How to Break a Civilization

What This Essay Describes

There is a way of defeating a people that requires no permanent army.

You do not need to hold their forts. You do not need to garrison their cities forever. You need only to reach the generation that has not yet formed its convictions, and hand them an account of their own past written by someone who requires them defeated. Do this for forty years and the work is finished. The people will then defend the story of their own worthlessness with more energy than they ever defended their temples — and they will go on defending it long after the last foreign soldier has sailed home.

The method has been named in our own century, by defectors describing what their services did to other countries. But the naming came late, and the name matters less than the anatomy. What matters is that this is a procedure: it has stages, the stages run in order, each one makes the next one cheaper, and every stage is documented in the correspondence, the minutes and the published policy of the men who carried it out.

What follows is the procedure as it was run in Bhārata, in the order in which it was run. Every element that is usually discussed as a separate grievance — the persecution of the Brāhmaṇa, the abolition of Saṃskṛta, the destruction of the village school, the invention of caste, the ridicule, the manufactured leadership — is not a separate grievance. Each is one stage of a single operation, and none of them makes sense in isolation.

To see why the operation was necessary at all, you must first see what it was aimed at.

What Was There

Hold the year 1750 in your mind, and be specific about it.

Bhārata produces roughly a quarter of everything manufactured on earth. Her textiles clothe the world from Java to the Atlantic ports; Dhaka muslin is woven to a fineness that European machinery will not match for another century and a half. Her steel — wootz, crucible-forged — is the raw material of the finest blades in the Islamic world, and will shortly sit on Michael Faraday's laboratory bench in London, studied and not replicated. Zinc has been distilled at industrial scale at Zawar in Mewar for five hundred years, the first place on earth where this was ever done. Her shipwrights at the Bombay Dockyard build teak vessels so superior that the Royal Navy commissions them; one launched in 1817 is still afloat today in an English harbour.

Her surgeons perform nose reconstruction by a pedicle flap technique set down in the Suśruta Saṃhitā. In 1794 an English observer writes the operation up in a London magazine; a British surgeon learns it from that account in 1814; modern plastic surgery begins. Her physicians practise smallpox inoculation as a settled, organised, seasonal art, described in detail to the Royal College of Physicians in 1767 by a Company official who had stood and watched it work.

Her mathematicians in Kerala derived infinite series expansions for pi, sine and cosine in the fourteenth and fifteenth centuries — two to three hundred years before Newton and Leibniz — and set them down with demonstrations in the Yuktibhāṣā. Her grammarians produced, in Pāṇini's Aṣṭādhyāyī around the fourth century BCE, a complete generative description of a natural language in some four thousand rules: a formal system so far ahead of anything else on earth that twentieth-century linguists and computer scientists would recognise in it the ancestor of their own notation.

And across her villages there is a school. Not a few schools. A school in the ordinary village, in tens of thousands of villages, paid for out of local revenue and community endowment, teaching writing, arithmetic, accounts and the local literary tradition — attended by children of every community, and in which, across large parts of the country, Brāhmaṇa children are a minority of the students.

Now hold 1947 beside it. Literacy at about sixteen per cent. Life expectancy in the twenties for decades. Manufacturing share under two per cent. Famine on a scale with no parallel in the subcontinent's recorded history — tens of millions across a century and a half, including three million in Bengal four years earlier, in wartime, while grain moved out of the province. The village schools gone. Saṃskṛta reclassified as a dead language.

And more than a million Indians shipped overseas under indenture between 1834 and 1917 — to Mauritius, British Guiana, Trinidad, Natal, Fiji, Malaya — to cut cane in other men's colonies, under a word the world learned to use for them: coolie. The descendants of the people who built the ships and wove the muslin were exported as field labour. Settled

That is the distance to be explained. Two hundred years. And the economic ruin, enormous as it is, is the smaller half of the story — because wealth can be rebuilt, and India is rebuilding it. What was done to the Indian mind has not been undone, because the instrument that would have to do the undoing is itself the thing that was captured.

Here is how it was done, in order.

Stage One: Seize the Account

The first move is to take control of how a people's past is described, and to make your description the one their own rulers are trained on.

In 1817 James Mill published The History of British India in three volumes. He had never been to India. He knew no Indian language — not Saṃskṛta, not a single vernacular. He regarded both facts as advantages, on the ground that distance conferred objectivity.

His conclusion was that Hindu civilization was rude, barbarous, priest-ridden and low; that its literature was worthless; that its science was borrowed; that its people were incapable of governing themselves; and that its history properly began with the arrival of Europeans.

The book became the standard text at Haileybury, the training college of the East India Company's civil service. For generations, the men who administered Bhārata arrived having been formed by a book about India written by a man who had never seen it and could not read a word of its languages. Settled

Mill did one thing more, and it is the most consequential act of intellectual sabotage in Indian history. He divided India's past into three periods: Hindu, Muslim, British. Not Mauryan, Gupta, Cōḷa, Vijayanagara. Not by dynasty, region or epoch — by religion. At a stroke, a continuous civilizational history was reorganised into a sequence of religious occupations, in which the Hindu period was the primitive substrate, the Muslim period an equivalent and parallel civilization, and the British period the arrival of reason.

He did not work alone. Charles Grant's Observations on the State of Society among the Asiatic Subjects of Great Britain (1792) had already supplied the missionary template: the Indian was depraved, and his depravity was the work of his religion and its priests. William Ward's compendium of 1811 industrialised the genre.

And the argument carried a structural necessity, which is why no volume of contrary evidence ever dented it. A missionary enterprise that is failing must account for its failure, and it cannot concede that the tradition it is failing against is intellectually serious. So the explanation had to be a conspiracy — a cunning sacerdotal class holding an otherwise willing population in bondage.

Stage Two: Sever the Nerve

Having established that the priest is the enemy, you may now attack the thing he carries — and you attack it in the same motion, because they are not two targets.

On 2 February 1835 Thomas Babington Macaulay wrote his Minute on Indian Education. Its arguments are famous. Its admissions are what matter. He conceded that he had no knowledge of Saṃskṛta. He then declared that a single shelf of a good European library was worth the whole native literature of India. He proposed to form a class of interpreters between the rulers and the ruled — persons Indian in blood and colour, but English in taste, in opinion, in morals and in intellect. Lord William Bentinck adopted the position by resolution on 7 March 1835. Settled

What that resolution did in practice: it ended state support for instruction in Saṃskṛta. The stipends that sustained students of Saṃskṛta learning were discontinued. Public money for education was redirected wholly to English-medium instruction in European knowledge. The Orientalists — H. H. Wilson and others, who argued for supporting Indian learning on its own terms — lost the argument, and lost it to a man who had read none of the material in dispute.

Why Saṃskṛta, and why this early in the sequence?

Because Saṃskṛta was not a liturgical relic. It was the operating language of Indian intellectual life — the medium in which grammar, logic, jurisprudence, astronomy, mathematics, medicine, poetics, metaphysics and statecraft were argued across the entire subcontinent; the one language in which a scholar in Kaśmīra and a scholar in Tañjāvūr could dispute a point of Nyāya and both be understood.

It was the interoperability layer of the civilization.

And serious composition in it was vigorous into the seventeenth and eighteenth centuries. The new intellectuals of the early modern period were producing original work in Navya-Nyāya, Mīmāṃsā and grammar right up to the eve of colonial dominance. Settled The claim that Saṃskṛta died a natural death of internal exhaustion, and that the British merely certified the corpse, is not supported by the record of what was being written in it. Its collapse tracks colonial intervention, not intellectual senility. Contested in the literature — the correlation itself is not in doubt.

Cut the link language and you do not merely lose a language. You lose the network. Regional traditions fall out of contact. A civilization that had held one high-bandwidth channel between its intellectual centres for two and a half thousand years is left with fragments that can no longer address each other — and then, a generation later, into that silence, English arrives as the only available link language. And with English arrives the entire apparatus of Stage One, pre-installed, unexamined, and now indispensable.

This is why the attack on Saṃskṛta and the attack on the Brāhmaṇa are one operation.

To the man who carried it, Saṃskṛta was not a subject. It was a discipline of the whole life: rising before dawn for decades to hold a text exactly, in the correct metre, with the correct accent, transmitted by jaṭā, ghana and krama recitations engineered specifically so that error is detectable and corruption impossible. That is why the Ṛgveda survives intact from a depth of antiquity at which every other human civilization's record is rubble and inference. It survives because a specific community treated its exact preservation as the purpose of a human life, and accepted poverty to do it.

