is of quite a special type (1962: 5).


Leach draws our attention to the system, and to the need to examine economic interdependence—an issue, as he notes, all too frequently ignored in much of the litera­ture on caste. It is not clear to us, however, why Leach urges us to explore the economic dimensions of caste solely in Durkheim's terms. To describe the caste system as an ‘organic system’ is somewhat tautological, at least to a functionalist, and it provides little insight into how the caste system differs from other social systems (all of which, presumably, are organic systems’). The only difference that Leach can point to is that in the caste system ‘the element of competition among the workers has been largely excluded.’ There is evidence, as we shall soon see that this suppo­sition does not always hold up. Thus, Leach offers important in­sights—for example, he advises us to view caste “as a system of labour division” rather than only as a system of ritual and social hierarchy and separation—but his vision is limited, perhaps be­cause he does not have at hand the kind of analysis of the units of the system that we have just gone through.



In the light of that analysis, we are able to inquire: what kind of economy do we find when autocephalous verbands, arranged in a complex stratified society, engage in economic exchanges? Technology is the important variable. This is a world of rocks and riven, sticks and steel, of sun, air and starlight, of galaxies, atoms and molecules. Man is but a particular kind of material body who must do certain things to maintain his status in a cosmic material system. In recent years, two approaches have been championed (and of course attacked) by anthropologists interested in economy and society. One of these approaches is customarily referred to as eco­logical anthropology and derives from the writings of Leslie White, Julian Steward, and others. The second has come to be called the substantives approach to economic analysis, and derives from the work of Karl Polanyi and his associates. We have found it advisable to make use of both of these approaches, however uneasy they fit, in our efforts to understand the economy of caste.[xviii]


First, therefore, we shall consider the South Asian ecosys­tem—the interaction between humans and the physical environ­ment of the subcontinent—leaning heavily upon the insights of Marvin Harris, the anthropologist who has made the most signifi­cant contribution to our understanding of this topic. Then we must turn to the workings of the South Asian economy, and it will be necessary to examine the arguments of Karl Polanyi, because the best analysis of South Asian economic processes available to us is that of Walter C. Neale, a student of Polanyi's and a ‘substantiv­ist’ economist.



But Neale wrote as an economist, not as an anthropologist, and so to complete the work of this chapter we will have to turn to the ethnographic record. Specifically, we will use Kathleen Gough's account of economic transactions in the early days of the village of Kumbapettai to see how caste functions as a socioeconomic system. Throughout this chapter, our concern is with both econ­omy and society, with human institutions and with the exchange of goods and services, with the material universe and with the ideo­logical one. The task is to fit data and theory together, so that the two levels of concern become integrated.


A quick flourish of the flag of eclecticism is in order here. It is certainly instructive to note that while the second half of the nineteenth century witnessed an increasing polarization in the social sciences between materialists (such as Marx and his stu­dents) and functionalists (such as Durkheim and his students), the second half of this century has witnessed a curious rapprochement between the two polar theories. The students of Durkheim urged us, as does Leach to view society as an organism, analogous to a living crea­ture. Leslie White has reminded us, however, that any living or­ganism is invariably imbedded in the material universe of matter, motion, and energy: Plants are, of course, forms and magnitudes of energy. Energy from the sun is captured by the process of photosynthesis and stored up in the form of plant tissue. All animal life is dependent, in the last analysis, upon this solar energy stored up in plants. All life, therefore is dependent-upon photosynthesis.


With this biological dictum to guide him, White could turn to society-as-organism with a materialist perspective. For, if every biological organism is basically an energy-capturing system, then so too is every social or cultural organism. White concludes that “culture is a mechanism for harnessing energy” (Ibid.: 369)—and suddenly the seemingly unbridgeable chasm between the two theo­retical positions has ceased to exist! At least, once we agree that a social organism is an “energy-capturing system” then it must follow that the social organism is bound by the laws of the material uni­verse, and most particularly by the laws of thermodynamics, and is therefore amenable to materialist analysis. White arrives swiftly at the following: We can now formulate the basic law of cultural evolution: Other factors remaining constant, culture evolves as the amount of energy harnessed per capita per year is increased, or as the efficiency of the instrumental means of putting this energy to work is increased.[xix]


