top of page

Who May Drink From the Tank? — Mahad, the family and the langar, and the long opening of India's institutions

6 hours ago
31 min read

Blossomed, Not Assembled · Volume V of VI · Parts Twenty to Twenty-Three

The earlier volumes argued that India cannot copy its way to development, that its inheritance must be sifted rather than worshipped, and that its institutions were designed, written down and at least once revised by people who did not build them.

Most older Indians live with their children, and that is a real strength. The same homes, castes and temples also locked people out of water, land and food. Mahad and the langar show who opened them, and why care that forbids exit is not care.

In Brief: For readers who value India's family and community life and want to know what in it a developed India should keep. The article shows the real strength of the household as India's welfare system, with the evidence on how older Indians live, and the real cost: the same family, caste and temple locked people out of water, property and food. Through Mahad in 1927, the opening of Hindu family law and the taking back of Sikh gurdwaras, it asks who keeps an institution, who may use it, and who can force it open.

India's families, temples and communities have done more of its caring than its states ever did, and a developed India should keep that care while forcing open every room those same bodies locked, because care that forbids leaving is not care. Most older Indians still live with their children. The same bodies kept people from water, property, learning and food, and they did not open those rooms on their own. This article follows three openings: a water tank at Mahad, the Hindu family in law, and the Sikh community kitchen. It finds that each needed a movement, the state and a slow change in what people do, and that a right written in law is not yet a right in use. It leaves open how much of the old care can survive once everyone inside it is free to leave.

Opening — The Day a Crowd Drank

On 20 March 1927, in the small Konkan town of Mahad, a crowd walked through the main streets to a public tank called the Chavdar tank. They were people of the castes treated as untouchable, gathered for a conference with B. R. Ambedkar. At the tank, Ambedkar drank, and the people with him drank. In his own later account, it was the first time they had drunk that water, though their town had already declared it public.¹

The town answered quickly. A rumour spread that the crowd meant to enter a temple next. Men of the higher castes attacked them. Ten years later the Bombay High Court described it in one plain sentence: "They were assaulted and beaten by the caste Hindus and there were criminal prosecutions."²

Then came the cleaning. The higher castes held a ritual to purify the tank, which in their eyes the drinkers had defiled.³ Later accounts say priests chanted while 108 pots of a mixture of cow dung, cow urine, milk and curds were poured into the water.⁴˒⁵ I have not found a record from 1927 that gives those details. The water had to be made clean again because certain people had touched it.

The tank was not private. The legislature of the Bombay Presidency had resolved in 1923 that public watering places must be open to the untouchable castes.³ The Mahad municipality had passed its own resolution on 5 January 1924 opening this very tank.¹ The law was already on the side of the drinkers. The town was not.

Mahad is usually told as a story about caste, and it is one. I think it is also a story about a question older than caste and wider than water. Who keeps an institution, who may use it, and who can force it open? A tank, a kitchen, a family and a gurdwara all have keepers and users, and the two are not always the same people. The gap between them, and the wisdom that grew on both sides of it, is the subject here.

Part Twenty — A Society Thicker Than Its State

Who looked after people before the modern state?

Mostly, not the state. The previous volume in this series stood at a different tank, in the Tamil village of Uttaramērūr, whose tank committee was chosen by drawing names from a pot, under rules carved on a temple wall around 921 CE.⁶ The design there was the work of a few, an assembly with a royal officer beside it. The keeping was the work of many, every year, for generations, by people who were not present when the tank was dug.

Design says who sits on the committee. Keeping decides who drinks. On this subcontinent much of the keeping was done by bodies between the person and the king. The guild governed its trade and supported its members. The temple fed pilgrims, taught students, maintained tanks and lent to villages. The village assembly kept its roads and water. Networks of caste and community housed the migrant and arranged marriages for the orphan. Before all of these stood the family.

The state was there, and at its best it acted. Aśoka, who ruled most of the subcontinent in the third century BCE, says in his second Rock Edict that he arranged "two types of medical treatment: medical treatment for humans and medical treatment for animals." Where medicinal herbs did not grow, he had them brought and planted. "Along roads I have had wells dug and trees planted for the benefit of humans and animals."⁷

An emperor digging wells is a good emperor. He was still not the one who taught the child, lent to the farmer or fed the widow on an ordinary day.

The roof and the house

I call this a society thicker than its state. It helps explain why so much survived the fall of so many kingdoms. When a dynasty collapsed, the guild still wove, the temple still taught, the assembly still met and the family still fed its old. The state was the roof. Society was the house.

The picture is too neat, and I keep it only with two corrections. Kings did more than shelter. They granted land to temples and villages, appointed officers and stepped into local affairs, and the Uttaramērūr rules were made with a royal officer present. State and society were woven together. And the bodies that outlived dynasties did so, I think, mostly by changing, doing different work under each new power while keeping their names.

Much of the development thinking India adopted in the twentieth century assumed the reverse. It expected the state to be the house, and society to move in. Development is made by citizens and by the bodies between them and the government, not by government alone.

Catholic social teaching calls the general rule subsidiarity. Pope Pius XI put it sharply in 1931. It is "an injustice and at the same time a grave evil and disturbance of right order" for a larger body to take over what smaller bodies can do.⁸ A task should go to the smallest body able to do it, and larger bodies should help the smaller ones, not replace them. A tank committee chosen in the village is subsidiarity cut in stone.

Smallest is not the same as most legitimate. A small body is also the easiest for a dominant family or caste to capture, as Mahad will show. So the rule this article uses has two halves. Subsidiarity for provision, constitutional supremacy for rights. The smallest able body should run the tank. No body, however small or old, may decide who drinks from it.

