"A caste is an enclosed class": is caste a cage with no way in and no way out?

One of the commonest ways to explain caste is to compare it with class. A poor person can, at least in theory, become rich, and a rich person can go bankrupt. Caste, the comparison runs, is different. It is fixed at birth and kept in place by marriage within the group. You cannot buy your way into a higher caste or lose your caste by losing your money. It is a cage with no way in and no way out.

This explainer, the second of three, tests that idea against B. R. Ambedkar’s earliest work on caste, written when he was a 25-year-old student in New York.

What Ambedkar argued

In May 1916 Ambedkar read a paper to an anthropology seminar at Columbia University, later published as “Castes in India”. He set out to find the one thing that makes a caste a caste.

Classes are open; castes are closed. “Class and caste, so to say, are next door neighbours, and it is only a span that separates the two. A caste is an enclosed class” (p. 88). Societies always divide into groups, he wrote. “The unnatural thing about these sub-divisions is that they have lost the open door character of the class system and have become self-enclosed units called castes” (p. 91).

Marriage is the lock. Ambedkar tested the leading definitions of caste of his day and kept one feature: “endogamy is the only characteristic that is peculiar to Caste” (p. 84). Indian clans had long married outside the clan, so making castes meant “the superposition of endogamy on exogamy” (p. 84).

The lock had to be forced. A closed group, he argued, must keep its numbers of marriageable men and women equal, and so it disposed of the “surplus woman” and “surplus man”. Sati, enforced widowhood and girl marriage were, in his reading, the means: “Strict endogamy could not be preserved without these customs, while caste without endogamy is a fake” (p. 88). Our brief on the 1916 paper sets out that argument in full.

Some closed the door; others were shut out. Why did other groups become castes? “Some closed the door: others found it closed against them” (p. 91). Some copied the Brahmans, who closed themselves first. Others were enclosed “because others in enclosing, themselves have closed them out” (p. 94).

The idea it is paired with: open and closed classes

The comparison between caste and class is Ambedkar’s own. He accepted that “society is always composed of classes”, whether “economic or intellectual or social” (p. 88). What made caste different was not that it ranked people, which classes also do, but that it shut the door between ranks. The test of the “self-enclosing” idea is whether that door is really shut both ways.

Where the idea holds

No way in. Ambedkar wrote that the person who breaks caste rules is “left to their own fate without having the alternative of being admitted into or absorbed by other Castes” (p. 93). Excommunicated people “have been quite willing to be humble members of some caste (higher by preference) if they could be admitted within its fold” (pp. 93–94). The other castes would not have them.

Money does not open the door. Debayudh Chatterjee, in a 2022 study of India’s communist movement, argues that “while one’s class position could be changed with economic upliftment, one’s caste is not dynamic. A Dalit still remains a Dalit and faces social prejudice even if they are able to climb up the class ladder” (p. 62). That was his charge against the communist parties, which treated Dalits as one more section of the poor.

Marriage is still policed. The rule Ambedkar found at the heart of caste is the one that village assemblies and families still enforce with violence, as the Supreme Court found in 2018 (part 1 of this series sets out the judgment). Ambedkar himself came to see intermarriage as the real solvent: “The real remedy for breaking Caste is inter-marriage. Nothing else will serve as the solvent of Caste” (Annihilation of Caste, §20 [5]).

Where the idea breaks

There is a way out, but it leads into another cage. Caste status can be lost. Ambedkar described exactly how: “Castes have no mercy for a sinner who has the courage to violate the code. The penalty is excommunication and the result is a new caste” (p. 93). The outcast is not freed from caste; he is enclosed again, in a new group. This is how, in Ambedkar’s account, castes multiplied “to a bewildering multiplicity” (p. 94). The cage is less a single locked room than a machine that keeps building rooms.

Leaving the religion is a door, of a kind. Ambedkar ended Annihilation of Caste in 1936 by announcing, “I have decided to change” (§26 [3]), and in 1956 he converted to Buddhism with his followers. But the state and society do not always treat conversion as an exit. When a Tamil Nadu convert to Islam sought certification in a Muslim backward-class community, he was told, in the authorities’ words recorded by the Supreme Court in October 2026, that he had “only converted to religion and not to the caste”. Under the 1950 Scheduled Castes Order, a Dalit who converts to a religion other than Hinduism, Sikhism or Buddhism loses Scheduled Caste status in law. The law treats the convert as having left that status behind; whether the convert leaves caste behind in daily life is a separate question, which the law does not answer.

Marriage across caste happens. It is rare, and it can be met with violence, but it is legal and it does occur, and Ambedkar saw it as the way to break caste, not as impossible. The enclosure is a rule that is enforced, not a law of nature.

Ambedkar’s own word is “enclosed”, not “enclosing”. He stressed that caste is made and kept by people: castes were “founded and maintained through these unnatural means” (p. 89), and “Caste, as I have explained it, is almost impossible to be sustained: for the difficulties that it involves are tremendous” (p. 94). A cage that must be constantly rebuilt can also be broken.

Why it matters

• Against the “class first” view. If caste were only an extreme form of class, rising incomes would end it. Ambedkar’s and Chatterjee’s point is that they do not.
• Law and conversion. The rules on who keeps Scheduled Caste status after conversion assume that caste stays behind when a person changes religion. Whether caste in fact stays behind is a question this piece does not settle; the published arguments of Dalit Christian and Dalit Muslim scholars are not yet read for this series.
• Women at the centre. In Ambedkar’s account, the enclosure is held by controlling women’s marriage and widowhood. Any account of caste that leaves out gender misses how the lock works.

The fine print

• Our framing. “Self-enclosing” is our word for the series. Ambedkar wrote of castes as “enclosed” and “self-enclosed units”.
• A theory of origins. The 1916 paper explains how caste began and spread. It is not a survey of how people live now, and it rests on older texts and scholarship.
• What we have not read. We have not read the survey research on intercaste marriage or on caste mobility, including Sanskritisation, for this piece.

Talking points

• If caste can be lost by excommunication but only into a new caste, is it a cage with no way out, or a system that turns every exit into a new cage?
• Should a Dalit who converts keep Scheduled Caste status? What does your answer assume about whether caste can be left behind?
• Ambedkar saw intermarriage as the real remedy, then concluded that people would not intermarry while the Shastras held sway. Which comes first: changing marriage, or changing belief?

The works

• B. R. Ambedkar, “Castes in India: Their Mechanism, Genesis and Development”, paper read at Columbia University, 9 May 1916; Indian Antiquary 46 (May 1917), pp. 81–95. Project Gutenberg transcription, cited by the journal’s page numbers.
• B. R. Ambedkar, Annihilation of Caste, 1936; third edition, 1944. Online edition by Frances W. Pritchett, Columbia University. Cited by section and paragraph.
• Debayudh Chatterjee, “Class versus Caste: The Conundrum of Dalit Politics and the Communist Movement in India (1926–2016)”, in Organizing Equality, ed. Alison Hearn and others (McGill-Queen’s University Press, 2022), pp. 54–71. Our brief
• Supreme Court of India, The Secretary to Government v. Sameer Ahamed N. and Others, 2026 INSC 1129, judgment of 15 October 2026.