Knowledge is Power: A Critical Genealogy of Epistemic Power in Colonial and South Asian History

 

 

I want to begin today not with an argument, not with a theory, and not with a date. I want to begin with a phrase. Three words. Five syllables.

 

"Knowledge is power."

 

You have heard it before. You have probably said it yourself. Perhaps a teacher wrote it above a blackboard. Perhaps a parent offered it as consolation when you were struggling at school. Perhaps you have seen it printed on a motivational poster, stitched onto a tote bag, hashtagged across social media. It is one of those phrases so thoroughly absorbed into the furniture of modern life that we have stopped noticing it.

 

And that, precisely, is the problem.

 

Because a phrase that everyone agrees with and no one examines is not a truth. It is an ideology. It is a story about the world so thoroughly naturalised that it has stopped looking like a story at all. And when stories stop looking like stories — when they pass themselves off as simple, obvious facts — they become the most powerful and the most dangerous kinds of ideas.

 

So today, we are going to do something that feels almost impolite. We are going to take this beloved phrase — this rallying cry of educators and revolutionaries and self-help gurus — and we are going to hold it up to the light. We are going to ask: where did it come from? What does it really mean? Who wrote it, and why, and in whose service did it operate? And we are going to ask, with some urgency, who has paid the price for its uncritical acceptance.

 

To do that, we are going to travel. Intellectually and historically. We will move from the court of King James I of England in the early seventeenth century to the ghats of the Ganges. We will move from the philosophy seminars of Paris in the 1970s to the villages of Bengal. We will visit a map that conquered a continent, a census that froze a society, a school policy that tried to remake millions of minds, and a burning book that said: enough.

 

By the end of this article, I hope you will never hear the phrase "knowledge is power" quite the same way again. Not because it is false — but because it is true in ways far more complex, far more troubling, and far more interesting than the motivational poster would have you believe.

Let us begin.

 

The Phrase and Its Problem

I want to start with a simple observation. Five words: "knowledge is power" — and yet, if you think about them carefully, they immediately raise far more questions than they answer.

Power to do what? Knowledge of what? And — most critically — whose knowledge, and whose power?

 

These are not pedantic questions. They are the questions that separate a slogan from an analysis. And it is analysis that we are here to do.

 

The phrase does not mean one thing. It has been used — simultaneously, in living memory — by Thomas Jefferson, who invoked it to justify republican civic education, and by British colonial administrators, who invoked it to justify the forced re-education of an entire subcontinent. It has appeared on the walls of village schools in post-independence India and in the operational manuals of the intelligence agencies that subsequently surveilled those same villages. It has animated liberation movements and justified oppression. It has been the slogan of the dispossessed seeking dignity and the instrument of the powerful seeking control.

 

That extraordinary promiscuity — the phrase's capacity to serve utterly opposed agendas — is not an accident. It is the key to understanding it. A phrase powerful enough to serve everyone has, woven into its fabric, a dangerous ambiguity. And that ambiguity, as we shall see, has not been innocent. It has had consequences that we are still living with today.

 

Now, before we get into the history — and we have a great deal of history to get through — let me make a methodological point. This article is not going to ask whether the statement "knowledge is power" is true or false. That is the wrong question, and it produces the wrong kind of thinking. Instead, we are going to ask a different set of questions: How did this particular formulation arise? In what historical conditions did it become dominant? What work has it done in the world? What has it enabled, and what has it suppressed?

 

This kind of inquiry is what the French philosopher Michel Foucault — one of the central figures we will encounter today — called a genealogy. Not a history of what really happened, but a history of how certain things came to seem natural, inevitable, and true. Genealogy, in this sense, is a kind of philosophical detective work. We are looking for the fingerprints of power on the surfaces of ideas that present themselves as neutral.

 

And nowhere on earth — let me say this now, at the outset, as a claim I will spend the rest of this article defending — nowhere on earth were those fingerprints more clearly pressed than in the Indian subcontinent under British colonial rule.

 

In British India, the identification of knowledge with power was not merely a philosophical conviction. It was a governing technology, a bureaucratic programme, and a cultural weapon. It was deployed with extraordinary sophistication and at extraordinary scale. And the story of how it worked — and how it was resisted — is one of the most illuminating and, I would argue, most morally urgent stories in modern history.

 

So let us begin at the beginning. Or rather, let us begin where the phrase itself begins.

 

A Genealogy of the Phrase

Before Bacon — Knowledge as Wisdom, Not Weapon

The first thing to say is that the association of knowledge and power is ancient. It is not a European invention. But the specific form it takes matters enormously.

 

Sun Tzu, the Chinese military strategist writing in the fifth century BCE, argued in The Art of War that the commander who knows both himself and his enemy need not fear the outcome of a hundred battles. Here, knowledge confers strategic advantage — but it is knowledge in the service of military mastery, bounded by the context of war. The ancient Greek tradition, from Socrates onward, associated knowledge — episteme — with a kind of authority over mere opinion. To know, in the Platonic sense, was to have access to a higher realm of reality inaccessible to those mired in sensory experience.

 

In the Islamic scholarly tradition of the eighth through thirteenth centuries — the period historians call the Islamic Golden Age, when Baghdad's House of Wisdom was the intellectual centre of the known world — knowledge held a sacred status. The Arabic word ilm, meaning knowledge or science, was understood as a form of closeness to God. The Prophet is recorded in the Hadith as having instructed the faithful to "seek knowledge, even unto China" — a remarkable injunction, given that China was the farthest conceivable destination from the Arabian Peninsula. The point was that the pursuit of knowledge recognises no geographical limit, and that its pursuit is a spiritual obligation.

 

And here in the Indian subcontinent itself, the ancient Tamil poetess Avvaiyar — one of the great literary voices of the Sangam period, roughly the third century BCE to the third century CE — composed what remains one of the most beautiful statements on the nature of knowledge ever written:

Katrathu kai mann alavvu; kallaathathu ulagalavvu. What is learnt is a handful of earth; what is unlearnt is the size of the world.

That is not a statement about power. It is a statement about humility — an acknowledgment of the vastness of what remains unknown. It valorises the pursuit of knowledge not because it gives you control, but because it opens you to the immensity of what you do not yet understand[i].

So what changed? How did the idea of knowledge shift from this orientation — toward wisdom, virtue, spiritual growth, humility — toward the explicitly instrumental claim that knowledge is power over the world?

The answer, in short, is: Francis Bacon. And Francis Bacon was very much a man of his particular moment.

 

Bacon and the Instrumental Revolution

Francis Bacon was born in 1561 and died in 1626. He was not merely a philosopher; he was a politician of considerable ambition, serving as Lord Chancellor of England under King James I before his career collapsed in a corruption scandal. He was also a visionary of a new kind of knowledge — a knowledge that did not contemplate the world but changed it.

 

The phrase we are examining today appears in his Meditationes Sacrae of 1597 in the Latin original: "Nam et ipsa scientia potestas est" — "For knowledge itself is power." He returns to the idea throughout his major works: The Advancement of Learning in 1605, the Novum Organum — his great methodological treatise — in 1620, and the unfinished utopian fiction New Atlantis in 1627. In each of these, the relationship between knowledge and power is not merely asserted but elaborated into a comprehensive political and philosophical programme.

