These are not abstract threats. They have dates and names and specific consequences — films that nearly vanished, traditions carried by a single pair of hands, manuscripts that survived only because someone refused, against all practical advice, to let them go.

This feature is about that refusal. About the decoy camera unit sent out to distract protesters while the real film was being made elsewhere. About the woman who stood up to perform when standing was forbidden. About the art painted on paper boxes by women with no institutional support, no grant money, no recognition — only the knowledge that if they stopped, it would be over. About the stories that crossed a border in the folds of a garment, or didn't cross at all and were lost, or crossed and were buried in family silence for fifty years before someone finally went looking.

The archive is not a building. It is a decision — made repeatedly, by specific people, under specific pressure, to keep going. What follows is an account of some of those decisions, and what they saved.

A second crew, fully equipped, unloaded camera pointed at nothing in particular, drawing the protesters away. While they followed the empty camera, the real film was being made elsewhere. This, before a single frame had been seen by a censor.

Garam Hawa — Scorching Winds — is based on an unpublished short story by Ismat Chughtai, the Urdu writer who had already spent a lifetime being told that the things she wanted to write about were not things that ought to be written about. Chughtai knew what suppression looked like. She had faced an obscenity trial in Lahore in 1944 for Lihaaf, a story whose crime was depicting, however obliquely, female desire. She was acquitted. She kept writing. The story she eventually gave Sathyu — never published, passed over to him to make into something — was about Salim Mirza, an Agra shoemaker, watching his world contract. His family leaving for Pakistan. His business crumbling. His city, the city of his ancestors, becoming quietly hostile. His refusal, held as long as it can be held, to leave.

The censor board, upon seeing the completed film, banned it. Instigation to communal dissension, they said — a phrase so breathtakingly inverted that it takes a moment to fully absorb. A film about a Muslim man's love for India was ruled to be a threat to India. It was held for eleven months. A Muslim Member of Parliament objected. LK Advani, then writing on films, made what the record diplomatically calls damaging remarks — without, it should be noted, having seen the film. Bal Thackeray asked to screen it. Sathyu, who knew Indira Gandhi, went to her directly. She watched it. She admired it. The certificate came through.

By then, the distributors had walked.

The film eventually released, found its audience, won the National Award for Best Film on National Integration — the same film that had been banned for threatening national integration — and entered the conversation where it has remained ever since. Balraj Sahni's performance as Salim Mirza is one of the great pieces of screen acting in Indian cinema. Quiet, immovable, slowly devastated. He finished dubbing the film and died shortly after. Garam Hawa was his last.

But here is the thing that the awards and the eventual recognition tend to obscure: the film almost didn't exist at all. Not because it was a difficult film to make — though it was — but because at every stage, someone decided that this particular story did not need to be told. The funders were reluctant. The locations were hostile. The censors said no. The distributors disappeared. What survived was not the result of a smooth institutional process. It was the result of Sathyu not stopping. Of Chughtai handing over a story she had chosen not to publish. Of Kaifi Azmi writing a qawwali — Maula Salim Chishti — that became, depending on who you ask, the finest in Indian film history. Of Balraj Sahni giving the performance of his life to a role that very nearly had no screen to appear on.

Ritwik Ghatak's story runs in a different register — quieter in its losses, more scattered, more heartbreaking for it. Where Sathyu's film fought its way into the world and won, Ghatak's body of work exists partly as a record of what didn't make it. His first feature, Nagarik, completed in 1952, was never released in his lifetime — the negative sat deteriorating on a laboratory shelf for twenty-five years before it was discovered, salvaged from near-decay, and finally shown in 1977, a year after Ghatak died. Three of his films — Kato Ajanare, Bagalar Bangadarshan, Ranger Ghulam — were abandoned mid-production, their reels acquired decades later by the National Film Archive of India after years of sitting with the West Bengal government's Department of Information and Cultural Affairs. The ownership of the negatives of Ajantrik and Bari Theke Paliye was disputed for so long that they could not be screened at a retrospective held in his honour. Subarnarekha was restored from a 35mm master positive, its colour grading supervised frame by frame, as recently as 2025 — more than fifty years after it was made.

