Start where the old code does: the Shudra and the untouchable were barred from learning — but be exact about which learning, because the exactness is the whole argument. Bühler’s public-domain Manu opens study of the Veda to “the three twice-born castes” alone (10.1) and defines the Shudra as “one birth only” (10.4); the harsher enforcement is in the Gautama Dharmasūtra, which orders molten tin poured in the ears of a Shudra who so much as hears the Veda recited (12.4). The bar was never on literacy as such — a Shudra could keep accounts — it was on the sacred and legal knowledge that conferred authority. So the plain question — who first opened a classroom to the people the śāstra shut out? — has an untidy, honest answer.
Before any statistics: a thousand years of the rule
No one counted literates before the nineteenth century, so for the whole classical and medieval span the evidence is not statistics but two other kinds — what the law prescribed, and what outsiders observed. Both point the same way. The historian R. S. Sharma, in Śūdras in Ancient India, traces how the Dharmaśāstra fixed the Shudra outside the twice-born who alone received the sacred thread and the Veda; the exclusion was not a stray verse but settled architecture. And you need not take a modern scholar’s word for it: Al-Biruni, the Central Asian scholar who studied India around 1030 CE, recorded from the outside that the Brahmins guarded the sciences of the Veda and that the Shudra and the outcaste were forbidden even to hear them recited — Gautama’s molten-tin rule still operating a thousand years after it was written. [Documented]
There was one place learning slipped the varna gate: the heterodox śramaṇa traditions. When the Chinese monk Xuanzang studied at Nalanda around 637 CE, he described a monastic university teaching grammar, logic, philosophy, medicine, mathematics and astronomy — admission by examination, not by birth, to monks drawn from across Asia. Buddhist and Jain monastic schooling was the real alternative to the Brahmanical monopoly, open in a way Vedic study was not. But keep it in proportion: it was monastic and elite, not a village school for the untouchable, and the orthodox order had no love for it — Nalanda itself was sacked around 1193. [Debated] on how far down the caste order that opening actually reached.
So across a thousand years the picture holds: the sacred-legal gate stayed shut, confirmed by an outside witness at 1030; the one breach was the monastic tradition the old order distrusted. But the record does not leap from Al-Biruni to the Company’s surveyors — six centuries of Sultanate and Mughal rule lie between, and they change the picture in one way while leaving it untouched in another.
The medieval middle: three worlds of learning, one gate still shut
From the thirteenth to the eighteenth century, learning in India ran in three separate channels. The first was new: the Islamic maktab and madrasa. The Delhi Sultans were serious patrons — Firoz Shah Tughlaq (r. 1351–88) alone endowed dozens of colleges and built the great Madrasa-e-Fīrūz Shāhī at Hauz Khas around 1352, and by the fourteenth century Delhi held hundreds of madrasas teaching Qur’an, law, logic and the sciences. Under the Mughals it deepened; Abu’l-Fazl’s Ā’īn-i-Akbarī (c. 1590) records Akbar’s court reforming the syllabus toward arithmetic, geometry, astronomy and medicine. But this was a new ruling-class literacy with its own gates of faith and rank — and it left the Hindu caste-gate entirely alone. [Documented]
The second channel was the old one, unchanged. Sanskrit learning continued in the tol, the pāṭhaśālā and the chatuṣpāṭhī — at Mithila, at Nadia in Bengal (a centre of Nyāya from the 13th to 16th century), at Kashi — and it stayed exactly what Manu had made it: birth-gated. The Maithil Brahmin identity of the period is described as a triad of learning, ritual purity and lineage — three words that mean the same thing. Six hundred years after Al-Biruni, the Veda and its sciences were still transmitted Brahmin to Brahmin, by blood. [Documented]
The third channel is where the excluded finally appear — not admitted to the first two, but building their own. The Bhakti and Sant poets made a devotional literacy in the spoken languages, deliberately turning from Sanskrit: Kabir the weaver, Ravidas the leather-worker, Chokhamela the Mahar untouchable barred from the very temple he sang to, Namdev and Tukaram in Marathi. That the people who carried medieval India’s most enduring literature were low-caste and untouchable — and had to compose in Hindi, Marathi and Punjabi because the Sanskrit door was shut — is itself the proof of the gate. And here the site keeps its own honesty: Ambedkar judged the Bhakti saints a disappointment, holding that they sought dignity within caste rather than its abolition, and so let the structure stand. The crack was real; it was not the demolition. [Documented] on the saints’ origins and vernacular turn; [Debated] on how much they changed.