Some scholars, such as Harry Pearson perhaps, might con­sider this statement more of an inference, or an hypothesis—or even a pious hope—than a “law”, but we need not trouble ourselves about that now. White's perceptions, we can note, led directly to much rewarding research and theoretical clarification. Anthropologists have returned to the long-evaded questions of cultural evolution because of White's work. This is important because culture has evolved, and continues to do so. We ignore that at peril to our discipline. And so, because of White, we understand more about the sequences and processes of cultural evolution than ever before. At the same time, anthropologists—as a direct result of the work of cultural ecologists—are becoming more and more con­vinced of the validity of Durkheim's perception, referred to by Leach in the passage cited earlier, that the institutions and elements of an “organic system” do indeed all fulfill “functional roles”. If this means, as it is usually taken to mean, that no social organism can survive over time if any of its institutions are massively dys­functional, then surely it must follow that the society's economic (ecological relationship with its material environment also cannot be dysfunctional. In other words, there cannot ever be an “irrational economy” continuing over a significant stretch of time. And this perception brings us back to South Asia, the econ­omy of which has long been characterized as “irrational”. Marvin Harris, in his penetrating paper, “The Cultural Ecology of India's Sacred Cattle” reviews a wide-ranging body of literature, all of which reflects a view that the economy of South Asia is now, and by implication has always been, wasteful, impractical, irra­tional, and—in short—uneconomic.[xx]


This view of the economy focuses on what is perceived as Indian misuse of cattle. To summarize points made in the works cited by Harris, the traditional argument is that in India there is, apart from an enormous and hungry human population, an addi­tional population of some two hundred million cattle. These animals, it is believed, are “worshipped” by Hindus and are therefore bred and kept alive for reasons having nothing to do with their propensities for producing milk or meat. They compete with hu­mans for scarce food resources; they stray freely and graze with abandon and without hindrance in. country rice field and urban fruit stand. And finally, so the litany runs, they constitute a total waste of meat—and this in a protein-poor land—since beef is for­bidden to Hindus. In challenging this traditional view of South Asian economy, Harris is, as always, explicit about his theoretical interests and concerns: In spite of the sometimes final and unqualified fashion in which “surplus”, “useless”, “uneconomic”, and “superfluous” are applied to part or all of India's cattle, contrary conclusions seem admissible when the cattle complex is viewed as part of an eco-system rather than as a sector of a national price market. Ecologically, it is doubtful that any component of the cattle complex is “useless”, i.e., the number, type, and condition of Indian bovines do not per se impair the ability of the human population to survive and reproduce. Much more likely the relationship between bovines and humans is symbiotic instead of competitive. It probably represents the outcome of intense Darwinian pressures acting upon human and bovine population, cultigens, wild flora and fauna, and social structure and ideology.[xxi]


Harris then proceeds to analyze the ecosystem, particularly in terms of the human/bovine interrelationship. We shall attempt to summarize some of his findings. When we refer to ‘the ecosystem of South Asia’ we are talking about the ways in which the enormous human population of the subcontinent (India, Pakistan, Bangladesh, etc.) provides, and has provided over the millennia, for its needs. Some 80 percent of the human food energy requirements of the area (Harris calls these “calorie rations”) derives from grain crops, primarily from rice and wheat, and the proportion of each sown reflecting local environmental conditions. Unlike horticulture, or kitchen garden­ing, in which humans can provide all the necessary labor, grain production almost everywhere requires the labor of animals in addi­tion to that of humans.  In South Asia, this has meant the partici­pation of bullocks (cattle or water buffalo) in the various stages of crop production: preparation of the fields, plowing, irrigation, har­vesting, and transporting.


Viewed therefore from the perspective of the ecosystem, Harris points out, the primary function of cattle in South Asia is to provide the crucial traction without which no crop of significance could be produced in the area. Citing the appropriate literature, Harris, demonstrates that working cattle and buffalo (that is, cas­trated males) are in minimal if not indeed short supply. Since oxen cannot replace themselves, a substantial herd of cows (and a lesser number of bulls) must always be present to replenish the herd of working animals. The primary function of cows, it follows, is to pro­vide replacements for working animals, not to provide milk or meat.


According to Harris, the next most important function of South Asian cattle population, from an ecological perspective, is to provide the fuel with which the grain crop is converted into food—that is, into something humans can eat: In India cattle dung is the main source of domestic cooking fuel. Since grain crops cannot be digested unless boiled or baked, cooking is indispensable (Ibid.: 53). Harris notes, of course, that the cattle population also pro­vides a substantial portion of the protein requirements of South Asian humans in the form of milk, milk products, and meat.' He argues, however, that if the quality or quantity of milk or meat production from Indian cattle is not up to that of the West, that fact must be equated with the traction and fuel producing qualifica­tions of the animals, which would be much less important in mod­ern Europe.[xxii]


Finally, Harris points out, cattle in South Asia make all their contributions to the ecosystem without drawing significantly upon energy resources needed by humans. After all, cattle feed in South Asia primarily upon fodder (straw and other crop residues) and waste‑land grass and shrubs, none of which can be used for food by humans.