A civilization is not its monuments. Buildings are rebuilt, territory recovered, dynasties replaced inside a century. A civilization is a transmission — knowledge, grammar, metre, ritual, law and memory handed across generations. Cut the transmission for three generations and nothing is recovered, because no one is left who knows what was lost. Egypt did not survive it. Persia did not. Mesopotamia did not. They are archaeology.

Stage Two, continued: The Carrier

Everything said above concerns the language. This section concerns the man who held it, because the two cannot be separated and because his story has been so thoroughly inverted that most Indians now believe the opposite of every fact about it.

Begin with what the work actually was.

A boy of seven or eight is given to a teacher. He rises in the brāhma-muhūrta, the hour and a half before dawn, and he will rise at that hour for the rest of his life. He performs sandhyāvandana at dawn, at noon and at dusk, every day, for sixty or seventy years, without exception, without audience, and without any possibility of reward. He learns his śākhā by heart. For the Ṛgveda that is 10,552 mantras in the saṃhitā alone — and the saṃhitā is only the beginning, because he must then master the vikṛti pāṭhas, the permutation recitations, in which the text is recited forwards, backwards, in interleaved pairs, in braided triplets. Jaṭā. Ghana. Krama.

Consider what those permutations are for. They are not ornament and they are not display. They are an error-correcting code, designed some three thousand years before the term existed, on the reasoning that if a text must cross an unknown number of centuries through human memory alone, then the transmission medium must make corruption arithmetically detectable. Change one syllable and the permutations no longer resolve. The redundancy is the checksum.

A ghanapāṭhin — a man who holds a Veda in the densest of these recitations — represents somewhere between twelve and twenty years of continuous work under a teacher. There are, today, perhaps a few hundred such men alive on earth. Contested on the count, since no full register exists; Settled that the number is in the hundreds and falling.

Now the part that ought to end the argument about oppression forever.

He was forbidden to profit from it. The sale of learning — vidyā-vikraya — is condemned in the dharmaśāstra as a degradation. He could not charge for what he knew. He taught the sons of the village for nothing, fed at their houses in turn or on bhikṣā, and lived on dāna and on the produce of an agrahāra grant given precisely so that he would not have to earn. Accumulation was proscribed to him. The śāstra prescribes for the Brāhmaṇa a life of aparigraha — non-possession — and the historical record shows that the prescription was, on the whole, honoured.

Understand the institutional design here, because there is nothing else like it in the world. A civilization took its most valuable asset — the entire body of its knowledge — and placed it in the hands of a class that was structurally forbidden to convert it into wealth or power. The men who held everything were guaranteed nothing. That is not a priesthood extracting rent. That is a civilization deliberately building its memory into people who could not sell it.

And when the endowments were resumed, the temple economies plundered, and the revenue base confiscated by the settlements described in the next stage, this design became the community's destruction. A merchant caste that loses its patrons finds new ones. A landed caste that loses a patron still has land. A class forbidden to accumulate, whose entire economy was the grant, whose entire skill was a discipline the new order had declared worthless — that class simply falls, in a single generation, from the centre of the civilization to destitution.

The arithmetic of the accusation

Set the charge of oppression against the numbers, and it does not survive a paragraph.

Across two and a half millennia the overwhelming majority of Indian kingdoms were not Brāhmaṇa. Maurya. Gupta. Cōḷa. Pallava. Cālukya. Rāṣṭrakūṭa. Pratihāra. Paramāra. Sena. Ahom. Vijayanagara. Bhonsle. The Brāhmaṇa-origin ruling houses can be counted on two hands across the entire span — Śuṅga, Kāṇva, Kadamba under Mayūraśarman, and the Peśvās, who governed under the Chatrapati's seal and not in their own name.

A class that spent twenty-five centuries at the acknowledged apex of a supposedly rigid hierarchy, and in all that time almost never took a throne, is not an oppressor class. It is something the modern political vocabulary has no word for: an authority deliberately held apart from power, which accepted the separation as its own dharma. The śāstra is explicit — he does not rule, does not bear arms as his vocation, does not accumulate, is expected to be poor and is expected to tell the king the truth at his own cost.

Ask the question in its simplest form and the accusation collapses. A community that never held the throne, never held the treasury, never held an army, was forbidden to trade and forbidden to accumulate, and stood at a small single-digit fraction of the population, is said to have held a subcontinent of hundreds of millions in subjection for three thousand years.

Whom he worships

There is one refutation of the whole accusation that requires no archive, no survey and no colonial despatch. It requires only that you look at what a Brāhmaṇa does when he faces the image and folds his hands.

If a class fabricates a theology to establish its own supremacy, that theology will put that class at the summit — not merely in the role of officiant, but as the object of devotion itself. This is not a subtle prediction. It is what priestly self-interest looks like everywhere it has actually existed. The god will resemble the priest.

Now look at Bhārata.

Rāma is Ikṣvāku, of the Sūryavaṃśa: a Kṣatriya prince. Kṛṣṇa is of the Vṛṣṇi Yādavas, and the childhood the entire devotional tradition dwells upon is passed among cowherds in Vraja — not among scholars, and not in any assembly of learned men. Between them these two absorb the overwhelming mass of Hindu devotion across every region and every language of the subcontinent, and neither is a Brāhmaṇa.

Śiva stands outside varṇa altogether — the ascetic of the cremation ground, ash-smeared, in the company of what the ritual order treats as impure, holding no place in the social scheme at all. Devī, as Durgā and as Kālī, is likewise not situated within it, and Kālī's associations are pointedly with what lies beyond the boundaries of purity. Hanumān, among the most worshipped deities in all of India, is a vānara, not even human within the itihāsa. Gaṇeśa, invoked first at the opening of every rite performed anywhere by anyone, has no varṇa either — and it is the Brāhmaṇa who must invoke him before he may begin.

The daśāvatāra does include two Brāhmaṇa figures, and honesty requires naming them. But consider what they are. Vāmana is a dwarf mendicant who arrives asking for three paces of ground. Paraśurāma is the Brāhmaṇa who takes up the axe — and he stands in the tradition not as a model but as a warning, the figure who departs from his own dharma, whose excess is the point of his story. Neither is an image of Brāhmaṇa worldly supremacy. One is a beggar and the other is a caution.

The pattern runs deeper than the choice of deity. It runs through the texts themselves.

The Gītā is spoken by a Kṣatriya to a Kṣatriya on a battlefield; no Brāhmaṇa is present, and none is needed. In the Mahābhārata the great Brāhmaṇa warriors — Droṇa, Kṛpa, Aśvatthāmā — stand on the losing side, and Aśvatthāmā's end is among the most terrible in the whole itihāsa. The Bhāgavata returns again and again to the humbling of Brāhmaṇa pride: Durvāsā's rage against the Kṣatriya devotee Ambarīṣa recoils upon Durvāsā himself. And Sudāmā, the Brāhmaṇa companion of Kṛṣṇa, arrives at Dvārakā as a starving supplicant with a handful of beaten rice, and it is God who washes his feet — a story in which the Brāhmaṇa's poverty is the whole moral, and it is not presented as an injustice to be corrected.

These texts were composed, transmitted, memorised, commented upon and recited for millennia by the very community they repeatedly humble.

And the ritual order says the same thing where it counts most. At Purī, one of the four dhāmas, the Daitāpati sevakas — a community of Śabara descent — hold hereditary ritual rights that Brāhmaṇas do not possess, including the central offices of the Nabakalebara. At the most sacred point of one of the holiest kṣetras in Bhārata, the Brāhmaṇa stands aside. Settled

At Paṇḍharpūr, the entire Vārkarī devotional literature — the abhaṅgas that Maharashtra has sung for seven hundred years — was composed overwhelmingly by non-Brāhmaṇas: Nāmadeva the tailor, Cokhāmeḷā, Janābāī the maidservant, Tukārāma the Kuṇbī. Brāhmaṇas recite them. And in the Tamil country, as already recorded, the Āḻvārs' hymns are chanted in the sanctum beside the Veda.

If he controlled the narrative, he told it against himself with remarkable consistency for a very long time.

What they would not do

And there is one fact the accusation cannot accommodate at all. They did not convert.