Sharing as a duty

The deepest of these bodies is the household, and here the old texts speak directly. The Manusmṛti, a law book composed around the start of the common era, lists five daily offerings a householder owes. The fifth is nṛyajña, the offering to human beings, and Manu defines it in two words, atithipūjanam, the honouring of the guest.⁹ Feeding the stranger at your door was not a kindness you might choose. It was a daily debt.

The Ṛgveda's hymn on giving puts the same idea more fiercely:

nāryamaṇaṃ puṣyati no sakhāyaṃ kevalāgho bhavati kevalādī He feeds neither friend nor companion. The one who eats alone eats nothing but sin. (Ṛgveda 10.117.6, my plain rendering, after R. T. H. Griffith's translation of 1896)¹⁰

The verse just before it gives a reason, and the reason is an economist's:

o hi vartante rathyeva cakrānyam-anyam upa tiṣṭhanta rāyaḥ Riches roll like the wheels of a chariot. They come now to one, now to another. (Ṛgveda 10.117.5, after Griffith)¹⁰

Today's giver may be tomorrow's beggar. On this reading, sharing is not charity. Sharing is insurance written as ethics. A family that feeds its old, and a village that feeds a traveller, are buying cover against the turn of the wheel that will one day come to them.

Isn't "sharing is dharma" just a slogan?

It is a slogan when it is said and not shown. The texts above do treat sharing as dharma, here meaning a duty that holds a household and a society together.

A verse is evidence of an idea, not proof that people lived by it. The same tradition that told the householder to honour every guest also kept rules, in its law books and in village practice, about who could eat with whom and who could touch whose water. When the India Human Development Survey asked households about untouchability in 2011–12, its follow-up question was whether a Scheduled Caste person could enter their kitchen or use their utensils.¹¹

The sharing the texts command was real, and it had walls. Mahad is what the walls looked like in 1927. Any honest use of "sharing is dharma" has to answer a plain question. Sharing with whom?

The old-age ledger

The household's sharing still carries great weight, and the numbers deserve careful reading. The Longitudinal Ageing Study in India, Wave 1, a national survey of more than 72,000 people aged 45 and over carried out in 2017–18, asked older Indians how they live.¹² Among those aged 60 and above, 40.6 per cent live with a spouse and children, and another 27.6 per cent live with children and others. About two in three older Indians, 68 per cent, live with at least one child.¹² Only 5.7 per cent live alone.

No budget line pays for that arrangement, and no pension fund backs it. The United States runs one of the world's largest public pension systems. Its 2026 Trustees Report projects that the main retirement trust fund will run out of reserves in the last quarter of 2032. Benefits would not stop. Incoming taxes would still pay about 78 per cent of what is promised, unless the law changes.¹³ A public system that falls short still pays most of the cheque. A family that falls short can pay nothing at all.

What living together does not prove

Living with a child is not the same as being cared for by one. The survey counts who shares a roof. It does not show who controls the income, who makes the decisions, or whether the old person is welcome or only tolerated. An older person in a joint household may be the one supporting it.

The better test is what older people themselves want: to choose where they live, to keep control of their own money and property, and to be heard. When family care fails, the law now offers a remedy. The Maintenance and Welfare of Parents and Senior Citizens Act of 2007 lets a parent who cannot maintain themselves claim a monthly allowance from adult children before a tribunal. It also lets a parent undo a gift of property made on a promise of care that was not kept.¹⁴

The Indian ledger also has a cost that appears in no account, and someone pays it. In the same survey, older women are more than three times as likely as older men to live alone, 9.0 per cent against 2.6 per cent.¹²

Inside the homes where old people are looked after, the looking-after is done mostly by women. India's Time Use Survey of 2024 found that women who did unpaid domestic work for their households spent about 289 minutes a day on it, against 88 minutes for men who did such work. Women who gave unpaid care to household members spent about 137 minutes a day on it, against 75 for men.¹⁵

Isn't the joint family the answer to the pension problem?

It is part of the answer, and only part. The family carries a great share of India's social security on a ledger no national account records.

A developed India should not treat that ledger as a relic that prosperity will dissolve, as it dissolved in many rich countries. It should support families as places of care, through housing, tax and labour rules that make it easier for generations to live near one another. And it should build the state's net for those whom no family holds: the widow without sons, the childless couple, the migrant whose children are in another state.

There is one condition, and the rest of this article is about it. Care that has to be kept up by forbidding anyone to leave is not care. Family care is not exploitation in itself. A family-care system that works only because women cannot refuse it is unjust. It is a debt paid by the person with the least power to refuse it.

Part Twenty-One — The Locked Rooms and the Long Road to Mahad

The same bodies that cared also coerced

The thick society that sustained India also coerced within it. The record is long, and it can be put in four lines:

  • The joint household that cared for the old also confined the young wife, often for most of her life.

  • The community that sheltered its members also shut others out.

  • The guild that protected its own kept outsiders from the trade.

  • The village that kept its tank kept some of its people from drinking from it.

The wisdom is the principle that care is best given by those bound to you, and that no one should eat alone. A second principle needs its own sentence. An obligation that is felt is often stronger than an entitlement that must be claimed, because it does not wait to be demanded. It is also easier to abuse, because no one can appeal against a duty. Obligation gives a society its warmth. Rights protect its members when the warmth is absent. A humane society needs both and cannot use one in place of the other.

The coercion is none of these. It is the rust on the wisdom, and rust can eat through the metal if nobody scrapes it off.

Mahad, step by step

Historically, the thick society did not open its own locked rooms. Mahad shows how long the opening took, and how many times the words came before the water.