 

Let me give you a sense of the flavour of Bacon's thinking. In the Novum Organum — the title means "the new instrument," a direct challenge to the Aristotelian logical tradition he sought to replace — he wrote two sentences that I want you to sit with:

"Nature to be commanded must be obeyed." And: "The sovereignty of man lieth hid in knowledge."

 

"The sovereignty of man lieth hid in knowledge." What a sentence! Not the goodness of man, not the wisdom of man, not the virtue of man — the sovereignty. Dominion. Control. This is a programme of mastery. And the object of that mastery is explicitly nature — the non-human world.

 

Now, I want to make sure we appreciate why this was a radical departure, not just a restatement of what everyone already believed. In the dominant Aristotelian tradition that Bacon was rebelling against, the purpose of knowledge was theoria — contemplation. You studied the natural world in order to understand the eternal forms that organised it. The goal was insight, wisdom, understanding — not manipulation or control. The good scholar was the one who best grasped the natural order; the idea that the scholar's job was to improve upon it, to redirect it toward human purposes, would have seemed impious.

 

Bacon's genius — and it was genius, whatever we may think of its consequences — was to insist that knowledge disconnected from practical effect was essentially useless. In a striking phrase, he compared scholastic philosophy to a spider spinning elaborate webs from its own substance — beautiful, intricate, and catching nothing of substance. He wanted a different kind of knowledge: knowledge that caught things. Knowledge that worked.

 

But here is the question we need to press on: worked for whom? And on what?

 

The feminist philosopher of science Carolyn Merchant, in her remarkable 1980 book The Death of Nature, argued that the Baconian project was saturated with metaphors of domination and control that were simultaneously about nature and about women. She quotes Bacon describing his method as putting nature "on the rack" to extract her secrets — and this is not a casual metaphor. The rack was a specific instrument of torture. The imagery of interrogating, binding, and forcing nature to yield her secrets runs throughout Bacon's writing. And nature, in this tradition, is consistently feminised. She is passive matter awaiting the imposition of form by active masculine reason.

 

This is not mere language. As Merchant argues, these metaphors structured actual scientific practice and its social organisation — the assumption that the natural world (and, by extension, the bodies of women, the lands of colonised peoples, the labour of workers) existed as resources to be known, catalogued, and exploited in service of the knowing subject.

 

Bacon also understood knowledge as fundamentally political and institutional. His New Atlantis — an unfinished utopian fiction that is one of the most influential texts in the history of science — imagines a utopian island called Bensalem, governed not by kings or priests but by a college of scientists called Salomon's House. These scholar-administrators collect information from across the world, conduct experiments in purpose-built laboratories, and direct their findings toward the improvement of the state. It is, in every essential respect, the blueprint for the modern research university and the modern scientific state — and it is, from the outset, explicitly a programme of power.

 

There is one more crucial dimension of the Baconian legacy that I want to highlight before we move on, because it becomes absolutely central to the colonial story. The historian and philosopher Mary Poovey, in her 1998 book A History of the Modern Fact, traces how the Baconian programme generated a new kind of epistemological object — the fact. Not an observation, not an interpretation, not a judgment, but a fact: something that appeared to be neutral, quantifiable, universally legible, stripped of the particular interests of whoever had produced it. The fact looked like nature speaking directly, without any human voice to mediate it.

 

This epistemological fiction — and it is a fiction, as Poovey demonstrates at length — proved to be one of the most powerful inventions in the history of governance. Because if the results of your investigations appear as facts, as neutral data, then when you use them to govern, you are not exercising power. You are simply responding to reality. You are not imposing your will; you are merely administering what the situation objectively requires.

Remember that principle. We will need it very soon.

 

Foucault: Power Makes Knowledge Possible

So far, we have Bacon's version of the relationship between knowledge and power: knowledge enables power. If you understand how nature works, you can harness it. This is the version most people have in mind when they quote the phrase, and it is not wrong — but it is radically incomplete.

 

The philosopher who gave us the most important tools for going deeper is Michel Foucault — a French thinker who died in 1984 and whose influence on the humanities and social sciences in the decades since his death has been, if anything, greater than it was during his lifetime. Foucault's crucial move is to flip the directionality that Bacon assumed.

 

Bacon says: knowledge enables power. Foucault says: power produces knowledge. Not the same thing at all.

 

In Discipline and Punish, published in 1975, Foucault wrote what has become one of the most quoted passages in modern academic writing:

"Power produces knowledge... power and knowledge directly imply one another; there is no power relation without the correlative constitution of a field of knowledge, nor any knowledge that does not presuppose and constitute at the same time power relations."

 

Let me unpack that, because it is dense. Foucault is saying that what counts as knowledge — what questions get asked, what answers are treated as valid, what categories are used to organise the world — is not determined by some neutral process of investigation. It is determined, at least in part, by power relations. The psychiatrist's authority to classify someone as mad is itself an exercise of power. The criminal justice system's production of criminal statistics is itself a form of governance. There is no position outside power from which one could simply observe the world as it is.

 

Think about what this means for colonialism. When British scholars and administrators went to South Asia and produced vast quantities of knowledge about Indian society — its languages, its religions, its legal customs, its caste system, its history, its flora and fauna — they presented this as a scientific enterprise: disinterested, rigorous, factual. Foucault would say: look again. What questions were asked, and what questions were not? Who was treated as a reliable source, and who was dismissed? What categories were used to organise the data? Who benefited from the knowledge produced?

 

The answers to those questions reveal not a neutral scientific enterprise but a systematic production of knowledge in the service of colonial governance. And this is the argument that the Palestinian-American literary critic Edward Said made, brilliantly and influentially, in his 1978 landmark work Orientalism.

 

Said argued that European scholarship about "the Orient" — that vast and frankly incoherent category that encompassed, at different moments, the Arab world, Persia, Turkey, South Asia, East Asia, and anywhere else that struck European writers as sufficiently exotic — was not disinterested. It was a discourse: a systematic body of representation that, over centuries of accumulation, produced "the Oriental" as a recognisable type — sensual, irrational, unchanging, incapable of self-governance — in relation to whom the European subject appeared, by contrast, as rational, dynamic, and naturally suited to rule.

 

Said's own words are precise:

"Orientalism is not an airy European fantasy about the Orient, but a created body of theory and practice in which, for many generations, there has been a considerable material investment."

 

Material investment: not merely ideas but institutions, careers, funding, archives, careers, policies. The Asiatic Society of Bengal. The survey reports of the Indian Civil Service. The ethnographic museums of London and Edinburgh. The comparative religion departments of European universities. All of these were simultaneously knowledge-producing and power-exercising institutions.

 

Now, Said's work has also been criticised — and we should be honest about this. Critics from Aijaz Ahmad to Homi Bhabha have argued that Said overestimated the coherence of Orientalism, underestimated the contradictions and resistances within it, and sometimes reduced colonised peoples to passive objects of European representation, thereby reinscribing the very dynamic he was critiquing. These are fair points. The colonised were not simply the objects of colonial knowledge; they were also its shapers, its interpreters, its subverters. We will return to them.

