Skip to navigation | Skip to content

Setting Caste Back on its Feet

Sumit Guha’s Beyond Caste is the most important synoptic study of caste since Louis Dumont’s Homo Hierarchicus.  Guha is an historian, not an anthropologist, but anthropologists should take note.  He has marshaled a vast array of evidence drawn from native and pre-colonial sources, rather than the more conveniently accessible colonial reconstructions that Dumont and others depended on, along with an up-to-date reading of historical literatures few anthropologists are aware of, to powerfully challenge both popular and anthropological common sense on the topic.

According to the common sense position, which Guha sees as having fully displaced earlier and more accurate understandings only in the twentieth century, caste is fundamentally a religious phenomenon (p. 20).  It is a ritual order, intrinsic to brahminical Hinduism.  But although only Hinduism offers theological justification of caste, caste is practiced throughout South Asia by all people irrespective of religious affiliation (Government of India 2007; see Viswanath 2015b).   The fact that Muslims and Christians in India are also divided into endogamous jātis has generally been sidelined, however, in ad hoc fashion as evidence of Hindu “influence” by those committed to a religious explanation of caste (e.g. Dumont 1980: 201–16).  Guha argues this puts the cart before the horse.  While it is true that caste differences have been culturally marked in a variety of ways, including through ritual notions such a purity and pollution, on Guha’s account, purity is better understood as just one among several idioms—another arguably more prominent one being that of honor—through which social precedence is expressed (p. 9; cf. van der Veer 1993: 34–5).  Precedence itself is not determined by such idioms, but by complex processes and conscious stratagems resting ultimately on power and domination.

There is nothing wrong with studying the liturgical orders through which social precedence was expressed and celebrated, or with attempting to reduce the complexities of the ideological texts of Brahminical Hinduism to a singular principle à la Dumont.  But one should not mistake the result of such an exercise for the very essence of caste as such.  At best it would give us only a distillation of one set of Hindu theological rationales for caste, not its sociological reality. Guha, indeed, questions whether caste even has a singular essence, a point to which we will return.  But the error of substituting a hermeneutics of Brahmin ideology for sociological analysis should be easy to grasp.  For as Guha points out, there is no caste (save perhaps the priestly Brahmin) whose ritual role is primary (p. 13).  Many castes have no ritual function at all, and among those that do, this function is a secondary one.  Even the “untouchables”—often posited by those who subscribe to a ritually determinist theory of caste as the structurally antithetical complement of the Brahmin—have never been defined primarily by their ritual function of removing impure substances.  The majority of untouchables (or Dalits, as they are now called) have never performed such functions, and among the minority that did perform ritually impure functions this was always a secondary occupation for all but a few (Rawat 2011).  Historically speaking, most Dalits were not ritual specialists, but hereditarily unfree agricultural laborers, slaves.  This is especially so in India’s wet rice-growing civilizational heartlands, where brahminical Hinduism was at its most hegemonic, the caste system most elaborate, and untouchability most intensely marked (Kumar 1965; Viswanath 2014)—exactly, in other words, where the ritual interpretation of caste should work best.

Contemplating this anomaly I once wondered, “Were ‘untouchables’ consigned to a life of hard agricultural labor on account of [a prior ritual] impurity, or was being coded impure and assigned polluting tasks simply part of what it meant to be under the total domination of others?” (Roberts 2008).  Guha does not focus on Dalits, but the account he provides makes clear that caste is not the outcome of a prior ritual logic, but a political and economic one.  Culture is not absent from Guha’s account, but his notion of culture is not that of “a reified entity, ‘above’ the fray of daily life, which somehow produces behaviour,” (p. 11). Representations are themselves a part of social reality, which enters into and reciprocally shapes, non-discursive aspects.  But symbolic markers of status (such as “purity”) are like any other resource insofar as they are not equally available to all and their use is rigorously policed (p. 13).  Culture does not, in other words, stand on its own.  Discursive systems are inherently susceptible to conflicting interpretations, and even the most rigorously formulated formal language cannot supply the rules for its own interpretation.  Ritual principles are applied only in a highly selective matter, and in ways that suit the interests of those in power.  Thus when castes have attempted to elevate themselves by adopting more “pure” practices, their efforts have been met with violence (p. 13), not celebration, by superiors who allegedly value purity over all things.  Where such attempts have succeeded, it is not because the upstarts more perfectly emulated the superior group’s cultural forms, but because they had already achieved a measure of economic and political power (Srinivas 1956; 1959).  Purity is therefore best understood as an idiom through which status is expressed, not its essence, as accounts that see caste as a ritual order would have us believe.