On 4 August 1923 the Bombay Legislative Council passed a resolution moved by the reformer S. K. Bole, that the untouchable classes be allowed to use all public watering places. In September the government told its departments to act on it.³ The Mahad municipality's resolution of January 1924 followed. Nothing changed at the water. The higher-caste residents kept the untouchable castes away, and the municipality did not enforce its own decision.³

Then came 20 March 1927, the drinking, the beating and the purification. Ambedkar called a second conference at Mahad for 25–27 December 1927. Before it could meet, a group of caste Hindus filed a civil suit on 12 December, claiming the tank as private property. On 14 December the court granted a temporary injunction barring Ambedkar and the untouchable castes from the tank.¹ A court order now stood between the drinkers and water their own municipality had opened.

The conference met anyway. With the matter before the court, the people did not march to the tank.⁴ On 25 December 1927, at that conference, a copy of the Manusmṛti was publicly burned. Ambedkar told the gathering why they had come: "We are not going to the Chavdar Lake merely to drink its water. We are going to the Lake to assert that we too are human beings like others."³

The case went on for ten years. The caste Hindus claimed an ancient custom of excluding the untouchable castes. The trial court held that they had not proved it, and the appeal court at Thana agreed.³ On 17 March 1937 the Bombay High Court dismissed the last appeal on the same ground.² The judgment settled one tank in one town. It did not make a rule for India.

Why did Ambedkar burn the Manusmṛti?

He burned it because it was a text often cited to justify the wall. By December 1927 the people of Mahad had been beaten, their water had been purified against them, and a court had barred them from a public tank. The Manusmṛti was a book that gave the exclusion its language of purity and rank. Burning it was a public refusal to accept that a text could decide who was human.

Readers who value the Manusmṛti will feel the force of that act, and they should. The same book is quoted in this article for its rule that a householder must honour every guest. Both things are in it. Earlier in this series I argued that an inheritance has to be sifted, not swallowed whole. Ambedkar chose fire after four years of polite resolutions had not opened one tank.

Didn't the Constitution end untouchability?

In law, yes. In use, not yet. The Constitution came into force twenty-three years after the drinking. Article 17 says: "'Untouchability' is abolished and its practice in any form is forbidden." Article 15(2) says that no citizen shall, on grounds only of religion, race, caste, sex or place of birth, be barred from "the use of wells, tanks, bathing ghats, roads and places of public resort" kept up by the state or dedicated to public use.¹⁶ The tank that Mahad's residents guarded became, in the country's highest law, a tank for everyone.

Parliament then had to keep writing. The Untouchability (Offences) Act of 1955 made it a crime to deny anyone the use of a well, tank or bathing ghat on grounds of untouchability. In 1976 it was strengthened and renamed the Protection of Civil Rights Act.¹⁷ In 1989 came the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act.¹⁸ Each law exists because the one before it was not enough.

The surveys show how far use still lags. In the India Human Development Survey of 2011–12, run by the National Council of Applied Economic Research and the University of Maryland across more than 42,000 households, 27 per cent of respondents said someone in their family practised untouchability.¹¹ That is more than one household in four, sixty years after Article 17. The figure counts only those willing to admit it to a surveyor. It is a floor, not a ceiling.

Three hands opened the tank

A right is a door. A door in the law can stand open for decades in a wall that is still shut in the village. Opening a locked room has taken three hands, and Mahad shows all three.

The first hand is the movement. People walked to water knowing they might be beaten, and they were. No law opened the Chavdar tank on 20 March 1927. A crowd did.

The second hand is the state. Only a public power can turn a claim into a rule that binds the whole country and punish those who break it. The court in 1937 opened one tank in law. The Constitution in 1950 made the same opening a right of every citizen.

The third hand is society itself, which has to change what it does. That hand is the slowest, and the survey of 2011–12 says it has not finished.

Is the state the hero here?

No. The state is a chisel, and a chisel cuts both ways. At Mahad the colonial state passed a resolution in 1923 and did not enforce it. Its municipality opened the tank and then let the residents close it. Its civil court barred the drinkers from the water by injunction, and the case took a decade to end.

At the end of 2022, three of every four prisoners in India's jails, 75.8 per cent, were awaiting trial.¹⁹ A citizen needs the state's protection against the locked room, and protection against the state's own rod. A right of access that the local police will not defend is a sentence on paper.

Aśoka and the public standard

Aśoka knew the chisel too. His first Rock Edict forbids animal sacrifice in his domain. The fifth tells of officers he created thirteen years after his coronation, the dhamma-mahāmātras, officials of right conduct, who "work among all religions" and seek the proper treatment and release of prisoners.⁷ Here a public standard is set over local custom, and officers are paid to see it kept.

Here the likeness ends. Aśoka's standard was a king's. It bound his subjects, not himself, and the subject had no court in which to argue back. The Constitution a citizen can invoke against a village is not a royal edict. What the two share is the idea that some rules must stand above the custom of any village, caste or family.

This is why the rule set out earlier has two halves. Subsidiarity decides who runs the tank. The Constitution decides who drinks. In a developed India a citizen must be able to call on the Union against a panchayat, a family, an employer or a platform, when care has become a cage.

Part Twenty-Two — The Family Opened by Law, Room by Room

The hardest case

The family is the hardest case. It is the institution this article most wants to protect, and it held the most locked rooms. In much of India, for most of history, a daughter left her parents' house at marriage with what her family chose to give her. She could rarely divorce, rarely inherit the land, and rarely leave her husband's house with her children.

There was no single old law to open. Before 1956 Hindu law varied by school, region and caste: the Mitākṣarā across most of India, the Dāyabhāga in Bengal, and in Malabar matrilineal systems under which, as the reformer Renuka Ray wrote in 1952, women held rights fuller than men's.²⁰

The opening was fought room by room, and not by one man. The All India Women's Conference had asked in 1934 for an inquiry into women's legal disabilities. Women members such as Hansa Mehta, G. Durgabai and Renuka Ray spoke for reform in the legislature.²¹ As Law Minister, Ambedkar steered the Hindu Code Bill, which would have given Hindu women rights to inherit property, to divorce and to choose in marriage. It was introduced on 9 April 1948 and stalled by fierce opposition inside and outside Parliament.²¹ On 27 September 1951 Ambedkar resigned from the Cabinet.

