Dumont, Guha, and the Danger of Stereotyping India


“…the caste system should be seen as less ‘exploitative’ than democratic society. If modern man does not see it this way, it is because he no longer conceives justice other than as equality.”

These were words once written by Louis Dumont, a French scholar and the author of the famous Homo Hierarchicus: The Caste System and Its Implications, a book long considered as one of the most authoritative works on the caste system. It has also been heavily criticized, and I do not think it will retain its special position much longer. One of the recent publications that knocks down many of Dumont’s myths is Beyond Caste: Identity and Power in South Asia, Past and Present by Sumit Guha.

As this is The Diplomat’s Asia Life section and not an academic journal it may seem strange to compare two academic texts. But this piece is about contemporary Indian society and how it is being perceived from the outside. Dumont can serve as a warning to foreigners dealing with India in many contexts, from researchers to journalists, tourists to think tank analysts.

Enjoying this article? Click here to subscribe for full access. Just $5 a month.

The Bond of Bread and Daughter

In modern India, it would seem that many castes in a given region establish the proximity or separation between them by the rules concerning commensality and intermarriage, and the rules concerning the division of roles and professions (although professions are not so strictly divided between castes and inherited within castes in today’s India, particularly not in cities). In other words, the rules of a given caste establishes who can eat with whom and who can marry whom. In modern Hindi, a part of this system is described by the phrase roti-beti ka rishta: “the bond of bread and daughter.” These aspects were also noted by Dumont, who tried to depict the relations between castes in a locality using a few cases described in tables.

For Dumont, however, social supremacy in the caste system is closely associated with status, and status is highly related with ritual purity, although he never claimed these are the only important aspects. The notions of social status and purity are indeed visible in the rules of “bread and daughter”: many castes refuse to eat with certain other communities because they consider the others as ritually impure. So far, even many of Dumont’s critics would likely agree with him that in this regard the rules of purity and religious rituals are important in creating social bonds and divisions. What the French author had been putting special emphasis on, however, is that the Hindu religion, and the rituals and purity rules that come with it, are central to the establishment and functioning of the caste system. To put it in a very simplified way, many would have it that purity was one of the justifications to ostracize certain communities, a tool used by those who had power: power to dominate, to distribute roles and to ostracize others. For Dumont, however, purity was much more than this: it was a part of the backbone of the social order; it was one of the rules, and not just the justification for rules.

This is perhaps best shown in Dumont’s description of the jajmānī system. A member of the dominating, land-owning caste in a given village would pay a priest (Brahman) to arrange for certain religious rituals, sacrifices to gods. That person, in whose name the sacrifice was undertaken, is called a jajmān in modern Hindi, and hence the system around the person was called jajmānī. According to Dumont, many members of the dependent castes (which did not own land and had to serve the dominant castes) offered their particular services to the dominating caste family which arranged for the ritual. Their services were thus treated, according to the French social scientist, as payments in kind, as their repayment to the dominant cast for the sacrifices which the members of the dominants caste was arranging for. Seen this way, ritual had a central role in supporting hierarchy and distributing roles.

Dumont also made a few general and judging remarks that show his partiality with regard to the caste system. For example, he claimed that: (1) in the Hindu view, unlike the Christian one, an individual is a part of the world, (2) social hierarchy is natural for the societies worldwide, and hierarchy is needed today, and not only on the social level, and (3) the caste system should be seen as less “exploitative” than democratic society.

Many of Dumont’s observations were of significance and his book still should be perceived as an important part of research on Indian society. That said, however, it must be pointed out that he had formed a very essentialized perspective of the caste system. As in the process of metal casting, he was casting the castes into very rigid and seemingly changeless forms.

Dumont’s view has been taught for decades. Generations of Indologists graduated with Homo Hierarchicus as part of their curriculum, and inadvertently some of them looked at Indian society through Dumont’s lens, at least partially. But at the same time Dumont’s approach had been challenged right from the very start (and indeed, views contrary to his had been raised even before he formulated his theories). Certain academics have been ringing the alarm bells for years, pointing out, among others, that if you take up Dumont’s telescope and look at contemporary Indians, the stars do not add up to constellations; the neat and attractive model of the society partially falls apart.

The King is Naked

Sumit Guha’s book was one of the fatal blows dealt to an already embattled theory. While Dumont claimed that hierarchy functioned partially thanks to culture (as social status was based on it), Guha stresses the importance of power (political and economic) in hierarchy. This does not mean that Dumont completely ignored power or that Guha is completely ignoring culture. It’s only that in Guha’s case the balance of importance tilts from culture towards power.

Sumit Guha pointed out, among others, that jajmānī is not the dominant system around which the exchange of services takes place in a village society, and it does not seem as prevalent in previous centuries. Not only has the exchange of services transformed into a system of paid services, but the past exchanges were apparently not centered around the ritual. It does not mean that the Brahmans and the rituals were not important — but what Guha says is that they were not essential and central for the functioning of the exchange of services.

Guha similarly shows that the notion of purity was not as important as Dumont had claimed it was. In the priestly narrative, the idea of purity was especially significant in establishing the social position of Brahmans and Dalits (untouchables). The Brahmans were considered the purest, and the Dalits the most impure. Not every caste in India shares this view, however, and, contrary to Dumont’s tables of castes, these communities cannot be simply arranged in line according to their perceived purity and impurity.