Under the Sultanates. Under jizya. Under the Portuguese Inquisition, which from 1560 made the practice of Hindu rite a criminal offence in Goa and criminalised Saṃskṛta itself. Under two centuries of a missionary and educational apparatus purpose-built to make the whole inheritance contemptible. Across all of it, the conversion rate among these communities remained negligible.

Think about what that means about the conspiracy theory of Stage One. If the position were what Grant and Ward and Mill claimed — a racket maintained for extraction — then the rational operator switches sides the moment the state changes and the incentives invert. And the incentives inverted completely, for centuries. Conversion meant relief from the tax, access to office, patronage, land grants, physical safety, and the end of persecution. Every material argument pointed one way, and pointed that way for six hundred years.

They did not go.

They took the poverty and they took the danger instead, generation after generation. And when staying became impossible they did not convert then either — they walked. They carried palm-leaf bundles and household deities down the Konkan coast out of Portuguese Goa, over the ghats into Karnataka, where the temples they founded to shelter rescued mūrtis stand today. They walked out of Kaśmīra seven times across six hundred years. They went east across the Bay, and finally across oceans, where the same rites are performed now in Bali, Mauritius, Trinidad, Fiji, Durban, Houston and Melbourne by the descendants of people who had every reason to stop and did not.

You do not need to kill the transmitter to break the transmission. You need only make his work worthless: remove the endowments that feed him, close the institutions that train him, reclassify his discipline as dead, and teach his own society that he was a parasite. All four were done. All four are documented policy.

But for eight hundred years before that, the simpler method had also been tried.

Stage Three: Confiscate the Common Inheritance

Stage Two removes the civilization's high scholarship. Stage Three removes the floor beneath it — and this is the stage almost no Indian has been told about, because it is the exact inverse of what the textbook says.

Britain did not bring education to India. Britain destroyed a functioning mass education system and left an illiterate country behind.

This is recoverable from the colonial administration's own surveys, and it was recovered — by Dharampal, who went into the British and Indian archives and published the documents in The Beautiful Tree: Indigenous Indian Education in the Eighteenth Century (1983). The title is Gandhi's. Addressing Chatham House in London in October 1931, Gandhi said India was more illiterate than it had been fifty or a hundred years before, because British administrators had come in and uprooted a beautiful tree. He was ridiculed for it. Sir Philip Hartog challenged him publicly; the establishment view was that the old man was romanticising. Dharampal spent years in the archives and produced the evidence. Gandhi was right, and the proof had been sitting in the Company's own files. Settled

What the surveys found

Between 1822 and 1825, at Sir Thomas Munro's direction, the collectors of the Madras Presidency conducted a district-by-district survey of indigenous education. The returns recorded roughly twelve and a half thousand schools with some one hundred and eighty-eight thousand pupils, and over a thousand institutions of higher learning. Munro's Minute of June 1826 assessed that about one boy in three of school-going age was in school, and noted explicitly that the real figure was higher, because the survey did not capture instruction given at home, which was extensive.

Now the detail that detonates the whole account of Stage One — and it is important enough to take on its own, below.

William Adam's three reports on vernacular education (1835, 1836, 1838) surveyed Bengal and Bihar and reached an estimate on the order of one hundred thousand village schools. The precise number is genuinely disputed; Adam reached it partly by inference, and it has been argued over ever since. Contested What is not disputed is the character of what he found: a dense, decentralised, locally financed network of elementary schools embedded in ordinary villages.

And the most damning testimony comes from a colonial officer with no nationalist sympathies whatever. G. W. Leitner, Principal of Government College Lahore, published History of Indigenous Education in the Panjab since Annexation and in 1882. His finding, from his own investigation, was that indigenous education in the Panjab had been extensive before annexation and had been wrecked by British administration — the state system had displaced rather than supplemented it, and the province was less educated in 1882 than before the British arrived. Settled as his stated conclusion.

The charge, and the record

There is one accusation that does more work than all the others combined, because it supplies the moral justification for everything the reformers did to the system they were dismantling: that learning in Bhārata was a Brāhmaṇa monopoly, guarded by exclusion, and that the other varṇas were kept deliberately illiterate.

It is the load-bearing claim of the entire colonial account of Indian society. And it is refuted by the colonial administration's own paperwork.

Who was actually sitting in the schools. The Madras collectors did not merely count schools and pupils. They recorded the community composition of the student body, district by district, and the returns are reproduced in Dharampal's appendices where anyone may read them.

Across the Tamil districts, Brāhmaṇa children were a minority of school students. Śūdra communities and those recorded as other castes together formed the great majority — on the order of seventy per cent and above in several districts, with the Brāhmaṇa share falling under a tenth in some. Settled as to the pattern; the precise district figures are in the returns themselves and vary considerably between the Tamil, Telugu and Malayalam-speaking regions.

Who was standing at the front of them. Adam's Bengal and Bihar reports establish the second half of the point. The village schoolmasters were not, in the main, Brāhmaṇas either. Teaching in the vernacular schools was carried on substantially by men of Kāyastha and other communities — it was a profession, distributed across society, not a priestly reservation. Settled

So the position on the eve of Macaulay was this: a schooling system in which most of the students were not Brāhmaṇa and most of the teachers were not Brāhmaṇa, described by the government that ran the survey as a Brāhmaṇa monopoly requiring liberation.

What the observers said

The testimony runs across a thousand years and does not come from interested parties.

Xuanzang, the Chinese pilgrim, travelling in the seventh century, described an educational culture rather than an educational caste: children beginning formal study at about seven with a standard primer, proceeding through the five vidyās — grammar, the applied arts and crafts, medicine, logic and adhyātma — under teachers who did not sell instruction but were maintained so that they could give it.

G. L. Prendergast, a member of the Governor's Council at Bombay, recorded in 1821 that there was hardly a village, great or small, without at least one school in which reading, writing and arithmetic were taught. He was describing the presidency he helped govern, and he was not making a case for Indian civilization. He was stating an administrative fact. Settled

A. D. Campbell, Collector of Bellary, reporting into the Madras survey in 1823, went further and identified the cause of the decline he was witnessing. Schools were closing, he found, not because of any native prejudice but because the people had been impoverished — the endowments and revenue assignments that had maintained them were gone, and parents who had once sent their sons for nothing could no longer sustain a teacher. The mechanism described in the previous section was recorded by the officer watching it happen.

J. M. Ludlow, writing in 1858, put the whole matter in a sentence that ought to be in every Indian textbook and is in none: in every Hindu village which has retained its old form, he observed, the children are generally able to read, write and cipher — but wherever the village system has been broken up, there the village school has disappeared with it. Settled as to his stated observation.

G. W. Leitner in the Panjab in 1882, as already recorded, found the same and said so from inside the government that had done it.

What the tradition itself says

And the śāstric position is not what the accusation assumes, either. The Veda states its own audience directly.

That is the Veda declaring its own speech to be addressed to the śūdra by name. It is not an obscure passage and it is not a modern reading.

Nor is the record of who the tradition actually elevated ambiguous. Vyāsa, who arranged the Vedas, was the son of a fisherwoman. Vālmīki is the ādikavi. Satyakāma Jābāla could not name his father. Kavaṣa Ailūṣa was the son of a dāsī and his sūktas stand in the Ṛgveda itself.

And the tradition did not stop doing this. The Āḻvārs and the Nāyanmārs came overwhelmingly from non-Brāhmaṇa communities, and the Divya Prabandham of the Āḻvārs is recited in Śrīvaiṣṇava temples beside the Veda itself — the tradition canonising the compositions of non-Brāhmaṇa teachers into its own liturgy. Tirumaṅgai. Nammāḻvār. Nandanār among the Nāyanmārs. Later, Nāmadeva the tailor, Cokhāmeḷā, Tukārāma the Kuṇbī, Kabīr, Ravidāsa, Basavaṇṇa's whole movement. These are not marginal figures tolerated at the edge of the tradition. They are at its devotional centre, and it is Brāhmaṇa hands that have transmitted, commented upon and recited their work for centuries.

A system built to hoard learning does not spend two thousand years canonising the teachers who came from outside the supposed hoard.

The inversion

Now the turn, and it is the sharpest reversal in this entire essay.

The stratification the British described was, in substantial part, the stratification the British produced.