So six medieval centuries add a second and third channel of learning, but the first gate — the śāstric one, on sacred and legal knowledge — never opens. What is still missing across the whole span, ancient and medieval alike, is a single number. The counting begins only when the Company’s surveyors arrive.
“But the old schools were full of Shudras” — read the survey to the top
The sharpest reply you will hear runs like this: the śāstric bar is a colonial myth. Before the British, indigenous schools were everywhere and open — a Gandhian scholar proved it. The scholar is Dharampal, whose 1983 book The Beautiful Tree compiled the East India Company’s own Madras Presidency education survey of 1822–25 and William Adam’s Bengal–Bihar reports of 1835–38. The right cites these to say caste never closed the classroom.
Concede what is true first, because it is true and this site does not fight the record. Those surveys are real, and their headline is genuine: at the elementary level, schooling was far broader than the “British educated India” story admits. In the Madras data, only about 24% of pupils were Brahmin, Kshatriya or Vaishya together, while roughly 65% were Shudras — and in Bengal and Bihar, Adam found children (and teachers) drawn from low castes like the Chandal, Kahar and Malo. A Shudra learning letters and accounts in a village school breaks no rule of Manu, and the numbers show plenty of them did. [Documented]
Now read the same survey to the top of the tree — the part rarely quoted. Dharampal’s own tables record that higher learning was a Brahmin near-monopoly, and monopolised most tightly in exactly the śāstric subjects: Theology, Metaphysics, Ethics, and Law. The shape is a gradient: broad at the base (reading, writing, arithmetic), Brahmin at the apex (the sacred and legal knowledge that conferred authority). That is not a refutation of the two codes — it is the two codes, drawn as a graph. Manu opened Veda study to the twice-born only; the colonial survey, taken two thousand years later, still finds theology and law in Brahmin hands. The bar had migrated from statute to structure, but its outline was unchanged. [Documented] on the caste gradient; [Debated] on how much of it to attribute to the old code versus later economics.
Two honesties keep this from overreaching. First, do not claim the untouchables were absent to the last child — Dharampal himself resists that, and the site follows the evidence over the slogan: the lowest castes were the most drastically under-represented, but “the most excluded” is the defensible claim, not “zero everywhere.” Second, the surveys measured access to schools, not the ritual ceiling — the molten-tin verse was never about arithmetic. Put the two together and the right’s favourite source stops arguing against the thesis: read only its first line and caste looks open; read it to the canopy and the old code is right there in the branches.
The colonial hinge: the Charter Act of 1813
For decades the East India Company kept missionaries out, fearing they would unsettle its trade. That changed with the Charter Act of 1813: a “pious clause,” pressed by William Wilberforce and the Clapham Sect, finally admitted Christian missionaries to India — and the same Act made the first state commitment to education, earmarking “not less than one lakh of rupees” a year.
[Documented] Missions had in fact begun even earlier where they could — a school in Madras in 1805, one at Visakhapatnam in 1806 — and after 1813 they expanded fast. By 1901, mission institutions in the Madras Presidency numbered 3,778, a large share of all schooling in the province, reaching lower-caste children and girls whom the old order had excluded by design.