 

But Said's central insight — that the production of knowledge about colonised societies was deeply and structurally implicated in the exercise of colonial power — is not merely defensible. It is one of the most important intellectual contributions of the twentieth century. And nowhere does it find more compelling historical illustration than in British India.

 

The British in India — Knowledge as Governance

A Trading Company Becomes an Empire

Let me now take you to the Indian subcontinent. And let me begin with a story that tells you a great deal about how knowledge and power were intertwined from the very beginning of British involvement in the subcontinent.

 

The East India Company was founded in 1600. This needs a moment's reflection, because the Company is one of the strangest entities in the history of the world. It was a joint-stock trading corporation — a business, with shareholders, a board of directors, and the primary goal of making a profit — that somehow, over roughly a century and a half, became the effective ruler of most of the Indian subcontinent. It commanded armies. It levied taxes. It administered justice. It conducted foreign policy. At its peak, it was governing the lives of perhaps 200 million people, and it did so not as a government but as a company.

 

Now, in the early days of the Company's engagement with India — the seventeenth century, when it was genuinely just a trading enterprise — its need for knowledge was commercial. Company employees had to learn local languages to negotiate contracts. They had to understand local legal customs to enforce agreements. They had to grasp the financial systems, the weights and measures, the seasonal rhythms of Indian commerce. This knowledge was practical, immediate, and acquired in direct relationship with Indian interlocutors who were perfectly capable of withholding it or manipulating it to their own advantage.

 

But everything changed dramatically on a June afternoon in 1757, at a place called Plassey — a mango grove in Bengal — where the young Company commander Robert Clive defeated the Nawab of Bengal, Siraj ud-Daulah, in a battle that was as much a piece of political theatre as a military engagement. Clive had bribed several of the Nawab's key commanders beforehand; the battle was decided before it began. But its consequences were enormous. By 1765, the Company had acquired the Diwani — the legal right to collect the tax revenues — of Bengal, Bihar, and Orissa. Overnight, it went from being a trading company operating in India to being the de facto sovereign of the richest province of the Mughal Empire.

 

And that transformation created a knowledge problem of staggering proportions. How do you govern a subcontinent of 200 million people — speaking dozens of languages, organised by an extraordinarily complex web of caste, clan, religious, and regional identities, observing a bewildering variety of legal traditions from Hindu dharmashastra to Islamic fiqh to local customary law, and organised through agrarian systems of considerable sophistication — when your total British personnel in India numbers a few thousand at most?

 

The answer, which the Company developed over the next several decades with remarkable thoroughness, was: you know it. You gather information. You classify. You categorise. You map. You count. You translate. You archive. And in knowing it — in making it legible — you make it governable.

 

The historian Bernard Cohn, in his posthumously collected essays Colonialism and Its Forms of Knowledge, identified what he called the British "investigative modalities" — the systematic ways of knowing through which India was made into an object of colonial governance. Let me walk you through them, because each one is a case study in the transformation of knowledge into power.

 

The Survey: Mapping as Act of Sovereignty

Have you ever thought about what a map is? Not what it shows — but what it does?

We tend to think of maps as neutral representations of pre-existing facts. The mountain is there; the map shows it. The river runs this way; the map records it. But this is profoundly misleading. A map is always a set of choices: what to include, what to exclude, what to name, where to centre, what to make visible. And those choices are never innocent.

 

The Great Trigonometrical Survey of India, initiated in 1802 under Colonel William Lambton and later expanded under George Everest — yes, the man after whom the mountain is named — was one of the most ambitious scientific enterprises in history. Working across decades, in conditions of extraordinary difficulty — monsoon rains, malarial jungles, burning plains — teams of British surveyors and their Indian assistants measured the entire subcontinent with extraordinary mathematical precision. The result was a set of detailed topographical maps covering virtually every inch of British Indian territory.

 

Now, these maps were presented as achievements of science. And in a certain sense, they were: the mathematical rigour of the survey was genuinely impressive. But the geographer J.B. Harley, in his foundational 1988 essay "Maps, Knowledge, and Power," makes a devastating argument: maps are never innocent documents. They are always political. They impose a vision of the world, centring some spaces and marginalising others, naming places in ways that assert particular claims of sovereignty.

 

And the Survey of India did this systematically. It produced maps in which India appeared as a coherent, bounded, unified space — a space defined entirely by British administrative and strategic imperatives, not by the political geographies, pilgrimage routes, sacred landscapes, or ecological understandings through which Indian people had, for millennia, organised their relationship to the land.

 

Think about what that means. Before the Survey, the Gangetic plain — the great river valley that is the heartland of Indian civilisation — was known to those who lived there through a dense network of sacred sites, seasonal flooding patterns, agricultural calendars, and local ecological knowledge. The rivers that defined it were understood in terms of their sacred significance, their seasonal behaviour, their role in local agriculture. This was sophisticated, detailed, deeply local knowledge — the kind of knowledge that farmers and priests and boatmen had accumulated over centuries.

 

The Survey replaced all of this with coordinates, contour lines, and revenue divisions. It was not that the local knowledge was wrong — much of it was extraordinarily precise in its own terms. But it was knowledge that served the wrong masters. It served the people who lived there. The Survey produced knowledge that served the people who governed from London.

 

The naming process was particularly revealing. The Survey systematically Anglicised or replaced indigenous place names — a process that might seem like a minor inconvenience but was actually, as the postcolonial theorist Homi Bhabha would put it, a form of epistemological violence. Indigenous place names were not arbitrary labels. They carried embedded within them local histories, mythological associations, ecological knowledge, and cultural memory. The name of a river in the Deccan plateau encoded stories about monsoon floods, about saints who had bathed there, about the families who had farmed its banks for generations. The Survey replaced all of this with a colonial nomenclature that was accurate to the map and empty of meaning to the people who lived in the territory it represented.

 

The maps were also, quite deliberately, instruments of military control. Matthew Edney, in his definitive 1997 study Mapping an Empire, demonstrates that the Survey's directors understood their work explicitly as an instrument of imperial domination. The detailed topographical surveys were classified; civilian access was restricted. They existed to make India permanently visible — and therefore permanently vulnerable — to British military and administrative power.

 

The Census: Counting as Constitution

If the Survey was the mechanism by which British India's physical space was made governable, the Census was the mechanism by which its social space was made governable. And the Census, in some ways, is the more disturbing instrument.

 

Let me give you some context first, because many of you may not be deeply familiar with the social geography of India that the Census encountered. Indian society in the eighteenth and nineteenth centuries was organised by a complex of overlapping social categories of which caste was central but by no means the only one. The term caste itself is a European importation — it comes from the Portuguese word casta, meaning lineage or breed, which Portuguese traders in the fifteenth century applied to what they encountered in India. But casta was a catch-all term for two quite different things.

 

First, there was varna — the ancient four-fold scheme of ritual hierarchy: the Brahmins, who were the priestly and scholarly class; the Kshatriyas, the warrior nobility; the Vaishyas, the merchant and agricultural class; and the Shudras, the service and artisan castes. And outside even this four-fold order were those whom the system classified as beyond the varna entirely — the people we now call Dalits, those designated as "untouchable," assigned the most polluting and socially marginalised roles, subject to forms of discrimination of extraordinary cruelty.