The indigenous school was free at the point of use because it was funded from land revenue and endowment. That is precisely why the cultivator's son could attend it. When the settlements confiscated that revenue, the free school died — and what replaced it was fee-paying instruction in a foreign language, in towns, accessible only to those with money and leisure.

Who could afford that? Not the village. A narrow urban propertied class could, drawn from whichever communities in each region already had commercial or administrative wealth. So education, which had been broad and local and free, became narrow and urban and purchased.

And the charge was then handed to Indian political movements as their founding grievance — most consequentially in the south, where an entire century of politics was built on the proposition that the Brāhmaṇa had kept the Tamil people from learning. The survey proving otherwise had been conducted in the same presidency, by the same government, a hundred years before, and left in a file.

The mechanism, which is never taught

These schools were not charity and not accident. They were funded, structurally, out of the produce of the land: assignments of revenue, rent-free grants, endowments attached to temples and maṭhas, community allocation. A village supported its teacher the way it supported its tank and its temple — as infrastructure. This is why schooling was effectively free at the point of use, and why the poor could attend.

The British revenue settlements — the Permanent Settlement of 1793 and the ryotwari systems that followed — reassigned exactly those revenue streams into the colonial treasury. The grants that fed the schoolmaster and the endowments that fed the pāṭhaśālā were resumed, absorbed, or rendered worthless.

Then the sequence closes on itself with an elegance that deserves study. Confiscate the revenue base of indigenous education. Wait a generation while the schools starve and shut. Measure literacy by a new metric and find it appallingly low. Publish the finding as proof of native backwardness. Offer English education as the remedy — priced beyond the reach of those who had previously been taught for nothing, and available in quantity only to the class being groomed as intermediaries. Then let the beneficiaries of that last step write the history of the first four.

By 1911 the census recorded literacy in the low single digits. By 1941, about sixteen per cent. And every Indian schoolchild since has been taught that this was the darkness out of which the British had begun to lift them, rather than the hole into which they had been pushed.

The Wood's Despatch of 1854 gave the policy its respectable theory: downward filtration. Educate a thin elite in English and knowledge would percolate to the masses. It never percolated. It was never designed to. What it produced was exactly what Macaulay had specified — a class of interpreters, cut off by language and formation from the people below them, dependent for their entire intellectual world on the rulers above them, and constitutionally unable to imagine that their own civilization had anything to teach.

Remember that class. It becomes Stage Six.

Stage Four: Manufacture the Fracture

A people who still think of themselves as one people can still act as one. So the fourth stage is to install a permanent internal division and persuade the population it is ancient.

It began with a word that exists in no Indian language.

Casta is Portuguese — breed, stock, purity of lineage. It arrived in the sixteenth century with men who applied it to something they had not understood. English took it as caste. There is no Saṃskṛta equivalent, because there is no such thing to name.

What existed was varṇa — a functional ordering of society by guṇa and karma, quality and vocation — and jāti, the local self-governing community, of which there were countless, which varied by region, changed occupation, split, merged, rose and fell across centuries, and which no central authority enumerated or ranked, because no central authority had the slightest interest in doing so.

Neither is caste. To render varṇa as caste is not translation. It is the smuggling of an entire colonial anthropology inside a single word. Settled

The tradition's own position on birth is not a modern apologetic. It is stated at the highest points of its own authority.

The Chāndogya Upaniṣad (4.4) records Satyakāma Jābāla, who cannot name his father and says so plainly, and whose truthfulness under humiliation is itself accepted by Gautama Hāridrumata as the qualification — the tradition adjudicating the question directly, in its oldest voice. The Vana Parva has Yudhiṣṭhira tell Nahuṣa that where truth, generosity, forbearance, good conduct and non-cruelty are found, there is the Brāhmaṇa, and that birth without them confers nothing. The Vajrasūcī Upaniṣad dismantles the birth criterion methodically, testing brāhmaṇatva against body, birth-group, knowledge, action and observance, and rejecting each in turn.

And the rsi-lineage is itself the argument. Vyāsa, son of Satyavatī. Vālmīki. Viśvāmitra, born Kṣatriya. Mahidāsa Aitareya, son of Itarā. Kavaṣa Ailūṣa, son of a dāsī, whose sūktas stand in the Ṛgveda. A system engineered to lock status by birth does not build its own scripture out of counter-examples to birth.

The outside witnesses agree. Megasthenes at the Mauryan court around 302 BCE counted seven groups, not four — and recorded that the philosopher class alone could be entered from any of the others, precisely the mobility the colonial account declares impossible. He recorded that Indians kept no slaves, and that no Indian army in his knowledge had ever marched out to invade another nation. Faxian, walking across Madhyadeśa in the fifth century, found a people prosperous, unregistered, moving freely, with capital punishment rare. Settled

So where did the thing we now call caste come from? From the ledger.

An empire must make its subjects legible in order to administer them. From the enumerations of the 1860s and the all-India census of 1871-72 onward, the colonial state set out to count Indians by caste. To count, you must fix. A fluid, local, negotiable identity had to become a single permanent inheritable attribute, recorded against a name and published with the authority of the state.

The machinery of the count

It is worth being exact about how the thing was actually built, because the mechanism is where the intent shows.

The box. From 1871–72 the census carried a caste column, and every Indian had to be entered in it. One person, one answer, one permanent word. That requirement alone did more damage than any doctrine, because it made a whole range of ordinary Indian realities administratively impossible. A man whose community was known by one name in his village and another in the district had to choose. A family whose occupation had changed over three generations had to declare one. A group that stood in different relation to its neighbours in two adjacent taluks had to be fixed as one thing everywhere. And a man who simply did not think of himself in these terms was assigned an entry by the enumerator, on his behalf, permanently.

Nothing in Indian society had ever done this. No king had done it. No śāstra had done it. The state now did it to three hundred million people, on paper, with the force of law behind the paper.

The multiplication. And once the counting began, the count exploded — not because India grew more complex, but because enumeration manufactures categories. Prinsep had produced one hundred and seven "castes" in the single city of Benares in 1843. By the census of 1901 the enumeration ran to something on the order of two thousand three hundred main castes and tribes across the subcontinent, with recorded sub-divisions running into the tens of thousands. Contested as to exact aggregates, since the definitions shifted from census to census and no two are strictly comparable — which is itself the point. A stable, ancient, immemorial structure does not change its own size every ten years depending on how the question is phrased.

The ranking. Then Risley, Census Commissioner in 1901 and architect of the Ethnographic Survey, took the further step that converted a survey into a hierarchy. He proposed to rank India's communities by official social precedence — and to establish the ranking on physical measurement. He measured skulls. He measured noses. He built a nasal index and concluded that a community's social standing varied with the proportions of its members' noses. This was published as the science of the Government of India in The People of India (1908). Settled

Indians were measured with calipers, like stock, and the results were printed as their place in their own society.

The feedback loop. And here the operation became self-driving, which is the mark of a well-designed one.

Communities understood immediately what was happening: a permanent official rank was being assigned, and it would govern access to office, to recruitment, to standing. So they organised to influence it. Caste associations formed across the subcontinent in the 1890s and 1900s for the express purpose of petitioning the census — to be recorded under a better name, claimed a higher origin, placed in a superior column. Journals were founded. Delegations were sent. Genealogies were commissioned.

The statute. And in the very same year that the first all-India census went out, the machinery acquired police power.

The Criminal Tribes Act of 1871 empowered the government to declare entire communities criminal — not individuals, communities, by descent. Those notified under it were registered, restricted in their movement, required to report, subjected to surveillance, and in many cases confined to settlements. Membership was by birth. A child born into a notified community was born a criminal in the eyes of the state, and would die one. The Act was extended in scope in 1876 and again in 1911, and by its later years applied to well over a hundred communities and millions of people.

It was repealed only in 1952 — five years after independence. The communities affected are known to this day as the denotified tribes, and they still carry the stigma the statute manufactured. Settled

The demographic weapon. The census was also used directly against the majority as a population. In 1911, under E. A. Gait, the administration circulated criteria for deciding which groups should be counted as Hindu at all — a set of tests which, applied, would have struck very large numbers off the Hindu total and redistributed them into separate columns. The proposal provoked immediate and widespread agitation, and it is instructive that the objection had to be organised at all: the size of the Hindu population had become something the enumerator could adjust by definition. Settled as to the circular and the controversy it caused.