The rules Dumont talked about were often negotiated and dynamic. “Society was never static in some ‘traditional’ mode, and the bitter contests for political power and social status that rend the villages today are not as novel as they may seem,” wrote Guha. One of the examples the author gives is of a Dalit autobiography which shows how, in a part of India, Dalits had to do some of the cleaning during marriages but were given the food leftovers as gratification for their work. The source captures the moment when a valiant Dalit cleaning women demanded from the host that her children were given a part of the regular meal from the marriage feast, and not the leftovers. Afterwards she and her family stopped collecting food leftovers during the upper caste marriage ceremonies. What we saw here was two people facing each other during a test of power; the notion of purity seemed to be important in the story, but it was not the only aspect, and jajmānī did not appear at all.

Dumont, Guha observes, collated various periods, regions and sources into nearly one vision of society. I could add here that this lack of chronology corresponded with Dumont’s focus on purity and ritual. If society is static, then the laws of purity can remain central. But if purity is one of the ways of justifying social domination, then it should mean the laws of purity can be changed of transgressed when the balance of power changes. At any rate, the more dynamic society is, the more status turns out to be a justification rather than the backbone of hierarchy.

There is No Homo Indicus. The Mistakes We Should Not Make.

So what were the mistakes Dumont made? First, his research was supposed to be anthropological but in practice it was mostly based on theoretical writings and on research conducted by other people. He was casting his eye on castes, and yet was mostly reading books, and not talking to Indians.

Books and articles are obviously important, but it is rather striking that his Homo Hierarchicus, considered a seminal work on the caste system, was written without conducting field research (although Dumont had done some earlier field research in India before writing his book). To be sure, Guha did not undertaken field research either, but he is a historian and he has been clear that he is recreating a historical social process, not researching contemporary Indian society.

Second, even Dumont’s use of Indian textual resources is quite limited. Apart from one novel and a little bit of classical Sanskrit treaties that can tell us some things about the ancient age, the bibliography of Homo Hierarchicus does not contain any positions written in Indian languages. The reference list also does not include any work in a contemporary Indian language, apart from one modern novel in Hindi (Karmbhumi by Premchand), which, curiously, Dumont quotes as an indication of contemporary social reality. Moreover, Western authors predominate over Indian ones in the bibliography.

Third, Dumont laid too much stress on the past as indicative of the present. For instance, his book was first published in English in 1970 (in French in 1966) and among its tables that show relations between castes one is based on a book that had come out in 1960 but another on a publication from 1931.

Fourth, Dumont wanted to show the entire caste system with its lower levels, and yet fell for the upper caste narrative. The whole focus on Brahmanic traditions and rituals as a bedrock of the social hierarchy does a great service to the conservative Brahmans. As they wanted to be considered as the highest class, they had created narratives and religious traditions that supported this view. But it does not mean that the Brahman supremacy was always uncontested and easily accepted.

It is a paradox because Dumont had sincerely tried to free himself from the Westernized view, to look at Indian society from the Indian perspectives. Eventually, however, he chose one of the Indian perspectives which was then essentialized and became a part of the Westernized view.

Fifth, Dumont seemed to wish everything in his view of the social order had been orderly. In his book, the castes appeared to fit in his neatly arranged tables. But social reality is not a puzzle — you cannot simply rearrange and assemble the pieces until you finally create a full picture with all the elements perfectly fitting each other. It is never so simple: social hierarchy is expressed on various levels, it is seen differently from various perspectives, it is often challenged and it keeps changing.

I think it is a common problem with the way we see the social world: we want to see society as orderly and predictive, we want it to be a like a clean house with all pieces of furniture in their places, but the reality always turns out to be more nuanced and complex.

Lastly, Dumont wanted to encompass everything, to create one theory of the caste system — while it may be debated if one caste system exists at all, maybe it would be better to speak of “caste systems.” Attempts at grand theories are certainly important for the progress of social science. But contemporary grand theories must be based on very exhaustive material, and this is increasingly difficult to achieve for a single researcher. Dumont’s work still enriched and energized debate on Indian society in many ways. But with this scope of material, his goal of creating a grand theory was perhaps too high to reach from the very start.

It is often better to satisfy ourselves with a very limited scope of material and research, to look for humble and not general conclusions, to dig one’s own well, rather than try to form an all-explaining and all-encompassing grand theory. Contrary to academic fashion, such strictly limited research can often turn out to be more accurate and valuable.

Has our approach to India changed since Dumont? In the case of some people and institutions – a lot; in the case of many others – not really. Like Dumont, many of us, so-called experts and researchers on India, prefer to read about Indians rather than talk to them; read in English rather than in an Indian language; speak of India and Indians, rather than of various communities and regions; project the history on the present because it allows to ignore change and offers more general conclusions; and attempt at grand theories or accept them rather than admit the existence of nuances and exceptions that do not fit into the theory.

Well, let us hope we get better, because every Louis Dumont will sooner or later face his Sumit Guha.

Sign up for our weekly newsletter
The Diplomat Brief