Is this really news?  After all, the example I just gave was formulated in a widely-cited article first published by M.N. Srinivas in 1956.  The importance of Guha’s argument is not just in the new historical data he brings to bear—of which there is a great deal—but in the new light this sheds on the best anthropological research of previous generations.  In the case of Srinivas, for example, Guha’s account allows us to grasp implications of the former’s arguments that Srinivas himself was unable fully to see.  For despite showing that power ultimately determined ritual status, not the reverse, Srinivas nevertheless persisted in seeing phenomena like the control of land and servile labor, merchant capital, the state, and sheer physical dominance as “secular criteria” that could “influence” caste, but were not intrinsic to it.  Srinivas, like the majority of anthropologists and lay commentators to this very day, remained imprisoned by the twentieth-century common sense that conceptualizes caste as a Hindu ritual order.

How did the current, Hindu-centric view of caste attain the status of common sense?  Guha warns readers that his book does not purport to be a history of anthropological thought (p. 16).  But he does offer some clues as to how this came to be.  Significantly, early European and West Asian observers did not perceive caste as a Hindu phenomenon at all.  Following indigenous usage they understood castes as bounded kin and political units existing among not just Hindus, but Muslims, Christians and other religions, as well as Dalits (who at that time had not yet been incorporated within Hinduism [Roberts 2015]).  For most of the period of British rule this older understanding prevailed.  Knowledge of local society was based on practical experience and direct contact with native leaders at the local level.  Citing Bernard Cohn and Thomas Trautman, Guha explains that it was only in the late 19th century that this early “field view” come to be replaced by a “book view” of caste, which privileged Brahminical Sanskrit texts (p. 32) and erroneously assimilated caste (jāti) to the defunct Hindu notion of varṇa (a theoretical division of ancient Aryan society into priests, warriors, merchants, and farmers).  The prioritizing of Hindu theological justifications of caste was ironically given a boost by Christian missionaries, who subscribed to a kind of religious determinism that pictured caste as a Brahmin conspiracy (see also: Viswanath 2010; 2014).  This was partially challenged by the rise of village ethnographies in the 1950.  These ethnographies provided important insights into how caste actually operates at a local level, and the best village studies highlighted the importance of political-economic power.  At the same time, the village ethnographers retained the late nineteenth-century baggage that had redefined caste as a Hindu religious phenomenon.  They also reproduced the orientalist assumption of an unchanging “traditional India,” which they believed had endured at the level of the allegedly “self-sufficient” village that had remained unchanged despite the comings and goings of kings (p. 35).  The pendulum swung back even more fully to the orientalist view with Louis Dumont.  Dumont’s account was, of course, challenged in the decades that followed (e.g. Marriott 1969; Lynch 1977; Dirks 1987; Bayly 1992).  But with few exceptions Dumont’s critics failed to question the basic assumption that caste was Hindu in essence.

For Guha, the most significant breakthrough of the post-Dumontian era was made by Nicholas Dirks (1987) and Susan Bayly (1992).  Both scholars were able to show that the caste order was indeed fundamentally a political one, whose terms were set by the state.  It is this key insight that Guha radically extends.  The problem with Dirks, on Guha’s account, is that he failed to break with the modern commonsense that sees caste as Hindu.  Only instead of centering this allegedly Hindu social order on the Brahmin, Dirks saw it as centering on so-called Hindu kingship.  Bayly’s breakthrough was in her rediscovery of the fact that not just Hindus, but, equally, Muslims and Christians participated in the “politically structured system of honor and patronage” of caste (pp. 37–8).  Yet Bayly, too, is charged with reverting to the Hindu-centric view in her later work (1999), where, despite programmatically addressing the significance of caste in non-Hindu communities she sometimes seems to “fall back into an idiom of caste status centered on purity and exaltation of and by Brahmans, which can only apply in broadly Hindu ritual contexts” (p. 38).