His resignation statement of 10 October 1951 called the Hindu Code "the greatest social reform measure ever undertaken by the legislature in this country." He warned that leaving untouched the inequality "between class and class, between sex and sex" while passing economic laws would "make a farce of our Constitution and to build a palace on a dung heap."²²

The Bill came back in pieces

The Bill returned in four parts: the Hindu Marriage Act of 1955, and the Hindu Succession Act, the Hindu Minority and Guardianship Act and the Hindu Adoptions and Maintenance Act of 1956.²¹ Women gained a right to divorce, a right to adopt and a share in their parents' separate property, each with limits. A married woman could not adopt in her own right until 2010, and she still needs her husband's consent.²³ In these Acts "Hindu" is a legal category, not only a faith. It includes Buddhists, Jains and Sikhs, and leaves out members of Scheduled Tribes unless the government directs otherwise.²⁴

One room stayed shut. Under the old Mitākṣarā law, followed across most of India, ancestral property was held in a coparcenary, a joint ownership that sons acquired at birth. The 1956 Act left daughters outside it. A son was born an owner of the family's ancestral land. A daughter was not. That room stayed locked until an amendment to the Hindu Succession Act, in force from 9 September 2005, made daughters coparceners by birth.

Even then the courts had to finish the job. Some judges held that a daughter's right applied only if her father was alive in 2005. On 11 August 2020, in Vineeta Sharma v. Rakesh Sharma, a three-judge bench of the Supreme Court ruled that the right arises by birth, whether or not the father was then alive.²⁵ In the words reported from the judgment, "the daughter shall remain a coparcener throughout life, irrespective of whether her father is alive or not."²⁵ More than half a century passed between Ambedkar's draft and the daughter's full share.

These are Hindu personal laws. Muslim, Christian, Parsi and tribal communities have their own laws and their own histories of opening, which this article does not try to tell.

Do daughters actually receive their shares?

Mostly, the evidence says, they do not, or not yet. Landesa and UN Women surveyed 1,400 rural women in Andhra Pradesh, Bihar and Madhya Pradesh in 2014, so their figures describe those places, not all of India. Only 13 per cent of the women whose parents owned land had inherited it or expected to. Sixty-one per cent said they had signed a no-objection certificate giving up their claim to family land.²⁶

A study by Bina Agarwal and colleagues, using village panel data from nine states for 2009–14, found that barely 16 per cent of women in rural landowning households owned land, and that women were "much more likely to acquire land as widows than as daughters."²⁷

The law opened the room. Many families have not walked their daughters into it. A daughter may sign a release in her brothers' favour because asking for her share would cost her the welcome of her parents' house. The law cannot see that signature for what it is. Another may hold land in her own name while her husband or brothers decide what is grown and who keeps the income. A title in law is not control in life.

Can a family be both care and cage?

Yes, and often at the same time. The family that looks after a grandmother may be the same family that will not let a daughter-in-law leave with her children. Calling one of these the family and the other an accident does not help. Both come from the same strength. People bound closely enough to care are bound closely enough to control.

The question is how to keep the first without the second. I think it can be put as a test for every body that stands between the person and the state, from a family to a gurdwara to a resident welfare association. A thick society must be strong enough to care for the person, and weak enough to let the person leave. In practice the test asks five things:

  • Can I enter it on equal terms?

  • Can I speak inside it and be heard?

  • Can I disagree without being punished?

  • Can I afford to leave it, and leave without losing my standing as a citizen, my property or my children?

  • Can I come back?

The economist Albert Hirschman called the two main responses to a failing organisation "exit" and "voice". Exit is leaving it. Voice is speaking up inside it.²⁸ A body that gives its members both has a chance of staying healthy. A body with voice but no exit becomes coercive, because those who cannot leave have little reason to be heard. A society with exit but no bodies becomes a crowd of strangers.

A right to leave is empty if there is nowhere to go. When one body is the only source of a roof, food, water or work, leaving is a word, not a choice. That is why income, housing and a public safety net matter to this test as much as law does. The last question, about coming back, is the one most charters forget. A daughter who leaves an abusive marriage should be able to return to her parents' house without shame.

Who is obliged to whom?

The warmth of obligation depends on who carries it. A child's obligation to an ageing parent is part of the wisdom. An obligation discharged entirely by the daughter-in-law, who did not choose it and cannot leave it, is the rust.

Passing the test means a woman's right to her own property, income and choices, and the right to leave a family, a community, an occupation or a faith. It also means treating the unpaid work of care as work. The Time Use Survey's minutes are a start. If women spend more than three times as long as men on household work, that work is part of the economy, and a developed India should count it, support it and stop simply assuming it.¹⁵

Is this anti-Hindu?

No. The Hindu family was opened largely by Hindus: by Ambedkar, who was born into it, by reformers, by the women who sued their brothers, and by a Parliament elected mostly by Hindu voters. A tradition that can amend its own family law in public, over seventy years, is showing the power of correction this series looks for. The test of the five questions applies to every community's institutions, and the next part applies it, just as strictly, to a Sikh one.

Part Twenty-Three — A Kitchen Taken Back

What is the langar, and why does it matter?

The langar is the free community kitchen attached to every gurdwara, the Sikh place of worship. It matters here because it is the clearest institutional answer on this soil to both halves of the problem. The hymn says the one who eats alone eats sin. The caste rule said you may not eat with certain people. The langar answers both, every day, with food.