It should be noted, of course, that Harris is concerned only with the physical needs of the human population in South Asia. We do not see them, as he discusses ecosystem, as socially or culturally distinctive; there is, for example, no mention of caste. No matter. Harris' ecological analysis is most revealing and illuminating. Fur­ther, it cannot be negated merely because a few of his arguments are open to challenge or modification. Some economists,' for example, have questioned whether the system is necessarily ideal: is a more effective use of the elements possible? Could fodder, say, be replaced in the cattle diet by some­thing else or could cow dung be replaced by another fuel so that fodder and dung could be available for use as fertilizer? Might not the cattle be bred into better milk and meat producers without necessarily lessening their traction contributions?


Functional analysis, which includes ecosystem analysis, al­ways risks the too easy conclusion that “all is for the best, in the best of all possible worlds”. But Harris' work raises triumphantly above all these cavils, for what he has shown us is how the system works and has worked, and for that we are grateful. But there are questions. What is the role of caste in all this? Why have so many students of South Asian society and economy insisted that the economy is “irrational”? Why have they not per­ceived the functioning ecosystem that Harris describes? One reply might be that while the ecosystem clearly func­tions in the way that Harris has outlined, the system and its functioning may not be all that clear to the people who participate in it; in functionalist terms, one might say that the South Asian ecosystem is latent, not manifest. Elsewhere (Klass 1966: 60; 1978: 100), we have discussed the Bengal villager who was unable to see that the bull calf he sold with much regret in the cattle market eventually returned to him, in principle if not in fact, as the not he purchased in the same market to pull-his plow. The farmer had an ideological aversion to hurting a calf in any way---and castra­tion would unquestionably hurt the animal-but he did require an ox for breeding not for plowing. Put another way, the farmer participated fully within his ecosystem, but had difficulty—for reasons growing out of his belief and value system—perceiving the process and its consequences'.[xxiii]


Such ideological barriers to contemplation of consequences can lead to difficulty, especially when persons in such a system are asked to explain the system to outsiders, or to respond to externally derived proposals for change. High-caste villagers, for example, do not permit their minds to dwell on the fact that all the dead cattle of the village are consumed by the lower-caste people of the village. Such upper-caste people cannot be expected to appreciate the im­portance of the contribution made by such meat to the diet of poorer villagers, and consequently offer few objections to proposals by outsiders for the disposal of carrion in ways that will prevent others from eating it.[xxiv]  We may wonder: How can people participate in an ecosystem and maintain it over time when they are ideologically unable to perceive or accept the system as it is? It sounds difficult, and it may well be for other societies, but it is apparently possible in the caste system because of the structural relations between the ele­ments. If this is so, we cannot deal adequately with ‘ecosystem’ without understanding ‘caste’


One might ask, of course, why the expert—and most particu­larly the economist—could not perceive the inherent “rationality” of the ecosystem, especially when he has Harris to guide him. After all, he is not blinkered by the villager's ideology. True, but scholars too have their ideological perspectives, and in a not too different way theoretical concerns and disciplinary interests can interfere with their perceptions. A glance now at some of the con­sequences of such differences in the particular case of the South Asian ecosystem may help us to understand why some anthropolo­gists have found it necessary to part company with classic eco­nomic theory.


Economists tend to focus on “market”—on the interplay of “supply” and “demand”—and surely this is understandable and commendable, given the nature and demands of their discipline. Anthropologists, on the other hand, given the nature of their discip­line, tend to focus on the small community and its inhabitants—on the actual agriculturalist, for example, and his fellow villagers—with all the needs and problems of such a community. From the perspective of the field anthropologist, the “national price market” can seem very remote.[xxv]


It should not surprise us, therefore, to find this divergence of perception underlying the debate over the “rationality” of cattle utilization in South Asia. Thus, the economist Alan Heston, in his paper, “An Approach to the Sacred Cow of India” (1971)—writ­ten, he tells us, to “support the traditional view,” that is, of the economist—takes sharp issue with the ecological interpretation advanced by Harris. Heston argues that a good program of culling—in other words, cow slaughter—would .probably result in a far more efficient system of cattle utilization for South Asia than the one described by Harris.


Heston presents evidence that 24 million cows out of the total Indian herd of 54 million would, if properly provided for—as they would be if the weak and useless ones were not around to compete for food and care—likely produce as much milk as the total South Asian herd does now! And, he goes on, a similar culling of weak and sick oxen would result in fewer but stronger oxen providing the same amount of traction power as at present. Since, according to Heston, larger, healthier animals produce more dung than scrawnier, sicklier ones, even dung production would not be much affected by the introduction of an effective culling procedure. The end result of such a procedure would, be: maintenance of milk, traction, and dung production at present levels; a concurrent saving in fodder; and a significant increase in farmland, for with fewer cattle to graze more land would be available for cultivation.