Macaulay’s ceiling
Do not romanticise the colonial side, though. Macaulay’s 1835 Minute — quoted verbatim elsewhere on this site — set the ceiling deliberately: a thin class “Indian in blood and colour, but English in taste,” not mass literacy. His theory was downward filtration — educate a sliver at the top and let learning trickle down. It did not trickle. [Documented]
Nineteen years later the stated policy changed on paper. Wood’s Despatch of 1854 — the “Magna Carta of English education in India” — repudiated pure filtration, called for a graded system from village school to university, and committed the state, in principle, to mass and vernacular education (it founded the education departments and, in 1857, the first universities). But a commitment on paper is not a school in a village: a generation later the Hunter Commission was still recording how little had reached the lowest castes. “The British educated India” is a myth; the more accurate sentence is that missions and Indian reformers opened schooling to the excluded, against both śāstric prohibition and a colonial state whose promises ran well ahead of its delivery. [Documented]
The demand is the evidence
Here is the cleanest proof the classroom was closed: reformers had to keep demanding it be opened. In October 1882, Jyotirao Phule submitted a written memorandum to the Hunter Commission — the Empire’s first great education review — asking for free and compulsory primary education for the children of the lower castes. The Commission’s own report, in turn, recommended that indigenous schools “be open to all irrespective of caste, creed and sex” and that the state give financial help to draw the depressed classes into education. [Documented]
Read those two facts as a logician would. You do not petition to open a door that stands open; a state commission does not recommend admitting a caste that was never barred. The 1882 record only makes sense against a classroom that excluded by design — which is precisely the old code’s design. The reform movement is not evidence against the śāstric bar; the movement exists because of it.
And getting in was only half the battle — the system then leaked the excluded back out. The Hartog Committee (1929), the Simon Commission’s education arm, put a number on it: of every 100 pupils in Class I in 1922–23, only 18 reached Class IV three years later — the rest “relapsed into illiteracy.” For children with no literate household to fall back on — overwhelmingly the low-born — that wastage hit hardest. Nominal admission was not the same as a share of learning. [Documented]
The southern thread: an older vernacular, the first remedy
The story so far is largely northern; the south ran differently, and the difference matters. Tamil carried a literary tradition parallel to Sanskrit and, at its root, broader. The Sangam poems — dated by linguists to roughly 100 BCE–250 CE — were composed by some five hundred poets drawn from across society: Brahmins were present but not dominant, alongside merchants, cultivators, a textile trader, blacksmiths, soldiers, kings and women. The rigid fourfold Vedic order had not yet closed over Tamil life. [Documented] on that social breadth; [Debated] on the legendary “three academies” later tradition built around the poems — those are a medieval story, not a datable institution. Kannada and Telugu then grew their own letters — Pampa (c. 939 CE) and Nannaya (c. 1025 CE) — though these first rendered the Sanskrit epics into the vernacular, widening access to their content more than breaking the caste of their authors.
By the colonial era the south had inverted that inheritance. Madras schools and colleges funded Sanskrit and Brahmanical texts far above the Dravidian vernaculars, and a tiny Brahmin minority held the overwhelming share of the new English education and public service — the 1912 figures charted on the reservation page (82% of sub-judges from about 3% of the people). A region with an old, broad literary culture now had the narrowest possible gate on the new learning. That gap lit the sharpest anti-caste politics in India.
And it produced the first remedy. Out of the Non-Brahmin (Justice) movement came the Madras Communal Government Order of 16 September 1921 — India’s first reservation, twenty-nine years before the Constitution — apportioning public posts across communities (non-Brahmin, Brahmin, Muslim, Christian and Anglo-Indian, and the Scheduled Castes). The device the new code would later make national was pioneered in the south, and it came straight out of the classroom-and-office grievance this essay traces. [Documented]
The backlash — from the national movement
Here is the part that complicates every side. Opening the classroom to women and the low-born drew fierce opposition — and not only from priests, but from within the nationalist mainstream.
[Documented] Bal Gangadhar Tilak, through his paper Kesari, opposed the Age of Consent Act (1891) — which raised the age of consent for girls — as illegitimate colonial interference in Hindu custom, and he attacked reformers who leaned on missionary or colonial support. Among his targets was Pandita Ramabai, a Sanskrit scholar who had converted to Christianity in 1883 and founded the Sharada Sadan to educate women, especially widows.
Be fair about the argument: Tilak’s stated principle — that reform should come from within Indian society, not be imposed by the coloniser — is not a stupid one, and this site does not caricature it. But the effect of that principle, in 1891, was to oppose raising the age at which a girl could be married, and to discredit a woman who was educating women. Principle and consequence have to be weighed together.