 

But below this overarching framework of four varnas, the actual operative social unit was the jati — an endogamous occupational group defined by birth, of which there were perhaps several thousand across the subcontinent, each with its own specific customs, dietary rules, marriage practices, and occupational specialisations. The relationship between varnas and jatis was complex, contested, and varied enormously across regions. A jati that was considered high-status in one region might occupy a very different position in another. And caste was not static: communities migrated, converted, adopted new occupations, claimed new ritual identities. The boundaries were permeable and constantly negotiated.

So what happened when the colonial Census encountered this reality?

 

The Census commissioners' response was, in a word: simplification. They attempted to impose standardisation — to reduce the thousands of local jatis to manageable all-India categories, ranked in a clear hierarchical order. Every person in India had to be assigned to a single caste identity, which was then recorded in the Census register and became, in effect, an administrative fact about that person.

 

The sociologist Nicholas Dirks, in his meticulous 2001 study Castes of Mind: Colonialism and the Making of Modern India, makes the crucial argument: this process transformed caste from a fluid, contextually negotiated social practice into a rigid, state-registered identity. The Census did not merely describe caste as it found it. It constituted caste in a new form — fixed, ranked, permanent, state-sanctioned.

 

And the consequences were profound. By recording each individual's caste in official returns, the colonial government created an administrative reality that then became the basis for further administrative acts. Revenue was assessed by caste. Military regiments were organised by caste. Most infamously, the Criminal Tribes Act of 1871 — to which I will return in a moment — designated entire communities as hereditary criminals on the basis of their supposed caste identity.

 

But perhaps the most consequential effect of the Census's approach to social enumeration was what it did to the Hindu-Muslim relationship — an effect that would ultimately contribute to one of the greatest catastrophes in modern South Asian history: the Partition of 1947.

 

Here too, the Census imposed a clarity that reorganised how people understood themselves. The Census required every person to declare a religious identity — Hindu, Muslim, Sikh, Christian, or one of a small number of other categories. There was no category for "both", no category for "neither", no category for the many forms of syncretic practice that had, for centuries, blurred the boundaries between these supposedly distinct traditions.

 

Think about what this erased. In pre-colonial South Asia, the boundaries between Hindu and Muslim practice were far more porous than the communal politics of the late colonial period would suggest. Sufi shrines were visited by both Hindus and Muslims; the devotional traditions of the Bhakti movement — with its emphasis on personal religious experience over ritual orthodoxy — had produced cultural forms that crossed sectarian lines. Many Hindu castes incorporated elements of Islamic practice and vice versa. The sharp enumeration of the Census replaced this lived complexity with binary categories — and then made those categories the basis of political representation.

 

The Morley-Minto reforms of 1909 established separate electorates for Muslims — the principle that Indian Muslims constituted a separate political community whose representation in legislative councils should be calculated on the basis of their proportion in the Census. The effect was to make the relative size of religious communities a matter of direct political consequence. And once you make counting people by religion politically consequential, you create a powerful incentive for political organisations to invest in identity mobilisation — to tell people that who they are, religiously, is the most important political fact about them.

 

The Muslim League, founded in 1906, and the Hindu nationalist currents within the Congress were both, in different ways, products of the political logic that colonial enumeration had created. The road to Partition — to the division of the subcontinent in August 1947, which produced one of the largest and most violent mass migrations in recorded history, killing somewhere between 200,000 and 2 million people, depending on which historian you consult — that road runs, at least in part, through the Census office.

 

This is what it means to say that knowledge is power. Not that knowing things is generally useful — everyone agrees with that. But that the specific categories through which a powerful state chooses to know its subjects shape, constrain, and sometimes destroy the lives of those subjects. The Census was not a neutral scientific instrument. It was a political one. And the politics it served were colonial politics.

The School as Battlefield

The Great Education Debate

We come now to what is, in my view, the most nakedly revealing episode in the history of colonial knowledge-power in India. It is an episode that took place not on a battlefield but in a committee room, and its principal weapon was not a gun but a memorandum. But its consequences were at least as lasting as any military conquest.

 

I am talking about the controversy over the language and content of colonial education that reached its climax in the 1830s. To understand it, we need to meet two intellectual camps that were, in the 1820s and 30s, locked in an acrimonious argument about what kind of education the British should fund in India.

 

The first camp were the Orientalists. Their intellectual programme had been developed in the late eighteenth century, most powerfully by the extraordinary figure of William Jones — a polymath linguist who had arrived in Calcutta as a judge in 1783 and immediately immersed himself in the study of Sanskrit. Jones founded the Asiatic Society of Bengal in 1784 — the first major institution for the systematic scholarly study of Indian civilisation. And in 1786, he made one of the most consequential linguistic discoveries in history. In a presidential address to the Society, he observed that Sanskrit bore systematic resemblances to Greek and Latin — resemblances too regular to be accidental — and proposed that all three languages must have descended from a common ancestor. This discovery inaugurated modern comparative linguistics and, eventually, the identification of what we now call the Indo-European language family.

 

Now, Jones and his Orientalist colleagues were not without their blind spots and their imperialism. They gathered and translated Indian texts in ways that served the Company's legal and administrative purposes. Their framing of Indian knowledge was shaped by European categories that frequently distorted what they described. But they operated from a conviction — genuine, if sometimes paternalistic — that Indian intellectual traditions were sophisticated, profound, and deserving of serious scholarly attention. They argued that British-funded education in India should work through and with these traditions — through the classical languages of Sanskrit and Persian — rather than against them.

 

The second camp were the Anglicists. And their position was, to put it mildly, less generous. They argued that Indian classical learning was not merely inferior to European learning — it was effectively worthless. The proper vehicle of education for Indians should be the English language, European science, and European history. India should not be educated through its own traditions; it should be educated out of them.

 

The dispute came to a head in 1835, when Thomas Babington Macaulay — then serving as the Law Member of the Governor-General's Council under Lord William Bentinck — composed the document that has come to be known, with good reason, as one of the most consequential and most contested texts in the history of British India: his Minute on Indian Education.

 

Macaulay was not a modest man. He was one of the most brilliant writers of Victorian England — his History of England remains a landmark of the genre — and he knew it. He brought to the question of Indian education the same combination of magisterial confidence and spectacular ignorance that characterises so much of British imperial thought at its most dangerous. He had never visited India before taking up his appointment. He knew no Sanskrit and no Arabic. And he composed his Minute on the language and content of Indian education with the following extraordinary opening gambit:

 

"I have no knowledge of either Sanskrit or Arabic. But I have done what I could to form a correct estimate of their value... I am quite ready to take the Oriental learning at the valuation of the Orientalists themselves; and I have never found one among them who could deny that a single shelf of a good European library was worth the whole native literature of India and Arabia."

 

I pause here, because I want to make sure we appreciate the full audacity of that passage. Macaulay is saying: I have not read the texts I am dismissing. I do not know the languages in which they are written. But I have heard from people who do know them that they are inferior — and since those people know them far better than I do, and even they concede the inferiority, the matter is settled. This is, as a logical argument, extraordinary. It is the argument of a man using second-hand testimony to confirm a conclusion he had reached before he asked the question.