The Sikh tradition traces the langar to Guru Nanak, the first Guru, in the sixteenth century. The third Guru, Amar Das, made eating in it a condition of meeting him, a rule given as pahile paṅgat pāchhe saṅgat, first the row, then the congregation.²⁹ Paṅgat means the line of people seated on the floor to eat. Whoever came to see the Guru, of whatever caste or rank, first sat down in the row and ate the same food from the same kitchen.

Sikh tradition holds that the Mughal emperor Akbar visited Guru Amar Das at Goindwal and sat in the row to eat before meeting him.²⁹ I have found no Mughal record of it, so I tell it as tradition.

A designed institution

The langar is not a feeling. It is a designed institution with three working parts. It has a rule, the shared row. It has a funding system, dasvandh, the practice of giving a tenth of one's earnings to the community's common fund, which the Sikh tradition traces to the Gurus who built Amritsar.³⁰ And it has a labour system, sevā, voluntary service, in which people who could afford to pay others cook, serve and wash plates themselves.

At the Harmandir Sahib in Amritsar, often called the Golden Temple, the kitchen's manager said in 2013 that it fed about 100,000 people on an average day.³¹ Its purpose was to break a social wall by making people perform the opposite of the wall, every day, in public.

Does the row reach the village?

Not always. The test applied to the Hindu family applies here too. The row in the langar is an ideal kept every day. Sikh society outside the row has kept caste. The sociologist Surinder Jodhka records that Dalit Sikh communities such as the Mazhabis and Ramdasias built gurdwaras of their own, a trend he found strongest in the Doaba region of Punjab. Fieldwork from the 1950s that he cites found Mazhabis drawing water from a separate well and seated in separate rows at village feasts.³²

In 2018 the president of the SGPC, the body that runs the historic gurdwaras, voiced concern at the rising number of caste-based gurdwaras. A newspaper report that year said most Punjab villages had four or five gurdwaras divided by caste and sect.³³

The langar's ideal is real, and so is the practice that falls short of it. One row in the kitchen can stand inside a village with five gurdwaras. The ideal is the tradition's own standard, and it is the measure by which its practice should be judged.

The capture

Institutions can be captured, and the langar was. By the early twentieth century many of the great gurdwaras had passed into the hands of hereditary custodians, the mahants. Some treated the shrines and their endowments as private property, and some had the backing of the colonial government. The kitchen still ran. The question was who owned it.

The answer came from a movement. On 15–16 November 1920 a gathering of Sikhs in front of the Akal Takht in Amritsar formed the Shiromani Gurdwara Parbandhak Committee, the SGPC, to take the gurdwaras back.³⁴ On 20 February 1921, at Nankana Sahib, the birthplace of Guru Nanak, the men of Mahant Narain Das killed a band of reformers who had come to take control of the shrine. The Sikh encyclopaedia entry I read puts the dead at about 200.³⁴

The movement went on, largely without violence, through years of arrests. On 9 July 1925 the Punjab Legislative Council passed the Sikh Gurdwaras Act, which came into force on 1 November 1925 and handed the historic gurdwaras to an elected Sikh body. Its first election was held on 18 June 1926.³⁴

The state did not simply arrive at the end. The same government had backed the mahants. It declared the SGPC an unlawful body in October 1923 and fired on volunteers at Jaito in February 1924, before it gave way.³⁴

What was corrected

The sequence is institutional correction in its full form. It runs from capture to protest, sacrifice, law and a new body chosen by election. What was corrected was control, not purpose. The mahants had captured the endowments. They had not rewritten the row. The movement took back the question of who keeps the kitchen.

A movement bled first, and the law came after. As at Mahad, the state stood on both sides. It first protected the keepers, then wrote down an opening that people had already paid for.

The new keepers and their own gate

The story does not end in perfection. An elected community body is wider than a hereditary custodian. It is also narrower than the republic. Its voters are not all Sikhs. They are Sikhs in Punjab, Himachal Pradesh and Chandigarh who enrol under the Act and meet its conditions, such as keeping unshorn hair.³⁵ Its gate has two disputes. The first is who may vote.

Sahajdhārī Sikhs are those who follow the faith but do not keep the full outward forms, such as uncut hair. The Gurdwaras Act made an exception that let them vote.³⁶ On 8 October 2003 the central government issued a notification removing their vote. On 20 December 2011 a full bench of the Punjab and Haryana High Court struck the notification down, holding that only the legislature could make such a change.³⁷ Parliament then did. The Sikh Gurdwaras (Amendment) Act of 2016 removed the sahajdhārī exception and made the change effective from 8 October 2003.³⁶

The second dispute is whether the electorate gets to vote at all. The general elections to the SGPC's house were last held in 2011, though they are meant to come every five years. Voter enrolment for a new round began in 2023 and was extended again and again.³⁸ A poll expected in June 2025 was not held.³⁹

At the end of June 2025 the post of Chief Commissioner for Gurdwara Elections fell vacant when its holder reached retirement age.⁴⁰ A retired civil servant writing on 27 October 2026 reported that the post was still empty and no general election had been called.⁴¹ I found no official notice of an election up to that date.

The five questions, applied to the kitchen

Applied to the langar itself, the test passes almost every question. Anyone may enter and sit in the row. No one is asked their caste, religion or rank at the kitchen door. Anyone may leave and come back.

Applied to the body that keeps the kitchen, the answer is mixed. Who may speak, by voting, has been narrowed by law. Whether the members can be heard at all depends on elections that have not been held for fifteen years. An elected body that does not hold elections is drifting toward the failure it was created to correct.

Applied to the villages around the kitchen, the first question fails wherever a gurdwara belongs to one caste. The Sikh ideal does not excuse that, any more than the Hindu guest rule excused the closed tank.