Heston's proposal for culling the total cattle herd undoubt­edly makes sense from the economist's perspective—that of the “national price market”—but from the anthropologist's perspective there are suddenly many problems. Joan Mencher, in her com­ment on Heston's paper, begins by noting that she finds it “amaz­ing for several reasons”. Heston considers an isolated aspect of the economic system without refer­ence to the rest of the system; he attempts to deal with the economics of cattle without considering patterns of cattle ownership, patterns of land ownership and tenancy, and the facts of agriculture as it is actually prac­tised in India's villages.[xxvi]


Mencher, an anthropologist, invites us to consider the impact of a culling program of these dimensions—in which almost half the national herd is destroyed or otherwise removed—upon the prototypical village household. In such a household, she re­minds us, traction for the fields derives from the efforts of one team of two oxen, however ‘sickly,’ and milk in the diet derives from ‘one scrawny cow,’ who may yet surprise us with just one more calf. Other anthropologists commenting on Heston's paper make similar points (see Nag 1971, Harris 1971, and Horowitz 1971). Heston's proposal may be excellent from the perspective of the national economy, but it is obvious to any anthropologist, on the •basis of his field observations, that what may be economically sound on a national level can spell total economic disaster to the very villager the program was designed to help. It should no longer surprise anyone, therefore, that anthropologist—since the days of Bronislaw Malinowski—have chaffed at the approach to economic analysis provided them by the formal discipline of economics. One can understand why many anthro­pologists (if by no means all) turned eagerly to the analytic ap­proach to economics suggested by Karl Polanyi (1957). Any attempt here to explore in depth the differences between (and among) anthropologists and economists on the knotty issues of ‘economic anthropology’ would take us too far from the concerns of this work.  Here it is useful and sufficient, to cite one anthropologist on the subject. Conrad M. Arensberg—one of the coeditors, along with Polanyi and Pearson, of Trade and Market in the Early Empires—devotes considerable attention to the differing perspectives of the two disciplines, noting: Not free human nature, nor free individuals, nor even any hard and fast psychological attributes of man, within his biological and physiological limits as an animal, give the anthropologist his starting point. In treating any culture patterns, even those of economic institutions, the anthropolo­gist selects ... ‘patterns of interaction’.... anthropology is deeply corn-mined to this priority of social patterns in any scheme for the understand­ing of the substantive economies of the human record. Our discipline is preconditioned to derive specific motivations, whether ‘economic’ or otherwise, from such arrangements, rather than from abstract human na­ture or needs (1957: 100). Clearly, Arensberg is pointing us in the direction of struc­tural analysis. He is implying that the approach to economic analy­sis of Marcel Mauss is more useful to anthropologists than that of Adam Smith—and he is reminding us that our primary concern is with patterned interaction and exchange.


[i]       K.L. Sharma, (ed), Social Inequality in India, New Delhi, India, Rawat Publications,1999, p. 27.

[ii]       K.L. Sharma, Op.Cit., p. 35.

[iii]      Pauline Kolenda, Caste in Contemporary India: Beyond Organic Solidarity, California, 1978, p. 30.

[iv]      Morton, Klass, Op.Cit., p. 77.

[v]       Morton, Klass, Op.Cit., p. 183.

[vi]      K.L. Sharma, Op.Cit., p. 45.

[vii]     Morton, Klass, Op.Cit., p. 55.

[viii]     Pauline Kolenda, Op.Cit., p. 43.

[ix]      N.K. Dutt, Op.Cit., p. 40.

[x]        M.N. Srinivas, (ed), Caste: Its Twentieth Century Avatar, New Delhi, Penguin, 1996, p. 60.

[xi]      M.N. Srinivas, The Cohesive Role of Sanskritization, New Delhi, OUP, 1989, p. 39.

[xii]     M.N. Srinivas, Op.Cit., p. 82.

[xiii]     K.L. Sharma, (ed), Caste and Class in India, Jaipur, Rawat Publications, 1994, p.75.

[xiv]      M.N. Srinivas, Cohesive Role., p. 74.

[xv]      M.N. Srinivas, ibid, p. 65.  Also see N.K. Dutt, Op.Cit., p. 70.

[xvi]     Marc Galanter, Competing equalities, New Delhi, OUP, 1984, p. 173.

[xvii]    Morton, Klass, Op.Cit., p. 129.  See for more information J.H. Hutton, Op.Cit., p.85.

[xviii]   Morton, Klass, ibid., p. 130.

[xix]     Morton, Klass, ibid, p. 118.

[xx]      N.K. Dutt, Op.Cit., p. 80

[xxi]     Marc Galanter, Op.Cit., p. 146.

[xxii]    Marc Galanter, ibid, pp. 145-50.

[xxiii]   Marc Galanter, ibid, p. 151.

[xxiv]    M.N. Srinivas, Caste in Modern India, p. 85.

[xxv]    Marc Galanter, Op.Cit., p. 115.

[xxvi]    K.L. Sharma, (ed), Op.Cit., p. 231.