 

But the Minute goes further than mere dismissal. It articulates, with uncomfortable clarity, exactly what colonial education was for. Macaulay did not want to educate Indians for their own benefit, their own flourishing, or the cultivation of their own capacities. He wanted to produce — and I am going to quote him at length here, because the precision of his language is exactly the point:

 

"We must at present do our best to form a class who may be interpreters between us and the millions whom we govern; a class of persons, Indian in blood and colour, but English in taste, in opinions, in morals, and in intellect."

 

"Indian in blood and colour, but English in taste, in opinions, in morals, and in intellect." That is not an education policy. That is a programme of cultural replacement. It is an attempt to use the educational apparatus to remake human subjectivity — to produce people who inhabit Indian bodies but think, feel, and judge in English terms. To make, in the terminology of postcolonial theory, colonial subjects: people whose capacity for independent thought and self-governance has been replaced by the habits and dispositions of the coloniser.

 

Macaulay's Minute carried the day. Lord Bentinck approved its recommendations. English became the medium of instruction in Company-funded schools and colleges. And the education system that followed — reinforced by Wood's Dispatch in 1854, which established the universities of Calcutta, Bombay, and Madras in 1857, modelled on the University of London — systematically marginalised or destroyed the indigenous systems of learning that had served Indian society for centuries.

 

What Was Destroyed

This is a point worth dwelling on, because the standard colonial narrative treats this substitution as simply the replacement of backward knowledge by advanced knowledge — the natural march of progress. But the evidence does not support that story.

 

Dharampal, an Indian historian whose 1983 compilation The Beautiful Tree drew on colonial survey reports to document pre-colonial Indian education, found something deeply inconvenient for the colonial narrative. In certain regions of India — parts of what are now Tamil Nadu, Andhra Pradesh, and Punjab — the village-based indigenous education system had a higher rate of participation, and in some cases a higher degree of accessibility to lower-caste children, than the colonial system that replaced it. The beautiful tree of the title comes from a speech by Mahatma Gandhi, who spoke of colonialism having uprooted a flourishing educational system and replaced it with something alien and withered.

 

The indigenous systems that were marginalised or destroyed included not just Sanskrit scholarship — the intellectual tradition that Western Orientalists were most interested in — but an entire ecosystem of practical, local, and vernacular knowledge. Agricultural knowledge embedded in local farming traditions. Medical knowledge in the hands of village practitioners. Mathematical knowledge in local accounting systems. Ecological knowledge carried by forest-dwelling communities who knew, with extraordinary precision, the seasonal, botanical, and hydrological dynamics of their particular environments.

 

None of this knowledge could be expressed in the categories or the languages of the colonial educational framework. And so, systematically and thoroughly, it was pushed to the margins — deprecated, underfunded, officially ignored, and, in many cases, eventually lost.

 

Gandhi's Dissent

The most powerful contemporary critique of this educational programme came from a man who had himself been produced by it — a man who had trained as a barrister in London, who spoke and wrote English with extraordinary facility, and who turned those capacities, with devastating effect, against the system that had cultivated them.

 

Mahatma Gandhi, writing in Hind Swaraj — Indian Home Rule — in 1909, made an argument that was, in the colonial context, almost shockingly radical:

 

"To give millions a knowledge of English is to enslave them... Is it not a painful thing that, if I want to go to a court of justice, I must employ the English language as a medium; that, when I become a barrister, I may not speak my mother-tongue, and that some one else should have to translate to me from my own language?"

 

Gandhi's critique was not anti-intellectual. He was not arguing against education. He was arguing against a specific form of education whose purpose was alienation — the production of people who were strangers to themselves, who could only access the institutions of power through the medium of a foreign language and a foreign intellectual tradition.

 

He developed this position into a concrete programme called Nai Talim — Basic Education — which sought to integrate craft knowledge, agricultural practice, and vernacular learning into a curriculum that could be delivered in village schools rather than urban universities. Knowledge, for Gandhi, should be rooted in material life and local context — not imported wholesale from an alien civilisation.

 

Gandhi's Nai Talim was never fully implemented. But the argument it embodied — that the language of education is not neutral, that what you learn is inseparable from how and in whose tongue you learn it — has never been more urgent. We will return to it when we reach the present day.

 

When Knowledge Becomes a Weapon

The Criminal Tribes Act — Born into Guilt

I want to turn now to what I consider one of the most morally extraordinary episodes in the history of colonial knowledge-power: a piece of legislation so radical in its premises that it deserves extended attention. I am speaking of the Criminal Tribes Act of 1871.

The Act did something that, stated plainly, sounds almost impossible to believe in a legal system that claimed to be governed by principles of evidence and individual culpability. It designated approximately 150 to 200 specific communities — nomadic pastoralists, itinerant traders, certain forest-dwelling groups, various artisan castes — as hereditary criminal tribes. Members of these communities were required to register with local police. They were prohibited from moving beyond their registered districts without police permission. They were subject to arrest without warrant. And they were assigned all of this — not on the basis of anything they had done. On the basis of who they were. On the basis of birth.

Where did this extraordinary claim — that entire communities were, by heredity, predisposed to crime — come from?

 

It came from colonial knowledge. Specifically, it came from a genre of ethnographic literature — the "criminal tribe" literature — produced by colonial officers and missionary anthropologists over the preceding decades. These writers had developed elaborate typologies of Indian criminals, categorising them by method of operation, religious custom, and caste affiliation. Their books bore titles like Wandering Tribes of India and Criminal Castes and Tribes of Southern India, and they were genuine achievements of detailed observation — their writers had spent time with these communities, learned something of their customs, and produced detailed descriptions of their ways of life.

 

But the framework within which these observations were organised was racial and cultural determinism. The assumption was that certain communities were, by their nature, constitutionally inclined to crime — that their social organisation, religious customs, and kinship patterns produced criminals the way that a factory produces goods. Individual variation within the community was, in this framework, epistemologically invisible. If a member of a designated criminal tribe did not commit a crime, that was not evidence against the theory; it was evidence of cunning concealment.

 

This is a spectacular example of what happens when a particular kind of knowledge — ethnographic observation, taxonomic classification — is operationalised through the machinery of state power. The knowledge does not have to be false to be dangerous. Some of it was detailed and accurate in its descriptions of specific practices. But it was organised within a framework of interpretation that rendered it, when translated into law, an instrument of mass persecution.

 

The Act was repealed after independence in 1952, replaced by the Habitual Offenders Act. But its legacy — in the form of communities whose social marginalisation has been reinforced over generations by police harassment, displacement, and denial of educational and economic opportunity — persists to this day. There are estimated to be some 60 million members of what are now called "de-notified tribes" in contemporary India. Their poverty, their exclusion, and their continued stigmatisation as communities of thieves and criminals cannot be understood without attending to the colonial knowledge apparatus that created them.

 

The Manufacture of Hindu-Muslim Conflict

I need to address one more dimension of the abuse of colonial knowledge before we turn to resistance. And this one is perhaps the most politically sensitive, because it touches on something that is still, today, a live and frequently violent political question.