Guha’s own explanation of caste is a complex one, and it is not possible for me to do justice to its nuances here.  The basic idea, however, is that castes are both directly constituted by states, and, indirectly, when military brotherhoods, “tribes,” and other corporate groups bound by kin ties respond to the threats and opportunities of state power by establishing themselves as dominant castes over village clusters.  When they could not directly challenge kingly authority, these dominant groups entered into feudatory relations with kings, who in turn used them as local instruments of state authority and tax collection.  While kingdoms and particular castes come and go, the broad pattern of relations reproduced itself.  Guha here builds on Burton Stein (1969; 1985), whose work covers a period stretching from the ninth to the nineteenth centuries, which he shows applies across South Asia and not just in southern India where Stein himself worked.  Guha, however, follows other specialists in pointing out that Stein himself seriously underestimated the extent of state penetration into the locality and the importance of its tax-collecting apparatus (p. 49).

The same basic pattern applied not only to Hindu kings, but also Muslim and Buddhist ones, and was adopted as well by the Portuguese, Dutch, and finally the British.  And it did not matter to rulers whether the corporate bodies through which they managed their kingdoms were Hindu or Muslim.  The old order began to break down only in the latter days of British colonial rule, via a complex set of processes that turned on the arrival of print and the new self-conception of Indians as “putatively equal citizens of an ethnic nation that aspired to statehood” whose political significance depended in part on numbers (p. 176).  The story Guha tells of the new demographic politics is broadly similar to the one told by Cohn (e.g. 1987) and countless others, but with some key differences.  Guha does not see this transformation as a sudden rupture in which a “blissful continuum of premodern collective life,” was “suddenly and arbitrarily sliced up by colonial modernity” (p. 164).  Caste was a competitive, power-laden institution from the very beginning.  Nor was the colonial census entirely new.  Drawing on his own earlier research (2003), and that of scholars like Norbert Peabody (2001) and C. A. Bayly (1996), Guha shows that caste-wise censuses were a well-known and important tool of rule in pre-colonial kingdoms.

These censuses did not, as Arjun Appadurai (1993) has claimed, pertain only to taxation and not caste identity.  While it is true that pre-colonial censuses were primarily concerned with the practical matter of taxation, Appadurai’s distinction, Guha argues, is based on an anachronistic understanding that treats corporate identity and economy as distinct, and in which “taxation is run by the Treasury and social regulation is the business of Health and Human Services” (p. 163).  Importantly,  pre-colonial censuses substantiate Guha’s core claim that caste was not a Hindu institution: they did not distinguish between Hindu or Muslim castes, nor did they conceive of the population as divided into competing religious blocs of Hindus and Muslims.  For as Norbert Peabody has shown, pre-colonial census takers did not perceive society as divided into horizontal communal blocs (i.e. Hindu and Muslim) but between so-called “clean” castes, which included both Muslims and Hindus jātis, on the one hand, and a subordinate population of servile castes, on the other.  It was only under the first colonial census that the Hindu–Muslim distinction displaced this earlier, caste/class division (Peabody 2001: 834–6), with Dalits thrown arbitrarily into the “Hindu” category (Roberts 2015; Viswanath 2015a).

Apart from categorizing populations on caste lines for purposes of taxation, financial imperatives were central to a second way in which states organized and regulated populations on caste lines.  Drawing on a rich textual record of quotidian statecraft from the pre-colonial period, Guha shows that the state was deeply involved in regulating the internal affairs of castes through the institution of dispute resolution.  Caste is fundamentally a structure of actual and potential kin, and therefore of inheritance.  Inheritance disputes were as ubiquitous in pre-colonial South Asia as they are today, and in regulating them the state involved itself directly in the internal workings of corporate caste groups.  This benefited the state in two ways.  Adjudicating such conflicts was a significant source of income for the state, since settlements invariably involved fines (pp. 130–5).  It also provided a means for the state to exert influence over dependents, extending down, through a series of intermediary office holders, even to families of the lowest, leatherworking castes (pp. 130, 135).  In short, state-imposed institutions of dispute resolution functioned in pre-colonial India as “structures of fiscal and socio-political surveillance and exploitation” that constituted caste, even at the domestic level, as a key node in the political economy of South Asia (p. 119).