The correction of 1925 is real. Whether it is still being kept is a live question. Together with the Tamil village that rewrote its election rules, it makes two recorded revisions on this soil, from two traditions, about a thousand years apart. Two are not a pattern. They are enough to show that correction has been done here, by people who did not build what they corrected.

Ending — Open in Law, Green in the Tank

The Bombay Legislative Council resolved in August 1923 that public water was for everyone. The Bombay High Court dismissed the last appeal against the drinkers on 17 March 1937, nearly fourteen years later.² In between, the drinkers had been beaten, the water purified against them, and a court order had barred them from it while the case ran.

The Constitution then wrote the same rule for the whole country. In 2011–12, when surveyors asked more than 42,000 households whether anyone in the family practised untouchability, more than one in four said yes.¹¹

The Chavdar tank has been open in law since 1937. In April 2026, a year before the centenary of the drinking, a reporter found its water stagnant and thick with algae, with sewage seeping in from nearby buildings. A local contractor told the reporter the water was so bad that one could not even swim in it. On 20 March 2026 the Chief Minister of Maharashtra broke ground on a Rs 55 crore project to purify it.⁴²

Every 20 March and every 25 December, people still come from across Maharashtra to stand beside the water a crowd was once beaten for drinking.

Words to keep

  • coparcenary, joint ownership of ancestral property under Mitākṣarā Hindu law, which sons acquired at birth and daughters only from 2005.

  • dasvandh, the Sikh practice of giving a tenth of one's earnings to the community's common fund.

  • Dāyabhāga, the school of Hindu law followed in Bengal, which differed from the Mitākṣarā on family property.

  • dharma, duty, the right order that holds a person, household and society together.

  • dhamma-mahāmātra, an officer of right conduct, appointed by Aśoka to see that his public standards were kept.

  • kevalādī, one who eats alone, in Ṛgveda 10.117.6 the person who will not share.

  • langar, the free community kitchen of every gurdwara, where all sit and eat together.

  • mahant, a hereditary custodian of a shrine, in Punjab the men who controlled many gurdwaras before 1925.

  • Manusmṛti, an ancient Sanskrit law book attributed to Manu, quoted both for its duties of hospitality and for its rules of caste.

  • Mitākṣarā, a medieval commentary on Hindu law that became the basis of family property law across most of India.

  • nṛyajña, the offering to human beings, one of a householder's five daily duties, discharged by honouring the guest.

  • paṅgat, the row in which people sit on the floor to eat in the langar, all at the same level.

  • sahajdhārī, a Sikh who follows the faith but does not keep the full outward forms, such as uncut hair.

  • saṅgat, the congregation of the faithful.

  • satyāgraha, holding to truth, a campaign of non-violent public action, as at Mahad.

  • sevā, voluntary service, offered without payment.

  • subsidiarity, the rule that a task should be done by the smallest body able to do it, with larger bodies helping, not replacing.

Sources

  1. B. R. Ambedkar, "The Revolt of the Untouchables", in Dr. Babasaheb Ambedkar: Writings and Speeches, vol. 5 (Education Department, Government of Maharashtra), ch. 21, pp. 247–258 (municipal resolution of 5 January 1924, "the Untouchables for the first time drank the water", suit of 12 December 1927, injunction of 14 December 1927). Text reproduced at https://surajyengde.com/wp-content/uploads/2023/07/THE-REVOLT-OF-THE-UNTOUCHABLES.pdf

  2. Bombay High Court, Narhari Damodar Vaidya v. Bhimrao Ramji Ambedkar, Second Appeal No. 462 of 1933, judgment of 17 March 1937 (Broomfield and N. J. Wadia, JJ.), AIR 1938 Bom 146. https://www.courtkutchehry.com/Judgement/Search/t/455024-narhari-damodar-vaidya-appellant-hash

  3. Supreme Court of India, Dr. B. R. Ambedkar gallery, "Mahad Satyagraha" (Bole resolution of 4 August 1923, government direction of 11 September 1923, the conference of 19–20 March 1927, the purification, the speech of 25 December 1927). https://main.sci.gov.in/AMB/mahadsatyagraha.php

  4. Sangharakshita, Ambedkar and Buddhism (Motilal Banarsidass, 2006), pp. 53–55, as cited in "Mahad Satyagraha", Wikipedia (the 108 pots). https://en.wikipedia.org/wiki/Mahad_Satyagraha

  5. Siddharthya Swapan Roy, "The Lake of Liberation", Outlook (8 April 2016) (108 pots, cow dung, cow urine, milk, ghee and curds, no source given). https://www.outlookindia.com/national/the-lake-of-liberation-news-296954

  6. Uttaramērūr inscription of Parāntaka I, regnal year 14 (920/921 CE), Vaikuṇṭha Perumāḷ temple, ed. Emmanuel Francis (DHARMA, 2023), after V. Venkayya (1908). https://dharmalekha.info/texts/INSUttiramerur00002 . See also K. A. Nilakanta Sastri, Studies in Cōḷa History and Administration (University of Madras, 1932).