 

The charge of "divide and rule" — the accusation that the British deliberately fostered religious and communal divisions among the Indian population as a mechanism of imperial control — is one that needs to be handled with care. The simple claim that the British "invented" Hindu-Muslim conflict is too crude and risks erasing the genuine theological and historical tensions between these traditions. But the specific forms taken by communal conflict in British India — the institutionalisation of separate religious electorates, the competitive politics of religious representation in Census returns, the administrative encouragement of communal organisations — cannot be understood without the colonial knowledge apparatus.

 

I have already mentioned the Census's role in hardening religious identities. But there is a further dimension. By making religious enumeration politically consequential — by establishing the principle that political representation should be allocated on the basis of religious population share — the colonial state created an incentive structure that rewarded religious mobilisation and punished cross-religious solidarity.

 

In this environment, what happened to the rich tradition of Hindu-Muslim syncretism that had characterised much of pre-colonial South Asian culture? It didn't disappear overnight. But it became politically inconvenient. Sufi saints who had been venerated by Hindus and Muslims alike became, in the increasingly communalised politics of the late colonial period, contested figures. Shared pilgrimage sites became flashpoints. The traditions of mutual participation in each other's festivals — which continued in many places well into the twentieth century — were increasingly viewed with suspicion by the organised communal organisations that the colonial political structure had brought into being.

 

To say this is not to blame the British entirely for Partition — the violence of 1947 and its causes are complex, and responsibility is widely distributed. But it is to insist that the colonial knowledge apparatus — the Census, the separate electorates, the administrative categories that hardened communal identity into political identity — created the conditions in which the worst outcomes became possible. Knowledge, here, was not merely power. It was the framework within which power produced catastrophe.

 

The Rebellion of 1857 — Knowledge, Rumour, and the Limits of Control

Let me take you now to a moment of crisis — a moment when the entire edifice of colonial knowledge-power came close to collapsing under the weight of its own contradictions. I am speaking of the Indian Rebellion of 1857.

 

In British historiography, this event is called the Indian Mutiny. In Indian nationalist historiography, it is called the First War of Independence. Both names are politically loaded; both capture something true. It began as a mutiny among Indian soldiers — sepoys — in the service of the East India Company, and it grew into something far wider and more complex.

 

The immediate trigger is famous: the introduction of new rifle cartridges that soldiers had to bite open with their teeth to load. A rumour spread through the military cantonments of northern India that the cartridges were greased with pork fat — offensive to Muslim soldiers — and beef tallow — offensive to Hindu soldiers. Whether the rumour was accurate is, historically, uncertain. The cartridges may indeed have contained animal fat; the British denied it, but their denials were not entirely convincing. What matters more, for our purposes, is what happened next.

Here is the question I want to put to you: was the cartridge rumour irrational?

 

The British thought so. They dismissed it as exactly the kind of credulity and fanaticism that justified colonial rule — evidence that Indians were not yet capable of rational thought, and that it was therefore not merely justified but necessary to govern them. But this interpretation is, itself, a form of colonial knowledge — a refusal to engage seriously with the actual content of the anxiety the rumour expressed.

 

Because the anxiety was not irrational. It was, in fact, very well-founded. Christian missionaries, with at least the tacit support of elements of the colonial administration, were active throughout northern India. The Governor-General Lord Dalhousie had openly stated his view that Christianity was the highest of religions and its spread in India was to be welcomed. Laws had recently been enacted facilitating conversion and protecting the property rights of converts. The educational policies of Macaulay and his successors were widely and reasonably understood as cultural and religious interventions — attempts to make Indians English in taste and intellect, which meant, in practice, to make them Christian in values.

 

And there was the doctrine of lapse. This was a policy under which, if a ruler of a princely state died without a natural male heir, the Company would "lapse" — annexe — the state rather than recognising an adopted heir. Dalhousie applied this policy ruthlessly, annexing the kingdoms of Satara, Jhansi, Nagpur, and Awadh in rapid succession. These annexations dispossessed not only their rulers but the entire network of institutions — temples, mosques, cultural establishments, scholarly communities — that the rulers had supported. The knowledge traditions and religious practices of these communities were directly threatened by the Company's territorial ambitions.

 

So when the cartridge rumour spread, it was not irrational. It was the expression, in a form available to people without access to official information, of a genuine and well-founded anxiety that the colonial state was engaged in a systematic assault on Indian religious and cultural life. The rumour was a form of knowledge — imprecise, potentially distorted, but pointing at something real. And the British dismissal of it as superstition was itself a form of ignorance — a refusal to take seriously the actual fears and experiences of the people they governed.

 

By June 1857, the mutiny had spread far beyond the military barracks. Agrarian communities, dispossessed landowners, urban craftsmen, and the residual court of the last Mughal Emperor in Delhi all joined what had become a broad-based uprising. It was suppressed by British forces — with considerable and often indiscriminate violence — by 1858. But it transformed British India permanently.

 

The East India Company was abolished. India came under the direct rule of the British Crown — the British Raj — governed by a Viceroy answerable to the Secretary of State for India in London. And the colonial state invested even more heavily in the institutions of knowledge-production: the Indian Civil Service examination system was reformed; the Survey of India and Census apparatus were expanded and systematised; the Archaeological Survey of India was founded in 1861, with the twin mandate of preserving India's historic monuments and establishing British authority over their interpretation.

 

Perhaps most importantly, the post-1857 colonial state adopted what its administrators called a more "scientific" approach to social management — informed by the vast ethnographic literature that had accumulated since the eighteenth century. A formal policy of non-interference in Indian social and religious custom was adopted. But this apparent withdrawal from cultural intervention was paired with ever more intensive knowledge-gathering about the customs in question. The state professed neutrality while its knowledge apparatus shaped the very phenomena it observed.

 

This is the paradox of mature colonialism at its most eloquent: the more knowledge the colonial state gathered, the more elaborate its claim to be governing in India's interests, and the more comprehensively it shaped Indian society in the image of its own administrative needs.

Fighting Back — Indian Responses to Colonial Epistemology

The Bengal Renaissance — Using the Master's Tools

I want to make sure I do not leave you with a picture of pure domination, because that picture would be both inaccurate and, in its own way, insulting to the millions of people who found, in the very instruments of colonial modernity, resources for challenging colonial rule.

 

The so-called Bengal Renaissance — a flowering of intellectual and cultural life in Calcutta, the capital of British India, in the late eighteenth and early nineteenth centuries — is the canonical example of this contradictory dynamic. Its most prominent figure, Ram Mohan Roy, is worth spending some time on.

 

Ram Mohan Roy was born in 1772 into a Bengali Brahmin family, and he died in 1833 in Bristol, England, where he had travelled to give evidence before Parliament on behalf of the Indian people. In the six decades between his birth and death, he remade the intellectual landscape of Bengal with remarkable energy. He founded the first Bengali-language newspaper, the Sambad Kaumudi in 1821. He campaigned tirelessly — and successfully — for the abolition of sati, the practice of widow self-immolation. He founded the Brahmo Samaj, a reformist religious and social movement that sought to combine the rationalism of the European Enlightenment with the philosophical core of the Vedanta — the ancient Indian tradition of non-dualist metaphysics rooted in the Upanishads.