The inheritance. The last full caste enumeration was Hutton's in 1931. But the categories did not lapse with it. They passed into the separate electorates of the 1930s, into the constitutional schedules of 1950, into the reservation architecture, and from there into the permanent structure of Indian electoral politics — where they now determine ministries, alliances and majorities.

Which means the classification built by Plowden, Ibbetson and Risley, ranked by nasal index, and lobbied over by associations that formed to lobby it, is today the single most durable institution the British left behind in India. It has outlived the railways, the Indian Civil Service, the Raj itself and every treaty of the period.

Nicholas Dirks, in Castes of Mind (2001), puts the conclusion in a line: the census had unprecedented effects on the very social realities it claimed merely to be recording. Settled Bernard Cohn made the same argument about enumeration and objectification. This is the mainstream scholarly account, not a Hindu apologetic.

The deepest fault line

Every fracture described so far runs between Indians and their own past. The last and most damaging one was cut through the living body of Hindu society itself, and it was cut deliberately, in stages, by men who wrote down what they were doing.

First the theory. The Ārya-Drāviḍa schema was not merely an account of ancient migration. It was a political instrument with a specific payload: if the higher varṇas were an invading fair race and the rest of the population the conquered dark aboriginals, then a very large part of Hindu society was not Hindu at all. It was a subject people, held in bondage by foreigners, awaiting liberation by a third party — and the third party, conveniently, was already administering the country.

Every element of that construction was manufactured in Europe. There is no invasion in the Indian record, no memory of one, no tradition of one, and no text describing one. Jambudvīpa is Sanātana Bhārata, and the Brāhmaṇa is integral to it and not an arrival within it. But as an instrument the theory was without equal, because it converted an internal social question into a racial and national one — and racial questions cannot be resolved, only fought.

Then the naming. On that premise, movements were constituted in the 1910s and 1920s under names that carried the theory in the first syllable — Ādi-Drāviḍa in Madras, Ādi-Āndhra from 1917, Ādi-Hindu in the United Provinces, Ādi-Dharm in the Panjab from 1926. Ādi: the original, the aboriginal, the people who were here before the others came. The names encode a European hypothesis about Indian prehistory as though it were the settled memory of the communities adopting it. Settled as to the naming and its premise.

Then the enumeration. The census machinery described above did the rest. Fluid, locally variable, regionally negotiated standings were fixed as permanent hereditary entries, ranked by an official in 1901 according to physical measurement, and published as the government's own statement of who stood where. What had been local and disputable became national and authoritative.

And note the direction of travel. Practices of exclusion in pre-colonial India were real in places, varied enormously between regions, differed between one village and the next, and were negotiated, contested and changed over time — communities rose and fell in standing across generations, which is why the historical record is full of jātis whose position is different in one century from the last. What the census produced was something that had never existed: a single, uniform, subcontinental, permanent ranking, identical in Tinnevelly and in Rohilkhand, backed by the state.

Then the field. The missionary enterprise, having concluded in the Stage One literature that Hindu society was a priestly tyranny over a captive population, drew the operational conclusion and directed its effort at the communities it judged most detachable. The mass-movement conversions of the late nineteenth and early twentieth centuries proceeded on exactly that reasoning, and the reasoning is stated plainly in the missionary literature of the period, which discusses the depressed classes as a field of opportunity in precisely those terms. Settled as to the strategy being explicitly articulated in the sources.

Then the constitution. And on 16 August 1932 the operation reached its logical end.

The Communal Award, announced by the British Prime Minister Ramsay MacDonald, granted separate electorates to the Depressed Classes. Understand what that instrument was. It did not mean reserved seats within a common body. It meant a separate roll, separate candidates and separate votes — a distinct political nation, constitutionally severed from the rest of Hindu society, permanently.

The precedent was not obscure. Separate electorates had been granted to Muslims by the Morley-Minto reforms of 1909, and everyone alive in 1932 could see where that instrument had led and was leading. To apply the same device to Hindu society was to begin the identical process again, from within.

Gandhi began a fast at Yerwada on 20 September 1932 against the Award. Four days later the Poona Pact was signed, replacing separate electorates with reserved seats inside a joint electorate — a considerably larger number of reserved seats than the Award had offered, but within one body, on one roll. Ambedkar signed it. Settled as to the sequence and the terms.

And then the inheritance. The Award was defeated; the logic was not. It passed into the schedules of 1950 and from there into the permanent architecture of Indian politics, where the categories Risley ranked now determine constituencies, ministries and majorities — and where the manufactured racial premise is still taught in schools as settled history.

So the fault line is real today. It is real in the way a surgical incision is real. Something was done to a living body, deliberately, with instruments, by people who published their methods; and the wound was then kept open, first because an empire found it useful and afterwards because an electoral system found it profitable.

None of which is an argument that nothing was ever wrong, or that no one ever suffered. It is an argument about who benefited from the wrong being described as total, ancient, racial and unalterable — and about who has kept describing it that way for a century and a half.

The tradition's own answer to the question was never the one the census assigned it. It is in the Veda's own declaration of whom it addresses. It is in Satyakāma Jābāla, whose father could not be named. It is in Vyāsa and Vālmīki. It is in Nandanār and Tiruppāṇ Āḻvār, whose compositions are recited in the temples to this day; in Cokhāmeḷā, in Ravidāsa, in Nāmadeva, in Basavaṇṇa. These are not figures the tradition tolerated at its margin. They stand at its devotional centre, and it is the descendants of Brāhmaṇas who have recited and transmitted and commented on their work for six hundred years.

A civilization does not do that with a people it considers separate.

Max Müller was central to the diffusion of the philological frame, and he was entirely open about the object of his Veda scholarship. Writing to his wife in December 1866, he described the Veda as the root of the Indian religion, and said that showing Indians what that root was would be the only way of uprooting everything that had grown from it over three thousand years. Settled The letter exists and is not disputed. He wrote in similar terms to the Duke of Argyll, then Secretary of State for India.

Caste is not varṇa. Caste is not jāti. Caste is a colonial administrative artefact — built by census, ranked by calipers, and handed back to us as our heritage. Settled on the mechanism of construction; Polemical only in the force of the naming.

Stage Five: Make Them Ashamed

The four preceding stages are structural. The fifth is atmospheric, and it does the work that policy cannot: it converts a defeated people into an embarrassed one.

Argument is for the administrator. For everyone else there is ridicule.

The nineteenth century produced an industry of it. The exhibition, the postcard, the ethnographic photograph, the travelogue, the missionary lecture with lantern slides. The fakir. The fire-walker. The man with the basket and the flute. The civilization of Pāṇini and Mādhava and Suśruta became, in the visual vocabulary of the world, a snake charmer.

Nothing about that image was accidental. It was manufactured, distributed and repeated until it became the world's default picture of Indians — and, more importantly, until Indians themselves began to wince at it.

That wince is the whole objective. When the target adopts the caricature and feels the shame on the aggressor's behalf, the operation no longer requires the aggressor's participation. It runs itself.

Everything was subjected to it. Dress — the dhoti as absurd, the turban as comic, the bare torso as savage. Food. Bathing. Ritual. The number of the gods. The forms of the images. The languages, described as barbarous jabber by men who had not learned them. The women. The marriages. The temples, catalogued as obscene by administrators who had come from a country that burned women as witches inside living memory.

The genre had a long tail and an international reach. A century after Ward, Katherine Mayo's Mother India (1927) — an American production timed with precision to answer Indian demands for self-rule with a catalogue of Indian sexual and social degradation — sold enormously across the West and did exactly the work required of it. Gandhi's assessment, that it read like a drain inspector's report, was accurate and did not slow it down by a single edition.

And note where Stage Five lands in the sequence. It arrives after the schools have been taken and the classical language cut off. The Indian who might have answered the ridicule with evidence no longer has access to the evidence. He cannot read the Aṣṭādhyāyī. He has not heard of the Yuktibhāṣā. He does not know that Munro counted the schools. He has been disarmed first, and mocked second, and he experiences the mockery as simply true.

The Ledger

Stages One through Five construct an accusation. This section records what the accusation licensed — and what had already been done, for centuries, by those who needed no accusation at all.

It is set down chronologically and without ornament. Every entry rests on a source, and in several of the earliest cases the source is the perpetrator's own chronicle, written in pride.