  7. Aśoka, Rock Edicts I, II and V, in S. Dhammika, trans., The Edicts of King Asoka (Buddhist Publication Society, Wheel 386/387, 1993). https://www.accesstoinsight.org/lib/authors/dhammika/wheel386.html

  8. Pius XI, Quadragesimo Anno (15 May 1931), paragraph 79. https://www.vatican.va/content/pius-xi/en/encyclicals/documents/hf_p-xi_enc_19310515_quadragesimo-anno.html

  9. Manusmṛti 3.70, Sanskrit text and translation. https://enjoylearningsanskrit.com/scriptures/manu-smriti/chapter-3/verse-70

  10. Ṛgveda 10.117.5–6. Sanskrit as given in the edition accompanying H. H. Wilson's translation, https://www.wisdomlib.org/hinduism/book/rig-veda-english-translation/d/doc839963.html and https://www.wisdomlib.org/hinduism/book/rig-veda-english-translation/d/doc839964.html ; English after R. T. H. Griffith, The Hymns of the Rigveda (1896), https://www.sacred-texts.com/hin/rigveda/rv10117.htm

  11. National Council of Applied Economic Research, "Biggest caste survey: One in four Indians admit to practising untouchability" (29 November 2014), reporting India Human Development Survey II (2011–12); see Amit Thorat and Omkar Joshi, IHDS working paper (2015). https://www.ncaer.org/news/biggest-caste-survey-one-in-four-indians-admit-to-practising-untouchability

  12. International Institute for Population Sciences et al., Longitudinal Ageing Study in India (LASI) Wave 1, 2017–18: India Report and fact sheet on the elderly (IIPS, Mumbai, 2020), "Family and Social Network" indicators for persons aged 60+. https://www.iipsindia.ac.in/sites/default/files/other_files/LASI_W1_Elderly_India.pdf

  13. Social Security and Medicare Boards of Trustees, A Summary of the 2026 Annual Reports (US Social Security Administration, 2026). https://www.ssa.gov/oact/TRSUM/index.html

  14. The Maintenance and Welfare of Parents and Senior Citizens Act, 2007 (Act 56 of 2007), sections 4, 7, 9 and 23. https://indiacode.nic.in/bitstream/123456789/19405/1/maintenance_and_welfare_of_parents_and_senior_citizens_act%2c2007.pdf

  15. Ministry of Statistics and Programme Implementation, National Statistics Office, Time Use Survey (January–December 2024), press release of 25 February 2025. https://mospi.gov.in/sites/default/files/press_release/PR_TUS_25022025.pdf

  16. Constitution of India, Article 15(2) and Article 17. https://indiankanoon.org/doc/609295/ and https://www.constitutionofindia.net/articles/article-17-abolition-of-untouchability/

  17. The Protection of Civil Rights Act, 1955 (Act 22 of 1955, originally the Untouchability (Offences) Act; renamed by Act 106 of 1976 with effect from 19 November 1976), section 4. https://www.indiacode.nic.in/bitstream/123456789/1544/1/A1955-22Eng.pdf

  18. The Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989 (Act 33 of 1989, 11 September 1989). https://www.indiacode.nic.in/handle/123456789/1920

  19. National Crime Records Bureau, Prison Statistics India 2022, as reported by LiveLaw. https://livelaw.in/news-updates/ncrb-releases-prison-statistics-india-report-2022-243973

  20. Renuka Ray, "The Background of the Hindu Code Bill", Pacific Affairs 25, no. 3 (September 1952), pp. 268–277. https://www.jstor.org/stable/2752804

  21. Chitra Sinha, "Rhetoric, Reason and Representation: Four Narratives in the Hindu Code Bill Discourse" (Centre for Gender Research, Uppsala University), drawing on her Debating Patriarchy: The Hindu Code Bill Controversy in India, 1941–1956 (Oxford University Press, 2012) (AIWC resolution of 1934, speeches of Hansa Mehta, G. Durgabai, Sucheta Kripalani and Renuka Ray, introduction of 9 April 1948, the four Acts of 1955–56). https://uu.diva-portal.org/smash/get/diva2:874167/FULLTEXT02.pdf

  22. B. R. Ambedkar, statement on resignation from the Union Cabinet, 10 October 1951, in Dr. Babasaheb Ambedkar: Writings and Speeches, vol. 14, part 2 (Government of Maharashtra), text as reproduced at Modern Rationalist. https://modernrationalist.com/dr-ambedkars-resignation-speech

  23. The Personal Laws (Amendment) Act, 2010 (Act 30 of 2010, 31 August 2010), substituting section 8 of the Hindu Adoptions and Maintenance Act, 1956. https://prsindia.org/files/bills_acts/acts_parliament/2010/the-personal-laws-(amendment)-act,-2010.pdf

  24. The Hindu Succession Act, 1956 (Act 30 of 1956), section 2 (application). https://indiankanoon.org/doc/685111/

  25. Supreme Court of India, Vineeta Sharma v. Rakesh Sharma, judgment of 11 August 2020 (Arun Mishra, S. Abdul Nazeer and M. R. Shah, JJ.), (2020) 9 SCC 1, as reported by LiveLaw. https://livelaw.in/top-stories/hindu-succession-amendment-2005-has-retrospective-effect-161260

  26. Landesa and UN Women, survey of 1,400 rural women in Andhra Pradesh, Bihar and Madhya Pradesh, press release (2014) (13 per cent, 61 per cent no-objection certificates). https://landesa.org/press-and-media/march-2-2014

  27. Bina Agarwal, Pervesh Anthwal and Malvika Mahesh, "Which women own land in India? Between divergent data sets, measures and laws", GDI Working Paper 2020-043 (University of Manchester, 2020). https://hummedia.manchester.ac.uk/institutes/gdi/publications/workingpapers/GDI/gdi-working-paper-202043-agarwal-anthwal-mahesh.pdf

  28. Albert O. Hirschman, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States (Harvard University Press, 1970).

  29. The Encyclopaedia of Sikhism, ed. Harbans Singh (Punjabi University, Patiala), "Guru kā Laṅgar" (origin with Guru Nanak, pahile paṅgat pāchhe saṅgat under Guru Amar Das, the tradition of Akbar's visit to Goindvāl). https://eos.learnpunjabi.org/GURU%20KA%20LANGAR%20(LIT.,%20LANGAR%20OR%20REFECTORY%20OF%20THE%20GURU).html