 

Roy was, in short, a man who had absorbed European thought deeply enough to use it as a critical tool — but whose commitment to Indian intellectual traditions was equally genuine. He was not simply an Anglicist in Indian clothing. He insisted that the Sanskrit Upanishads contained a philosophical rationalism — a form of monotheistic, anti-idolatrous, deeply reasoned engagement with questions of existence and consciousness — that was every bit as sophisticated as anything in the European tradition. He translated Upanishads into English and Bengali. He challenged both the colonial dismissal of Indian learning and the orthodox Hindu defence of practices he regarded as corruptions of the tradition's deepest insights.

 

Roy represents one of the central strategies of the colonial encounter with knowledge: reclamation and reinterpretation. Not rejection of the colonial framework, but engagement with it on terms that asserted the equal dignity of Indian intellectual heritage. It was a strategy that required enormous confidence and enormous learning — and Roy had both.

 

Ambedkar — The Most Radical Critique

But the most radical and, in my view, the most intellectually rigorous engagement with the politics of knowledge in colonial India came not from the Bengali bhadralok — the educated upper-caste gentlemen of Calcutta — but from a man who was born into the most marginalised community in Indian society: a Dalit, from the Mahar caste of Maharashtra, who went on to earn advanced degrees from Columbia University and the London School of Economics, became the principal architect of the Indian Constitution, and subjected both colonial epistemology and its indigenous counterparts to a critique that remains unsurpassed.

 

I am speaking of Bhimrao Ramji Ambedkar — B.R. Ambedkar — whose life story is one of the most remarkable in modern history and whose intellectual legacy is one of the most inadequately recognised.

 

Ambedkar understood something that many of his upper-caste nationalist contemporaries were, for obvious reasons, reluctant to acknowledge. The colonial knowledge system was not the only oppressive epistemic framework operating on the Indian subcontinent. The knowledge traditions of classical Hinduism — above all the Manusmriti, the ancient legal code whose prescriptions defined the social and ritual order of caste — were themselves instruments of oppression. They had, for centuries, prescribed severe restrictions on Dalit access to education, public space, and religious practice. They had authorised violence against Dalits on the basis of their ritual pollution. They had, in effect, constructed a world in which an entire group of human beings was permanently and cosmically inferior.

 

Ambedkar's response to this was, in 1927, at the Mahad Satyagraha — an event whose symbolic weight I want to make sure you fully appreciate — to lead a march of Dalits to a public water tank in the town of Mahad from which they had been excluded by caste custom. They drank from the tank. It was a simple physical act. And then they publicly burned a copy of the Manusmriti.

 

Think about what that burning meant. It was not merely a protest against a specific law. It was an epistemological act: a refusal to accept the authority of a knowledge tradition that had denied the humanity of Ambedkar and everyone like him. It was a declaration that the sacred texts of a tradition are not beyond critique — that the question "whose knowledge?" applies to the Manusmriti as much as it applies to Macaulay's Minute.

 

In his 1936 text Annihilation of Caste, Ambedkar argued with forensic precision that caste was not, as its defenders claimed, an organic social arrangement suited to Indian conditions, but an ideological construction sustained by religious authority and enforced by economic and physical violence. The category of caste was not a social fact but a political instrument — one that served the interests of those at the top of the hierarchy by constructing a theological justification for the exploitation of those at the bottom.

 

This argument anticipated, in important respects, the later work of Foucault and Bourdieu on the relationship between symbolic violence — violence exercised through cultural and symbolic systems rather than direct physical force — and the reproduction of social inequality. Ambedkar reached these conclusions not through European philosophy but through a combination of the Buddhist and rationalist traditions and the painful evidence of his own life.

 

And here is the final, beautiful, tragic irony of Ambedkar's engagement with knowledge-power. He was, throughout his life, a tireless advocate for Dalit access to education — not because he thought the colonial education system was adequate or just, but because he understood that in the world as it was actually organised, literacy, degrees, and legal expertise were the only weapons available to those the system had placed at the bottom. He used the master's tools with devastating effect. And at the end of his life — in 1956, weeks before his death — he converted to Buddhism, along with half a million of his followers, in a mass ceremony in Nagpur. It was, among other things, an epistemological act: a search for an intellectual and spiritual tradition, available within the subcontinent's own history, that did not carry the hierarchical baggage of Brahminical Hinduism.

 

The Swadeshi Movement and the Reclaiming of Technology

One more form of epistemic resistance deserves attention: the Swadeshi movement that followed Lord Curzon's partition of Bengal in 1905. Swadeshi means "of one's own country." And at its heart was a claim about indigenous knowledge — specifically, about Indian technological and scientific capability.

 

Swadeshi activists pointed to something that the colonial narrative of Indian backwardness had carefully elided: the extraordinary sophistication of Indian manufacturing traditions, most spectacularly illustrated by the fine muslins of Dhaka. These fabrics — so fine that Mughal records described them as "woven air" or "running water" — required techniques of cotton cultivation and weaving that were among the most refined in the world. They had been among the most prized trade goods of the pre-colonial Indian Ocean economy, exported across the world and admired in courts from London to Beijing.

 

By the early nineteenth century, the Dhaka muslin industry was effectively dead. British tariff policy had imposed duties on Indian textiles exported to Britain while allowing British mill-produced cloth to flood Indian markets duty-free. The industrial revolution had made British cloth cheaper; the tariff policy ensured that Indian cloth could not compete. The weavers of Dhaka — thousands of families whose craft knowledge had been accumulated over generations — were ruined. Some accounts from the period, which are not regarded as reliable in all their details, record that their thumbs were cut off; what is documented is the destruction of their livelihoods and, with it, the technological knowledge they embodied.

 

Economic historians including Prasannan Parthasarathi, in Why Europe Grew Rich and Asia Did Not, and Utsa Patnaik, in her calculations of colonial economic drain, have argued that what the British liked to call Indian backwardness was in large part a product of deliberate colonial deindustrialisation — the systematic destruction of Indian manufacturing capacity to create captive markets for British goods. This is knowledge-power operating in the economic domain: the colonial state used its legislative and tariff-setting powers to make certain kinds of indigenous knowledge economically unviable, and then pointed to the resulting poverty as evidence of Indian incapacity.

What Was Inherited — The Postcolonial Condition

In August 1947, the British left India. Or rather — they transferred formal sovereignty to two new states: India and Pakistan. The flags changed. The governments changed. The faces at the top of the administrative pyramid changed. But something else did not change, and this something is the subject of our final section.

What happens to the structures of colonial knowledge-power when formal colonial rule ends?

The answer is: they persist. Sometimes they are modified, sometimes they are contested, sometimes they are partially dismantled. But the categories, institutions, and habits of mind through which colonial power had organised knowledge do not disappear with the departure of the colonisers. They are, in many cases, inherited — because they are embedded in the very architecture of the state that the postcolonial government inherits.