The witness in the classroom
Statistics have a face. The man who would draft the Constitution’s equality clauses had himself been an untouchable child in a colonial school, and he wrote down what that meant. In Waiting for a Visa (1935–36), Ambedkar recalls that admission was not inclusion:
I knew that in the school I could not sit in the midst of my classmates according to my rank, but that I was to sit in a corner by myself… I had to squat on a piece of gunny cloth which I had to carry to school and take back with me… The teacher would not touch it, and the servants would not touch it.
And even water was rationed by caste:
The tap was on the upper floor. One of the servants of the touchable class had to open it. Unless he opened it, it was not possible for me to quench my thirst. In my case the permission of this servant was necessary… The situation can be summed up in the statement — no peon, no water.
[Documented] This is not a pamphlet’s claim about the old order; it is a first-person witness, written before independence by the man who then wrote the remedy. Ambedkar’s works are public domain in India. The child who carried his own gunny sack is the same jurist who put Article 17 — the abolition of untouchability — into the founding text. The essay’s whole argument runs through that one life.
The two codes on the classroom
The contest resolves, as always, into the two codes. The old one forbade the Shudra to study and a woman to be independent. The new one makes the opposite an enforceable right: Article 15(4) allows special provision for backward classes in education, Article 45 / 21A make schooling a duty and then a fundamental right, and Article 46 directs the State to promote the education of the weaker sections. What missions, Phule, Ramabai and later Ambedkar fought for against śāstra and empire alike, the Constitution simply guarantees: the classroom is not reserved by birth.
What the numbers show — where numbers exist
Ask the natural question — what was the literacy rate under the old code? — and the honest answer is that there wasn’t a measurement. Pre-colonial India ran no census and counted no literates, so any figure put on the Dharmaśāstra era would be invented. What is documented is not a rate but a prohibition (Manu 4.80–81, 10.4). The measured record only begins with the colonial census, around 1881. So we start the line there, and leave the era before it honestly blank.
Literacy in India, 1901–2011 — and the caste the old code barred
Overall literacy — ~5% (1901) → 18% (1951) → 74% (2011), ~80% today
Scheduled Castes — the caste the old code barred by law: ~10% (1961) → 66% (2011)
[Documented] Census of India literacy series (Census 1901–2011; SC literacy 1961–2011). Pre-1881: no measurement — the old code's bar is a legal prohibition, not a rate.
Read the two lines together and the whole thesis is in one picture: under the old code the excluded had no measured literacy and a written ban on learning; a century after the census began — and, decisively, after 1950 — the very caste the śāstra shut out of the classroom crossed from near-nothing to two in three literate. The line is not finished, but its direction is the argument.
Now anticipate the sharpest objection, because it is a fair one: literacy rose for everyone — the national figure already contains SC, ST and OBC, so isn’t the Scheduled-Caste rise just general development, not anything the Constitution did? Answer it honestly, in two parts.
First, read the gap, not the level. In 1961 Scheduled-Caste literacy (~10%) was barely a third of the national rate (~28%); by 2011 it (66%) was nearly nine-tenths of it (74%). The formerly-excluded did not simply rise with the tide — they closed the distance. A rising tide lifts every boat by the same amount; it does not make the lowest boat catch up to the others. Convergence is the signature of a barrier removed, not merely of general growth.
Second, a group under a written legal prohibition cannot rise at all while the prohibition holds. The near-zero start is not underdevelopment — it is enforced exclusion, on the record in the śāstra itself. Something had to lift the bar before any curve could begin.
Be equally honest about what this does not prove. How much of the catch-up to credit to the Constitution’s guarantees, how much to reservation, how much to missions and reform movements, and how much to broad modernisation is a genuine, unsettled question — and this site does not pretend to settle it. But the shape is not in dispute: the bar came down, and the barred caught up. [Documented] on the prohibition and the convergence; [Debated] on the exact causal weights. (Scheduled Tribes track the same path — roughly 8–9% in 1961 to 59% in 2011.)