Sindhu, 712. The Arab conquest under Muhammad bin Qasim. The Chach Nāma, the chronicle of the conquest itself, records the reduction of Debal and Brahmanabad, the slaughter of the defenders, and the enslavement of the survivors. The accounts of what was done to the Brāhmaṇas of Brahmanabad come from the conquerors' own record, not from any later Hindu complaint. Settled as to the chronicle's contents.

Nālandā, Vikramaśilā, Odantapurī, c. 1193–1200. Bakhtiyar Khalji's destruction of the great vihāras of Magadha. The Tabaqāt-i-Nāsirī of Minhāj-i-Sirāj — again the conqueror's chronicle, not the victim's — records that the shaven-headed men were put to the sword, and that when the library was found and men were sought who could explain the books, none remained alive who could read them.

Kaśmīra, 1389–1413. The reign of Sikandar, called Butshikan, the idol-breaker. The Rājataraṅgiṇī continuations of Jonarāja record the systematic destruction of temples, the prohibition of rite, and the flight or death of the Paṇḍits. The Kashmiri tradition counts seven exoduses across six centuries. This was the most complete of the early ones.

Benares and Mathurā, 1669. Aurangzeb's farmān ordering the destruction of the temples and schools of the infidels, recorded in the Maāsir-i-Ālamgīrī of Sāqi Mustaʿid Khan — the court's own official history, written in approval. The Kāśī Viśvanātha and the Keśavadeva temples were destroyed under it. The Brāhmaṇas of Kaśmīra, under a concurrent campaign of forced conversion, made the appeal in 1675 that ended with Guru Tegh Bahadur's execution at Delhi — a sequence in which a Sikh Guru gave his life for the right of Brāhmaṇas to keep their thread. Settled

Goa, 1560–1812. The Inquisition. Temples across Ilhas, Bardez and Salcete had already been demolished by state order in the preceding decades. The tribunal criminalised Hindu rite, prohibited Saṃskṛta, seized children for conversion, and conducted the auto-da-fé. Its records were burnt on its abolition — which is itself evidence, and which is why A. K. Priolkar's reconstruction in The Goa Inquisition (1961) had to be built from survivors' accounts and outside testimony. The Sāraswat Brāhmaṇas carried their deities out by boat and on foot down the coast to Karwar, Gokarna and Mangalore. They did not convert. They walked out with the mūrti and began again in the forest. Settled

Malabar, 1921. The Moplah violence in Ernad and Valluvanad: mass killing, forced conversion, and the flight of tens of thousands of Hindus, with Brāhmaṇa families prominent among those targeted. Documented at the time by observers who held no Hindu-nationalist commitment whatever. Settled

Telangana, 1946–48. The Razākārs of the Ittehad, under Qasim Razvi, operating as an irregular militia across the Nizam's dominions. Village raids, systematic terror against the Hindu population and its priestly and landholding families, and the massacres at Parakāla in September 1947 and Bhairanpally in 1948. Settled as to the pattern of atrocity; Contested as to totals, since the archives were not kept for our benefit.

Maharashtra, January and February 1948. And here the ledger crosses into independent India.

After Gandhi's assassination on 30 January 1948, organised mobs attacked Chitpāvan Brāhmaṇa homes and businesses across Pune, Satara, Sangli, Kolhapur, Belgaum and Nashik. Houses burnt. Families driven from villages their ancestors had lived in for centuries. Property destroyed on a scale recorded in the thousands of premises. The offence was sharing a jāti with the assassin.

This was a pogrom on Indian soil, in a Congress-administered province, within weeks of independence, against a community whose collective guilt consisted of a name. Death tolls remain Contested and were never seriously investigated — which is itself the point. It does not appear in the textbooks. Most educated Indians have never heard of it.

Tamil Nadu, from 1916. Not a massacre but an extended political programme: the Non-Brahmin Manifesto of 1916, the Justice Party, the Communal Government Orders, and the Self-Respect Movement from 1925, which together constructed the Brāhmaṇa as the alien enemy of the Tamil people. The argument rested entirely on the Ārya-Drāviḍa racial fiction imported from European philology — which is to say, Stage Four supplied the weapon. That the Justice Party was congenially loyalist, and that the colonial administration found the division exceptionally useful against the nationalist movement, is not a coincidence to be explained away. Contested as to intent; Settled as to the pattern of encouragement. The demographic result is visible in the state's out-migration.

Marichjhāpi, January and May 1979. Bengali Hindu refugees from East Bengal, largely from the Namaśūdra community, who had settled on an island in the Sundarbans, were blockaded and then forcibly evicted by the Left Front government of West Bengal. Food and water were cut off. Boats were sunk. The killings and deaths from starvation and disease during the blockade and eviction have never been officially enumerated, and no prosecution followed. Contested on the toll, which ranges across an order of magnitude in the sources; Settled on the blockade, the eviction and the absence of any accounting.

Kaśmīra, 1989–90. The final exodus. Tika Lal Taploo shot in September 1989. Justice Neelkanth Ganjoo in November. On the night of 19 January 1990, loudspeakers from the mosques of the valley instructed the Paṇḍits to leave. B. K. Ganjoo, found hidden and killed. Sarwanand Koul Premi and his son, abducted and killed in April. Girija Tickoo, abducted, raped and killed in June. Between one and three and a half lakh people left their homeland within months. Contested on the figure; Settled on the fact.

They have not returned. This occurred not under any invader but under the Republic of India, in living memory, under a central arrangement that watched it happen, and it produced no national reckoning of any kind.

What is actually being lost

And the cost is not primarily a cost to Brāhmaṇas. It is a cost to Bhārata, and it is being incurred right now, this year, while this is read.

Entire Vedic śākhās have gone extinct within living memory or stand at one or two surviving reciters. The Śākala remains strong; others do not. Whole disciplines — śrauta ritual practice, the deeper reaches of Mīmāṃsā, Navya-Nyāya, the astronomical and mathematical traditions, regional schools of grammar and poetics — are held by an ageing handful of men, most of them without students. Manuscript collections rot in family houses because no one in the next generation can read the script they are written in.

This knowledge is not stored in buildings. It is not stored in libraries. It is stored in trained human beings, and when the last man who holds a recitation dies untaught, that recitation is gone from the universe. Not damaged. Not endangered. Gone, permanently, in a way that no discovery and no funding and no growth rate will ever reverse.

Stage Six: Build the Vessel

By the 1880s the operation faced its own success. Two generations of Macaulay's policy had produced exactly the class it was designed to produce: English-educated, professionally trained, articulate, ambitious — and shut out of the senior positions of the government they had been taught to admire. That is a dangerous compound. Educated men denied advancement do not stay quiet.

The response was not repression. Repression was expensive and, after 1857, had been shown to be dangerous. The response was to build a vessel: a respectable, constitutional, national-seeming organisation into which that energy could be poured, where it could be observed, moderated, and drained through petition.

The Indian National Congress was founded in 1885 through the agency of A. O. Hume, a retired officer of the Indian Civil Service — designed to provide a managed channel through which Indian political feeling could be expressed, monitored and contained rather than allowed to gather into something the state could not control. Contested in the historiography, which has argued the safety-valve question for a century, but resting on Hume's own correspondence and on the sympathetic account of his career left by William Wedderburn.

The founding facts are not contested, whatever one makes of them.

Its early sessions were addressed to the Crown, and its early resolutions professed loyalty to the Empire while requesting reform within it. Its presidency was held, in its early decades, by a series of Englishmen — George Yule in 1888, William Wedderburn in 1889 and again in 1910, Alfred Webb in 1894, Henry Cotton in 1904. An organisation founded by a retired imperial civil servant, presided over by Englishmen, petitioning the Crown for concessions, is not on its face an instrument of national liberation. Settled as to every fact in this paragraph.

And it worked as designed for two decades, until Indians who wanted something other than petition forced the question — and were expelled for it. The split at Surat in 1907 between the Moderates and the party of Tilak was precisely the collision between a vessel built for containment and men who refused to be contained.

But the deeper point is not about Hume, and it is not a matter of individual recruitment. Hume died in 1912. Gandhi did not join Indian politics until 1915. Nobody was recruited by anybody.