For anthropological readers, the third chapter of Beyond Caste, on the supposedly mutualistic and self-organizing village caste system of yore, will be of special interest.  I therefore devote the remainder of this review to a more detailed summary of its argument.  A convenient example of the position Guha speaks against is that of Walter C. Neale, who envisioned the Indian village as held together, not by force or even calculations based on personal interest, but a shared belief system.  In Neale’s words, there ‘was no bargaining, and no payment for specific services rendered. There was no accounting … and the whole produce was easily and successfully divided among the villagers’” (p. 94, quoting Neale 1957: 226).  Neale is not alone in this view.  To illustrate his point, Guha could have quoted any number of anthropologists, such as Louis Dumont, who characterizes intra-village exchange in the Maussian language of “prestations and counter-prestations” (1980: 98–9).  The image Neale, Dumont and others rely on here is that of the fabled “grain heap” (cf Fuller 1989).  In Dumont’s words

[D]istribution on the threshing floor is essentially different from a market in that it takes place in virtue of the fact that everyone is interdependent… [W]e are not in the world of the modern economic individual, but in a sort of co-operative where the main aim is to ensure the subsistence of everyone. (Dumont 1980: 105)

Exploitation cannot exist in a caste society, according to Dumont, because, he believed, “an economic phenomenon [like exploitation] presupposes an individual subject,” whereas in caste society, “everything is directed to the whole… as part and parcel of the necessary order” (1980: 107).

In attacking the myth of the grain heap, Guha cites and builds upon the arguments of C. J. Fuller (1989), Thomas O. Beidelman (1959) and Peter Mayer (1993) (see also: Breman 1988).  But the pre-colonial sources Guha uses allow him to take the argument a step further.  Sumit Guha shows that the systems of village redistribution described by theorists from Maine to Marx to Gandhi (all of whom rely on 19th century colonial official Charles Metcalf) as evidence of village India’s autonomous self-organization—and a bulwark against state power—was in fact the product of state intervention.  Insofar as villages were corporate it was in order to facilitate revenue collection.  Far from existing prior to and outside the state, the internal organization of the Indian village—and indeed, the village itself—was largely constituted by pre-colonial Indian states (p. 86).  While village organization varied tremendously by region, and in time, Guha presents data strongly suggesting that those villages that most closely approximated the “ancient” form first posited by Metcalf were in fact those most determined by pre-colonial states (p. 112).  Where the state was weak or absent, villages were not integrated corporate entities and lacked “traditional” jajmani-type systems of intra-village exchange.  What had appeared to colonial observers and Gandhians as a source of resistance to state power was in fact its creature.  Insofar as villagers pooled grain in a “heap” for later distribution, it was because such arrangements had been mandated by revenue collectors to simplify the extraction process: “The common village threshing floor and grain heap sometimes seen… as the epitome of mutuality and harmony was, in fact, created by the need to satisfy the landlord or the king’s officer” (pp. 88, 91, 113).  And insofar as villagers cooperated with one another, it was not out of a primordial “community” solidarity, but because they were subject by precolonial states to collective punishment (pp. 85, 88).

What have been described as “share rents” were thus, in fact, a source of great exploitation (p. 94).  And even in the source Neale based his entire argument upon, clear evidence is available to show that the size of shares was not “fixed” by tradition, but determined by economic forces.  Preferential shares  (characterized by generations of anthropologists and colonial officials as “customary”) were in fact granted and determined by the state, Guha shows, and obtaining such shares was a matter of fierce competition.  In other words, “the jajmani system is… merely one example of widespread phenomenon—rent-seeking and the investment of resources in creating permanent, heritable sources of rents” (p. 112).  Apart from shifting alliances between individual village elites and pre-colonial states there was also “constant, if less dramatic, dickering over the quantity and quality of payments” (p. 107).  “Thus even where the village servant system had official recognition and support, it was continually restructured according to calculations of individual advantage” (p. 111).  Summarizing his argument on the Indian village, Guha writes