Consider the Census. Independent countries continued to conduct a decennial Census — and to enumerate the population by caste, religion, and language, using categories and methods continuous with the colonial apparatus. The political stakes of Census enumeration did not diminish with independence; in some respects they intensified. The Constitution's provisions for affirmative action — reservations in educational institutions and government employment for Scheduled Castes and Scheduled Tribes — made the official designation of these categories a matter of direct material consequence for millions of people. And the categories themselves were largely those first established by colonial administrators in the late nineteenth and early twentieth centuries.

 

The explosive politics of the Mandal Commission Report tells this story vividly. The Commission, appointed in 1978 and reporting in 1980, recommended reservations in government employment for Other Backward Classes — caste communities that were not designated as Scheduled Castes or Tribes but were nonetheless subject to significant social and economic disadvantage. The Commission's recommendations were finally implemented by Prime Minister V.P. Singh in 1990, and the result was nationwide protests — including self-immolations by upper-caste students who felt their prospects were threatened — that toppled the government. The politics of caste enumeration, in other words, remained explosive, contested, and consequential, a direct inheritance of the colonial apparatus.

 

Consider the educational system. Independent India and Pakistan continued to operate a higher education system structured around the English-medium universities founded under colonial rule — Calcutta, Bombay, Madras, Lahore — with the prestige hierarchy and the linguistic asymmetries largely intact. Access to elite institutions remained structured by class, caste, and language in patterns that reflected colonial precedent. Even today, the social premium attached to English-medium schooling continues to grow, creating a tiered system in which access to the globally dominant language of knowledge remains a marker and instrument of class privilege.

 

And consider the academy itself. The subaltern studies school — founded in the early 1980s by the historian Ranajit Guha, and including Gayatri Chakravorty Spivak, Dipesh Chakrabarty, and Partha Chatterjee — attempted to recover the history of those whose voices had been excluded from both colonial archives and nationalist historiography: peasants, women, workers, tribal communities. But as Spivak's celebrated 1988 essay "Can the Subaltern Speak?" asked with uncomfortable directness: is recovery even possible? The categories through which the subaltern's experience must be articulated — the language of rights, rationality, historical agency — are themselves constituted within the European epistemological framework. Even speaking on behalf of the excluded requires adopting the terms of their exclusion.

 

This is not a counsel of despair. But it is a recognition of the depth of the problem. Even resistance must operate within structures shaped by the power it resists.

 

The Digital Age and the New Asymmetries

We are approaching the end of our journey together. But I would be failing in my duty as a student of history if I did not bring the story to the present — because the entanglement of knowledge and power is not merely a historical phenomenon. It is happening now. And it is happening to us.

 

The global digital infrastructure of knowledge — the internet, search engines, social media platforms, AI systems — is overwhelmingly dominated by platforms and architectures developed in the United States and, to a lesser extent, Europe. Wikipedia, which for many billions of people is the primary freely accessible source of knowledge about the world, systematically underrepresents the histories, cultures, languages, and perspectives of the Global South. A 2018 study found that more than 80% of Wikipedia's editors were from Europe and North America. The histories, traditions, and perspectives of South Asia — let alone of Africa, Latin America, and the Pacific — are represented in proportion to the capacity and inclination of Western editors to attend to them.

 

The large language models on which AI systems are trained — including, it is fair to note, the systems that are increasingly being used in educational contexts, in legal systems, in medical diagnosis, in hiring decisions — are trained predominantly on English-language, Western-produced text. This means that their implicit frameworks for understanding the world, the categories they recognise, the relationships they identify, the assumptions they embed — all of these reflect, to a far greater degree than their creators usually acknowledge, a particular cultural, linguistic, and epistemological tradition.

 

This is not, I want to be clear, a technophobic argument. These are extraordinary and, in many respects, democratising technologies. But it is a Foucauldian argument: the question of what counts as knowledge, who produces it, and in whose image the systems of its organisation are built is never neutral. And if we do not ask these questions about the digital knowledge infrastructure, we risk reproducing, in a new form, exactly the asymmetries that colonial epistemology established — without the excuse of not knowing better.

 

Two centuries after Macaulay's Minute, the question of whose knowledge counts as knowledge, and whose power that knowledge serves, remains profoundly, urgently unresolved.

 

Conclusion: Giving the Phrase Its Weight Back

Let me try to bring this together.

 

"Knowledge is power." We began with this phrase. We have spent the last — well, a considerable time together — tracing its genealogy: from Francis Bacon's vision of man's sovereignty over nature, through Foucault's insight that power produces knowledge and knowledge produces power, through Said's demonstration that European scholarship about "the Orient" was a material instrument of colonial domination. We have watched it operate, in concrete historical detail, in the survey that remapped a continent, the census that froze a society's identities into administrative categories, the education policy that tried to remake millions of minds in a colonial image, and the legislation that designated whole communities as criminals from birth.

 

We have also watched it be contested, resisted, and subverted: by Ram Mohan Roy's insistence on the philosophical depth of Indian classical tradition; by Ambedkar's burning of a sacred legal code and his lifelong insistence on the right of the excluded to education; by Gandhi's critique of the colonial language of knowledge; by the Swadeshi movement's recovery of indigenous technological heritage; by the subaltern studies school's attempt to hear the voices that archives had silenced.

 

"Knowledge is power" is not a simple truth. It is a historically specific formulation, born in the intellectual revolution of early modern Europe, which elevated the instrumental domination of nature above all other intellectual values. In the hands of the British colonial project in South Asia, it became the justification for one of the most comprehensive exercises in epistemological violence in modern history. It was also, in the hands of those who used its tools against it — who took the British-funded schools and the colonial archives and the European universities and used them to argue for their own dignity and freedom — one of the animating principles of liberation.

 

Knowledge, in other words, is not only power-over. It is also power-for, power-with, and power-against. The same phrase that justified the Criminal Tribes Act was the phrase that B.R. Ambedkar wrote above his desk when he was studying in New York.

 

But — and this is my final point, and I want you to leave with it — the phrase is most dangerous when it circulates without this history. When it is taught, unexamined, in the schools of postcolonial societies as a simple motivational maxim — as though the relationship between knowledge and power were always benign, as though the question of whose knowledge and whose power never arose — it perpetuates the epistemic innocence that colonialism required and that its legacies continue to demand.

 

To recover the phrase's full weight — to hear in it both the promise of liberation and the menace of domination — is the beginning of the critical thinking that genuine education requires. Not an education that teaches you what to know. An education that teaches you to ask: how is this knowledge produced? Who benefits from it? What does it make possible? What does it make invisible? And whose voice is absent from it?

 

For those who have lived on the wrong side of the relationship between knowledge and power — whose languages were deprecated, whose customs were catalogued and controlled, whose children were educated in a foreign tongue, whose communities were designated criminal by the scratch of an administrator's pen — the phrase "knowledge is power" has never been innocent. Their understanding of it, won through historical experience rather than philosophical abstraction, may be the most accurate understanding of all.

 

 

---

[i] These pre-modern views, while emphasizing humility and spiritual growth, were not without their hierarchies. For instance, access to 'ilm' in the Islamic Golden Age or 'episteme' in Greek philosophy was often restricted by class, gender, or patronage, foreshadowing how knowledge could serve exclusionary power even before European instrumentalism