The point is structural, and it is the whole argument of this essay:

Consider the men who would negotiate the transfer of power. Gandhi: barrister, Inner Temple. Nehru: Harrow, Trinity College Cambridge, Inner Temple. Jinnah: Lincoln's Inn. The three men who between them determined the shape and the partition of the subcontinent were all barristers of the English Inns of Court, formed in English institutions, arguing in English, inside a constitutional idiom supplied entire by the power they were negotiating with.

That is not an accusation against any of them personally. It is the specification from the Minute of 1835, fulfilled on schedule. Macaulay asked for a class Indian in blood and colour but English in taste, opinion, morals and intellect, to stand between the rulers and the ruled. He got it. And in 1947 that class did not disappear.

It inherited the state.

Stage Seven: Wait for the Deliverer

Here is the stage that is least discussed and matters most, because it explains the shape of everything that followed.

A people who remember their own capability organise. They raise armies, fund campaigns, build institutions, take the field. That is what Bhārata did for as long as she remembered — and it is worth saying plainly who actually had to be defeated for India to be taken. Not the Mughal court, which was by then a pensioner. It was the Marathas. The Anglo-Maratha wars from 1775 to 1818 were the real conquest of India: the British had to break Hindavī Svarājya — the reconquest of Sanātana sovereignty carried by Śivājī, Sambhājī and the Peśvās — before the subcontinent was theirs. Settled as to the military history. That is precisely why Maratha history above all others has been rewritten in the vocabulary of Mughal chronicle and colonial administration, because the accurate account of it is the account of Indians who fought, and who very nearly won.

But a people who have been through the six preceding stages do not organise in that way. They cannot. They have been told for a century that their ancestors built nothing. Their scholarship has been declared dead, their schools have been closed, their social fabric has been enumerated into permanent factions, and their self-image has been reduced to a man with a snake. They are hungry — genuinely, physically hungry, in a country that fed the world. They are labourers on land they once owned and weavers without looms and indentured hands on other men's islands.

A population in that condition does not raise an army. It waits for deliverance.

The Empire had no need to arrange this. It follows from the six preceding stages the way water follows a channel. And it determines everything about the independence movement that actually emerged: its idiom was moral rather than strategic, its instrument was suffering rather than force, its theatre was the conscience of the ruler rather than the capability of the ruled, and its central figure was received by the masses not as a politician but as a mahātmā — a designation the masses conferred, spontaneously, because that is what a people in that condition require.

None of this is a charge against the man. It is an observation about the condition of the people who needed him, and about what that condition permitted. A movement founded on appeal to the ruler's conscience is a movement that has already conceded that the ruler has the power and the Indian has only the appeal. That concession was manufactured. It was manufactured deliberately, over four generations, by the policies described in Stages One through Five.

And the criticism came from within, not from outside. Motilal Nehru, C. R. Das, Lajpat Rai and Subhas Bose were all enraged when Non-Cooperation was withdrawn unilaterally in February 1922, at the height of mass mobilisation, over the events at Chauri Chaura. When Bose was elected Congress President at Tripuri in 1939, defeating Gandhi's own candidate, he was forced out within months. When the Congress chose a leader for the transfer of power in 1946, twelve of fifteen Provincial Committees nominated Patel and none nominated Nehru — and Nehru took the office. Settled as to the record; the interpretation is Contested and has been argued ever since.

And then there is the question of what actually ended the Empire.

The Second World War left Britain bankrupt, and that is the largest factor. But the immediate collapse of British confidence in holding India came from a direction the vessel had not channelled. The Indian National Army trials at the Red Fort from November 1945 produced mass unrest across the country, cutting across community lines. In February 1946 the ratings of the Royal Indian Navy mutinied — the rising spread to some seventy-eight ships and shore establishments and roughly twenty thousand men, from Bombay to Karachi to Madras to Calcutta, with sympathy strikes in the cities. Within weeks the Cabinet Mission was announced. Settled as to the chronology.

An empire that cannot trust its sepoys has already lost, and the men in London knew it. That is the fact the vessel was built to prevent, and it happened anyway.

Stage Eight: Let It Run Without You

The final stage requires nothing whatever of the occupier. He goes home. The operation continues, because it was never resident in him. It was resident in the curriculum and in the statute book, and both were inherited intact.

They left in 1947. They did not really leave. They had spent a century manufacturing a class that would carry the work forward without instruction, and in 1947 that class took possession of the state, the schools, the courts and the press. The Empire did not need to stay. It had already been internalised.

The statute book

Look first at what was written into law, because law is where a mentality becomes permanent.

The asymmetry of religious rights. Article 30 of the Constitution grants minority communities the right to establish and administer their own educational institutions. No equivalent right is extended to the majority. The practical consequence is that a Hindu institution may be regulated in ways a minority institution may not — an asymmetry that has been litigated for seventy years and has repeatedly been resolved against Hindu institutions. Settled as to the text and its effect.

State control of temples. From the Madras Hindu Religious and Charitable Endowments Act of 1951 onward, and in similar legislation across several states, Hindu temples and their revenues were placed under direct government administration. This power was inherited directly from the colonial state, which had appropriated temple endowments for its own treasury. No other community's places of worship are so controlled anywhere in the Republic. Seventy-nine years after independence, the revenues of Hindu temples remain in state hands, disbursed at state discretion, in a country that describes itself as secular. Settled

The Emergency, and the words inserted during it. On 25 June 1975 the government of Indira Gandhi suspended the fundamental rights of Indian citizens. Over the following twenty-one months, tens of thousands were detained without trial under preventive-detention law, the press was placed under prior censorship, the courts were curbed, and a mass sterilisation campaign was driven through the countryside with quotas and coercion, running to some eight million operations in 1976 and 1977 alone.

And in November 1976, in the middle of that suspension, with the opposition in jail and the press censored, the Forty-second Amendment inserted the words socialist and secular into the Preamble of the Constitution. Settled as to every fact in this paragraph.

The pattern in office

The individual cases are various. The pattern is not.

Kaśmīra, 1990. The exodus recorded in the Ledger above happened while the machinery of the Indian state watched. No commission of inquiry followed. No prosecutions of consequence followed. No rehabilitation of the displaced to their homeland has been achieved in thirty-six years. A republic that could not or would not protect a hundred thousand of its own citizens driven from their homes by terror, and then declined to investigate it, was not exhibiting neutrality. Polemical in the characterisation; Settled in the record of inaction.

West Bengal, 1977–2011. Thirty-four years of uninterrupted Left Front rule over a province that had been the industrial and intellectual capital of India. Its share of national industrial output collapsed; capital left; the university system that had produced Bose and Ray and Tagore's inheritors declined into a patronage machine. Marichjhāpi in 1979, Nandigram in 2007. Settled as to the economic trajectory; Contested as to how much of it is attributable to policy rather than to national trends — though the comparison with Maharashtra and Gujarat over the identical period is not ambiguous.

Tamil Nadu, continuing. The political lineage begun by Periyar has not been abandoned; it has been inherited. In September 2023 a serving Minister of the State Government and son of the Chief Minister, Udhayanidhi Stalin, publicly declared that Sanātana Dharma should be eradicated, likening it to dengue, malaria and coronavirus — diseases which, he said, must not merely be opposed but eliminated. He declined to retract, and repeated the position when challenged. Settled as to the statement and its date.

The habit of deference. And there is a subtler symptom, which is the most telling of all because it is not a policy but a posture. A republic founded on the abolition of subjecthood has spent much of its independent existence organising its largest political party around a single family, in which office has descended by blood and loyalty is professed to persons rather than to principles. Court behaviour did not end in 1947. It changed address.

That is not an accident of Indian politics. It is the seventh stage of this essay reappearing in republican dress — the deliverer, sought again, because the condition that produced the need for a deliverer was never treated.

The inherited shame

And here is the condition in which the descendants now live, which is the operation's true monument.

Consider a specific and very common Indian man. He is educated, often abroad. He is capable, articulate and successful. He can discuss the Greek and the Norse traditions with interest and no discomfort whatever. He finds the Japanese tea ceremony exquisite, Zen austerity profound, Gregorian chant sublime, a Catholic cathedral moving even though he believes none of it. He will travel a long way to see a European ruin and stand before it with reverence.

Put him in front of his grandmother's household rite and he is embarrassed. The chanting strikes him as superstition. The images strike him as gaudy. The purity observances strike him as primitive. He will explain, if pressed, that he is a rational person.