  30. The Encyclopaedia of Sikhism, ed. Harbans Singh (Punjabi University, Patiala), "Dasvandh". https://eos.learnpunjabi.org/DASVANDH.html

  31. Al Jazeera, "In Pictures: Kitchen that feeds 100,000 daily" (17 November 2013). https://www.aljazeera.com/indepth/inpictures/2013/11/pictures-kitchen-feeds-100000-daily-20131117124238293396.html

  32. Surinder S. Jodhka, "Sikh Religion and Contentions around Caste", Religions 15, no. 10 (2024), 1219, drawing on his "Caste and Untouchability in Rural Punjab", Economic and Political Weekly 37 (2002), pp. 1813–23. https://www.mdpi.com/2077-1444/15/10/1219

  33. G. S. Paul, "SGPC's 1 village, 1 gurdwara concept adopted by Dhurkot", The Tribune (15 March 2018). https://www.tribuneindia.com/news/archive/punjab/sgpc-s-1-village-1-gurdwara-concept-adopted-by-dhurkot-557841

  34. The Encyclopaedia of Sikhism, ed. Harbans Singh (Punjabi University, Patiala), "Shiromani Gurdwara Parbandhak Committee", entry by Ajit Singh Sarhadi (formation in 1920, Nankana Sahib, government support for the mahants, the ban of 12 October 1923, the firing at Jaito on 21 February 1924, the Act of 1925). https://eos.learnpunjabi.org/SHIROMANI%20GURDWARA%20PARBANDHAK%20COMMITTEE.html

  35. Chitleen K. Sethi, "How SGPC, the mini-parliament of Sikhs, conducts elections", ThePrint (30 November 2024). https://theprint.in/statedraft/how-sgpc-the-mini-parliament-of-sikhs-conducts-elections-its-inextricable-link-to-punjab-politics/2380569

  36. PRS Legislative Research, "The Sikh Gurdwaras (Amendment) Bill, 2016". https://prsindia.org/billtrack/the-sikh-gurdwaras-amendment-bill-2016

  37. Punjab and Haryana High Court, Full Bench (Surya Kant, M. M. S. Bedi and M. Jeyapaul, JJ.), Sehajdhari Sikh Federation v. Union of India, CWP No. 17771 of 2003, judgment of 20 December 2011, https://www.courtkutchehry.com/Judgement/Search/t/339915-sehajdhari-sikh-federation-vs-union . Outcome as reported in "HC restores voting rights of Sehajdhari Sikhs in SGPC polls", SikhNet (20 December 2011), https://sikhnet.com/node/32435

  38. "SGPC voter registration deadline extended for fourth time amid declining numbers", PTC News (31 July 2024). https://www.ptcnews.tv/punjab-2/sgpc-voter-registration-deadline-extended-for-fourth-time-amid-declining-numbers-4395062

  39. "Gurdwara panel likely to hold SGPC polls in June", Hindustan Times, Amritsar edition (31 January 2025). https://www.magzter.com/hi/stories/newspaper/Hindustan-Times-Amritsar/GURDWARA-PANEL-LIKELY-TO-HOLD-SGPC-POLLS-IN-JUNE

  40. K. B. S. Sidhu, "Chief Commissioner Gurdwara Elections Justice SS Saron Denied Extension: SGPC Polls Delayed", The KBS Chronicle (10 August 2025). https://kbssidhu.substack.com/p/chief-commissioner-gurdwara-elections

  41. K. B. S. Sidhu, post of 27 October 2026 on the SGPC office-bearers' election and the pending general election, The KBS Chronicle (post vacant since 30 June 2025, no general election since 2011). https://kbssidhu.substack.com/p/resolved-unanimously-and-what-thereafter

  42. Kasturi Walimbe, "Once a site of Dalit resistance, neglected Chavdar Tale to get Rs 55 crore facelift in centenary year", ThePrint (14 April 2026). https://theprint.in/the-fineprint/once-a-site-of-dalit-resistance-neglected-chavdar-tale-to-get-rs-55-crore-facelift-in-centenary-year/2904296/

How this was written

  • I read the court's 1937 judgment, Ambedkar's own account of Mahad in the Writings and Speeches and his 1951 resignation statement, the Constitution and the untouchability laws, the Hindu Succession Act, the 2010 adoption amendment and the senior citizens' maintenance Act, the Ṛgveda hymn and the Manusmṛti verse in Sanskrit with translations, Aśoka's edicts in translation, Renuka Ray's 1952 article, Surinder Jodhka's 2024 paper, and the LASI, Time Use, IHDS and US Trustees reports, as cited.

  • I could not open Anand Teltumbde's book on Mahad or Dhananjay Keer's biography in full. The 108 pots of the purification rest on later secondary accounts that agree with each other but cite no record from 1927. The fact of the ritual rests on the Supreme Court's Ambedkar gallery.

  • Some sources remain secondary because I could not open the primary: the NCRB undertrial figure and the Vineeta Sharma words (through LiveLaw), the resignation statement (through a reproduction), and the operative part of the 2011 sahajdhārī judgment, whose full text I could open only in part.

  • The state of the SGPC election is as reported up to 27 October 2026 by a commentator and by earlier news reports. I found no official notice either way.

  • The idea of a society thicker than its state, the three hands, the rule of subsidiarity for provision and constitutional supremacy for rights, and the five questions of the exit-and-voice test are my own constructions, drawing on Pius XI and Albert Hirschman. That the old bodies survived mostly by changing is my reading.

  • The conclusion would change if good evidence showed that daughters now receive and control their inheritance in most families, or that untouchability has largely vanished in practice since 2012, or that India's family care rests far less on women's unpaid time than the surveys say.

  • No imagined person appears in this article.

Recent Posts

See All

Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page