Elsewhere on this site, Al-Biruni’s eleventh-century account settles that caste is not a British invention. Good — but that verdict invites the honest follow-up, and it would be a cheap essay that ducked it: if the British did not create caste, what did they actually do to it? The answer is: quite a lot. They did not build the house, but they drew the map that everyone now argues over.
The census made caste the master category
From the first all-India census of 1872, and with growing ambition through 1901, the colonial state did not merely count Indians — it classified them by caste, and made caste the central organising fact of official knowledge about India. Every person was to be assigned a caste; castes were to be ranked; the village itself came to be understood, in the official mind, primarily as a theatre of caste relations.
[Documented] Under H. H. Risley, the 1901 census went further and ranked castes using anthropometry — measuring noses and skulls — on the pseudo-scientific assumption that caste rank tracked race. The method was junk; the effect was durable.
Fluid became fixed
This is the change that matters, and Dirks’s Castes of Mind makes the strong case for it. Pre-colonial jati identities were local, overlapping, and movable — a community’s standing could rise or fall over generations, and varied from region to region. The census froze this fluidity into fixed, ranked, all-India lists. Caste became a thing you were, singular and official, recorded by the state.
[Documented] One telling consequence: communities began petitioning the census to be recorded at a higher rank, forming caste associations to lobby for a better official label. Caste had become a currency of the modern state — something to be claimed, contested, and mobilised, in a way the older, looser order had not required.
The state that governed by classifying
Dirks calls the colonial apparatus an “ethnographic state” — one that ruled by cataloguing its subjects, and for which caste was the favourite lens. The gazetteers, the Tribes and Castes volumes, the census schedules: an entire machinery for turning a messy social reality into a governable table of ranks.
[Debated] Here is where honesty requires a flag. Dirks’s strongest formulation — that colonialism essentially made caste as we know it — is contested, and this site sides with the critics on the strong version. Al-Biruni and the Dharmaśāstra prove the substance is ancient; caste was not conjured from nothing in a census office. The defensible claim is narrower and still important: the British enumerated, ranked, rigidified and politicised an existing order into its recognisable modern form.
The tell: they stopped counting
If the census had merely recorded a neutral fact, recording it would have been harmless. It was not. By the 1920s, caste enumeration was fuelling caste mobilisation and political demands, and after the 1931 census the British largely abandoned the general counting of caste because of the political fire it had lit. [Documented] You do not stop measuring a thing that measuring leaves untouched.
And independent India inherited the anxiety: its own census, for decades, counted only the Scheduled Castes and Tribes, not caste at large — a silence only now being reopened.
The enumerating impulse, though, never really stopped. The Anthropological Survey of India’s monumental People of India project set out to catalogue every community in the country, assembling a working list of several thousand at the outset. The instinct to reduce a living social reality to a countable table of communities — the census’s founding move — outlived the census itself. [Documented] (drawing on K. S. Singh’s People of India volumes — cited, not reproduced).
Why both essays are needed
Read this beside the Alberuni piece and the full, honest sentence emerges: caste is ancient in substance, and substantially colonial in its modern form. Those who say “the British invented caste” have to answer Al-Biruni. Those who say “the British changed nothing” have to answer the census. Neither slogan survives, which is exactly the point.
There is a live coda. The reservation system — the new code’s engine of justice — runs on caste lists, and those lists descend in part from the colonial census’s categories. The Constitution had to build equality using, among its tools, the very enumeration the colonial state had built to rule. The old code supplied the hierarchy; the colonial state supplied the catalogue; and the new code has to make justice out of both.