He is not. A rational person is evenly sceptical. What he exhibits is not scepticism but selective embarrassment, and selectivity is the diagnostic fact. The reaction is not reasoned; it is trained. He was taught, through the whole length of his schooling and his reading and the films he grew up on, which traditions are to be found beautiful and which are to be found embarrassing — and the training worked so completely that he now experiences its output as his own refined taste.

Ask him what he is ashamed of and the loop closes. He cannot say. He has not read the Aṣṭādhyāyī. He has not heard of the Yuktibhāṣā. He does not know that his ancestors distilled zinc at Zawar for five hundred years before anyone else on earth, or that Europe learned to rebuild a human nose by watching an Indian do it, or that a British councillor in 1821 found a school in nearly every village he governed. He is ashamed of a civilization he has never been taught.

That is not an accident of his education. It is the design of it. Stage Three removed the evidence; Stage Five supplied the verdict. He was disarmed first and mocked second, and so the mockery met no resistance and became simply how things are.

The tell is the vocabulary. Watch which words are available to him and which are not. He can call a practice Brāhmaṇical without evidence and be understood as sophisticated. He cannot describe his own tradition approvingly in public without being suspected of something. He may say he is spiritual but not religious; he may not say he performs sandhyāvandana. He will describe the Purāṇas with a word borrowed from the study of dead Mediterranean religions, and will not notice that he has categorised his own living tradition as extinct while using it.

Every one of those constraints was installed. None of them is natural to a free people describing itself.

And observe what secular has come to mean here, since it is the word he reaches for most readily. In its ordinary sense elsewhere it denotes state neutrality: the government takes no side and administers no faith. In India it has come to denote something quite different — a state that directly administers the temples and revenues of one tradition while touching no other, that grants educational autonomy to every community except the largest, and that recognises no equivalent asymmetry as a problem requiring correction. The word entered the Preamble by the route described above, and has been used ever since to place the majority's inheritance permanently on probation, and it is defended most fiercely by those it disinherits.

The last symptom is the imitation. A society that has been taught to despise its own forms does not thereby become original. It copies. So one finds the architecture, the vocabulary, the political categories, the academic fashions and the moral postures of other countries adopted at second hand and a decade late, applied to a society they were never devised for, and worn with the particular anxiety of a man who fears above all things being thought provincial.

There is a word for this condition and Indians have always had it. Gulāmī. Not the political condition, which ended in 1947, but the older and more durable one, which did not.

The recovery is therefore not political and cannot be legislated. It is a matter of a man reading, in his own language, the things his great-grandfather knew — and discovering, with a shock that is very common among those who have done it, that he was never told any of this, and that the not-telling was not an oversight.

The frame in the mind

Beneath the statutes and the parties sits the thing that actually sustains all of it.

Mill's tripartite periodization still organises the school textbooks: Hindu, Muslim, British. The Ārya invasion is still taught across much of the country as settled fact, decades after its evidentiary basis collapsed. Brāhmaṇical has become a free-floating adjective for anything oppressive — usable without evidence, requiring no demonstration, immune from ordinary standards of proof, and functioning in educated Indian conversation as a marker of sophistication.

The census categories the British invented are now the load-bearing structure of Indian electoral politics — which means the single most consequential act of colonial social engineering has been made permanent by the free votes of free Indians.

And the cinema completed the work Ward and Grant began: the lecherous priest, the scheming astrologer, the cruel village pandit denying a boy his education. Not one film — a genre convention so settled that it operates as visual shorthand. No other Indian community has been assigned a standing cinematic villain-type, and none would be permitted one. Polemical The pattern is real, and seventy years of Hindi cinema will confirm it to anyone who looks.

This is the stage at which facts stop working. You can put Munro's returns, Adam's reports, Leitner's confession, the Tabaqāt's own admission that the readers had been killed, Risley's nasal index, Max Müller's letter, Megasthenes' seven classes and Bairoch's manufacturing tables in front of a man in this condition, and he will not update — because the frame that processes evidence has itself been replaced.

That is completion. A civilization arguing its enemy's case for free, indefinitely, with sincerity, and mistaking the argument for conscience.

Seventy-nine years is a long time to remain a subject of a state that no longer exists.

What Reversal Requires

Not resentment. Resentment is itself a symptom of the condition — it accepts the frame and merely inverts its emotional sign.

Reversal requires the specific work the operation was designed to prevent.

Recover the documents. Dharampal did it once, and it took years in the archives. The Munro returns, the Adam reports, the Leitner findings, the pre-colonial revenue records of school endowments, the district-level survey data — these exist, and every one of them refutes the story we were taught. They belong in the school curriculum, not in specialist monographs.

Restore the link language. Not as sentiment, and not as a compulsory school subject to be resented, but as what it was: a working medium for a civilization's own thought, in which its own texts can be read without a translator's frame standing between reader and page. Every Indian who reads the Gītā in English is reading a book edited by someone else's assumptions.

Rebuild the transmission while there is time. Every year the channel narrows, and every year the loss is permanent. The Ledger above records what was taken by sword and by statute; this is what is being lost right now, to nothing but neglect. Fund the pāṭhaśālās. Pay the reciters. Digitise the manuscripts while there are still men who can read them.

And name the procedure. Not to nurse a grievance, but because you cannot audit a frame you do not know you are standing inside. Every category through which the educated Indian currently understands his own society — caste, Aryan, Brahminical, the Hindu period, the beautiful tree that supposedly never existed — was manufactured, by identifiable men, for a stated purpose, and is documented in their own words.

That is not a conspiracy theory. It is a bibliography.

A civilization that produced Pāṇini's grammar, Mādhava's series, Suśruta's surgery and the exact preservation of the Ṛgveda across three thousand years was told, by men who could not read a line of its literature, that it was barbarous.

And it believed them.

It is time to stop.

Notes and Sources. Colonial primary material: James Mill, The History of British India (1817); Charles Grant, Observations (1792); William Ward (1811); T. B. Macaulay, Minute on Indian Education, 2 February 1835, with Bentinck's Resolution of 7 March 1835; Charles Wood, Despatch on Education, 1854; Thomas Munro, Minute on Indigenous Education, Madras Presidency, June 1826, with the collectors' returns of 1822-25; William Adam, Reports on the State of Education in Bengal and Bihar, 1835, 1836, 1838; G. W. Leitner, History of Indigenous Education in the Panjab since Annexation and in 1882 (Lahore, 1883); H. H. Risley, The People of India (1908) and the Census of India 1901; J. Z. Holwell, An Account of the Manner of Inoculating for the Small Pox in the East Indies (1767); The Gentleman's Magazine, October 1794, on the Indian rhinoplasty, with J. C. Carpue (1816); the Max Müller correspondence of 1866 and 1868; Katherine Mayo, Mother India (1927); William Wedderburn, Allan Octavian Hume (1913).

Indian primary material. Dadabhai Naoroji, Poverty and Un-British Rule in India (1901); R. C. Dutt, The Economic History of India (1902, 1904); M. K. Gandhi, address at Chatham House, London, October 1931; the published proceedings of the Indian National Congress sessions, 1885-1907.

Classical sources. Bhagavad Gītā 4.13 and 18.41-44; Chāndogya Upaniṣad 4.4; Mahabharata, Vana Parva, the Yudhiṣṭhira-Nahuṣa samvada; Vajrasūcī Upaniṣad; Pāṇini, Aṣṭādhyāyī; Suśruta Saṃhitā; Jyeṣṭhadeva, Yuktibhāṣā (c. 1530) and Nīlakaṇṭha Somayājī, Tantrasaṅgraha (1501). Foreign witnesses: Megasthenes, Indikā, as preserved in Strabo, Arrian and Diodorus; Faxian, travel record, early fifth century.

Secondary literature. Dharampal, The Beautiful Tree: Indigenous Indian Education in the Eighteenth Century (1983); Nicholas Dirks, Castes of Mind (2001); Bernard Cohn on enumeration and objectification in colonial India; Mike Davis, Late Victorian Holocausts (2001); Angus Maddison, The World Economy: A Millennial Perspective (OECD, 2001) and Contours of the World Economy 1-2030 AD (2007); Paul Bairoch, International Industrialization Levels from 1750 to 1980, Journal of European Economic History (1982); Utsa Patnaik on the colonial drain; William Digby, "Prosperous" British India (1901); A. K. Priolkar, The Goa Inquisition (1961).