Individualized jockeying for social and economic advantage was present at all times. It will be evident that I am deeply skeptical of attempts to trace socio-economic institutions to fundamental values. I have found considerable evidence to suggest that individuals systematically sought to modify and invent customs and institutions to their own perceived advantage, and that the patrimonial and early colonial state tried to derive fiscal and political advantage from these efforts. Intercaste relationships and the idyllic village itself existed through the constant contest of unequal powers. (p. 116)


I began this review by characterizing Guha’s Beyond Caste as the most important synoptic work on the topic since Louis Dumont’s Homo Hierarchicus.  As will be obvious to the reader by now, the two works represent diametrically opposed approaches.  Dumont’s visionary claims were admittedly extreme, but the underlying picture he sketched of caste as a Hindu, and therefore uniquely Indian, institution is representative of most anthropological work on caste.  For Dumont, Indian society had to have a unique essence, and that essence was caste.  Dumont wrote at a time when detailed accounts of how caste actually worked were threatening to dissolve its supposed uniqueness.  In the village ethnographies on which he drew, caste was beginning to appear far less exotic than earlier orientalist and missionary writings had made it out to be.  Rather than being the outcome of religious ideology, the data these studies had accumulated showed its features to be based on elements and processes not inherently different from those anthropologists had found elsewhere in the world.  If the empirical approach of village ethnographies were “logically carried out,” Dumont warned in 1958, “we should have to pretend … that India is a mere geographical entity [i.e., lacking a singular cultural essence] similar to Africa” (p. 50).   But to place caste within a broad continuum of social organizational forms is not, in reality, to deny any coherence to India as a civilizational area.

By titling his book Beyond Caste, Guha does not, on my reading, seek to abolish the term caste from our analytic vocabulary, but merely to signal his argument that “the effort to find a single, unified rationale for the internal workings and external relations of each of India’s thousands of [castes] is futile” (p. 1).  In the face of the evidence Guha provides, how might a Dumontian respond?  Herein lies the subtlety of Dumont’s argument.  For what he would likely say is that his account was never intended to describe sociological reality.  Its purpose is rather to fathom the specifically Hindu ideology through which the messy realities of everyday life and struggle are rendered coherent for those who participate in them.  Indians fully recognize that social life is in fact determined by struggles for power and calculations of interest.  It is simply that, according to Dumont, Indians themselves do not regard such matters as being of any importance (Lynch 1977; Roberts 2008).

The central weakness of Dumont’s argument is that it makes very strong claims about how natives really think.  Determining other people’s thoughts is a complex matter that requires, at a minimum, consideration of what they themselves have to say about it.  Yet the voices of living subjects are nowhere to be found in Dumont’s writings on caste.  His claims about native consciousness are based entirely on extrapolating from his own theoretical model, an abstraction of certain formal features of Hindu ritual behavior and from Sanskritic texts that were known to only a tiny fraction of inhabitants.  Further weakening his case, we have abundant ethnographic evidence that people did in fact see power struggles, and so on, as important, and that at least some Indians (e.g. Dalits) explicitly reject the views Dumont attributes to them.  But until now it has been possible for those who believe the religious, non-conflictual interpretation of caste to dismiss such discrepancies as the result of an historical rupture—the breakdown of a pre-colonial mutualism under the onslaught of modernity.  Evidence of harmony and consensus in the present day is thus taken as providing a direct window into the pre-modern mentality; evidence of the opposite, a measure of historical change, “a recent departure from traditional mutualism and ancient harmony” (p. 101).  The historical data Guha provides will make this line much more difficult to maintain.


References cited

Appadurai, Arjun. 1993. Number in the Colonial Imagination. In Orientalism and the Postcolonial Predicament: Perspectives on South Asia. Carol Appadurai Breckenridge and Peter van der Veer, eds. Pp. 314–339. University of Pennsylvania Press.

Bayly, C. A. 1996. Empire and Information: Intelligence Gathering and Social Communication in India, 1780-1870. Cambridge University Press.

Bayly, Susan. 1992. Saints, Goddesses and Kings: Muslims and Christians in South Indian Society 1700-1900. Cambridge: Cambridge University Press.

—————. 1999. Caste, Society and Politics in India from the Eighteenth Century to the Modern Age. Cambridge: Cambridge University Press.

Beidelman, Thomas O. 1959. A Comparative Analysis of the Jajmani System. Locust Valley, NY: J.J. Augustin.

Breman, Jan. 1988. The Shattered Image: Construction and Deconstruction of the Village in Colonial Asia. Dordrecht: Foris Publications.

Cohn, Bernard S. 1987. The Census, Social Structure and Objectification in South Asia. In An Anthropologist among the Historians and Other Essays Pp. 224–254. New York: Oxford University Press.

Dirks, Nicholas. 1987. The Hollow Crown: Ethnohistory of an Indian Kingdom. Cambridge: Cambridge University Press.

Fuller, C. J. 1989. Misconceiving the Grain Heap. In Money and the Morality of Exchange. Jonathan Parry and Maurice Bloch, eds. New York: Cambridge University Press.

Government of India. 2007. Report of the National Commission on Religious and Linguistic Minorities. New Delhi: Ministry of Minority Affairs.

Guha, Sumit. 2003. The Politics of Identity and Enumeration in India C. 1600–1990. Comparative Studies in Society and History 45(1): 148–167.

Kumar, Dharma. 1965. Land and Caste in South India. Cambridge Studies in Economic History. Cambridge: Cambridge University Press.

Lynch, Owen M. 1977. Method and Theory in the Sociology of Louis Dumont. In The New Wind: Changing Identities in South Asia. Kenneth David, ed. Pp. 239–263. The Hague: Mouton.

Marriott, McKim. 1969. Stratification. American Anthropologist 71(6): 1166–1175.

Mayer, Peter. 1993. Inventing Village Tradition: The Late 19th Century Origins of the North Indian “Jajmani System.” Modern Asian Studies 27(2): 357–393.

Neale, Walter C. 1957. Reciprocity and Redistribution in the Indian Village: Sequel to Some Notable Discussions. In Trade and Market Exchange in Rural India. Karl Polanyi, C. M. Arensberg and H. W. Pearson, eds. Pp. 218–236. New York: The Free Press.

Peabody, Norbert. 2001. Cents, Sense, Census: Human Inventories in Late Precolonial and Early Colonial India. Comparative Studies in Society and History 43(3): 819–850.

Rawat, Ramnarayan S. 2011. Reconsidering Untouchability: Chamars and Dalit History in North India. Indiana University Press.

Roberts, Nathaniel. 2008. Caste, Anthropology of. In International Encyclopedia of the Social Sciences. 2nd edition. William S. Darity, ed. Pp. 461–463. New York: Macmillan Reference USA.

—————. 2015. From Village to City: Hinduism and the “Hindu Caste System.” In Handbook of Religion and the Asian City: Aspiration and Urbanization in the Twenty-First Century. Peter van der Veer, ed. Pp. 237–253. Berkeley: University of California Press.

Srinivas, M. N. 1956. A Note on Sanskritization and Westernization. The Far Eastern Quarterly 15(4): 481–496.

—————. 1959. The Dominant Caste in Rampura. American Anthropologist 61(1): 1–16.

Van der Veer, Peter. 1993. The Foreign Hand: Orientalist Discourse in Sociology and Communalism. In Orientalism and the Postcolonial Predicament: Perspectives on South Asia. Carol Appadurai Breckenridge and Peter van der Veer, eds. Pp. 23–44. University of Pennsylvania Press.

Viswanath, Rupa. 2010. Spiritual Slavery, Material Malaise: “Untouchables” and Religious Neutrality in Colonial South India. Historical Research 83(219): 124–145.

—————. 2014. The Pariah Problem: Caste, Religion and the Social in Modern India. New York: Columbia University Press.

—————. 2015a. Silent Minority: Celebrated Difference, Caste Difference, and the Hinduization of Independent India. In Routledge International Handbook of Diversity Studies. Steven Vertovec, ed. Pp. 140–150. New York: Routledge.

—————. 2015b (forthcoming). Commissioning Representation: The Mishra Report, Deliberation, and the Government of the People in Modern India. South Asia: Journal of South Asian Studies 38(3).

Please join our mailing list to receive notification of new issues