Still from Ritwik Ghatak's Nagarik

Ghatak's films were not suppressed by censors. They were lost to indifference, to financial collapse, to the particular cruelty of a country that could not find the administrative will to care for what it had. A generation of Bengali cinema — one of the most formally adventurous bodies of work produced anywhere in the world in the twentieth century — came within very ordinary bureaucratic distance of disappearing entirely.

What we have now, we have because people refused to let it go. Ghatak's son Ritaban, working through the Ritwik Memorial Trust, restored the unfinished films. The World Cinema Foundation put Titas Ekti Nadir Naam back on screen. The NFAI, on what it describes as an acquisition spree, has been piecing together what remains of a cinematic inheritance that was never properly kept.

The question that sits underneath all of this — under the decoy camera unit in Agra, under the deteriorating negative on the laboratory shelf, under the eleven months Garam Hawa spent in censor limbo — is a simple one. What does it mean that these films survived despite everything? Not because of the state. Not because of the industry. But because individual people, with no particular incentive beyond the conviction that the work mattered, kept going.

Salim Mirza, at the end of Garam Hawa, joins a march. He has spent the entire film trying to stay still, to hold on, to remain in the house and the city and the life that are his. And then, finally, he moves — not to Pakistan, not away from India, but forward. Into whatever India is becoming. It is an ambiguous ending, as all honest endings are.

This is how Partition survived in Indian families. Not as history. As objects. As weight.

What Malhotra did, beginning in 2013 with what was initially her MFA thesis project and expanding into a decade of research across India, Pakistan and the diaspora, was to find the objects and let them speak. A yardstick once used to measure fabric. A pair of pearl necklaces stored in a bank locker whose key has been lost. A pocketknife. Kitchen utensils. Photographs so ordinary they are extraordinary — proof that someone had a kitchen, a table, a life in a place that the border then made unreachable. These things, she argued, were not decoration. They were archive. In countries where oral culture outweighs the official record, the object you carried across is sometimes the only evidence that you belonged somewhere.

The first surprise of Malhotra's work is that it needed to exist at all. Partition — the largest forced migration in human history, fifteen million people displaced, somewhere between one and two million dead — had gone largely undocumented in terms of personal testimony for half a century. Not because there were no survivors. Because they didn't speak. What Malhotra and those working alongside her — the Citizens Archive of Pakistan, the 1947 Partition Archive — discovered when they finally went looking was a specific, particular kind of silence. A silence born not only of suffering but also of shame. Shame arising from complicity. From intimate betrayal. From the things that were done and witnessed and never named again.

Manto named them. That was his problem.

He had been banned and censored through much of his career — for obscenity, for indecency, for writing about things that polite literature was not supposed to see. After Partition, he found he could not get his work published in India at all, not because of censorship exactly but because of something more diffuse and more damaging — a cultural unwillingness to look. He moved to Lahore. He wrote Toba Tek Singh in 1955, a story about a man in a lunatic asylum who, when told he must be transferred to India because Partition has happened, simply refuses — and ends up dying in the strip of land between the two countries, belonging to neither. The lunatics, Manto suggested, understood what the sane had decided not to. He died the same year at forty-two, alcoholism finishing what the literary establishment had begun.

Bhisham Sahni — whose brother Balraj, it is worth noting, gave the defining performance of his life in Garam Hawa — wrote Tamas in 1974, twenty-seven years after the events it described. Not because he needed that long to process them. Because that was roughly how long it took for the conditions to exist for such a book to be received. Tamas is a brutal, unflinching account of communal violence in Rawalpindi in 1947, drawn from Sahni's own experience as a Congress worker doing relief work during the riots. It won the Sahitya Akademi Award. It was adapted into a television film in 1988 that was challenged in court and very nearly kept off the air — instigation to communal violence, the familiar charge, twenty years after the book's publication, forty years after the events themselves.

The literature of Partition has always been in a negotiation with those who would prefer it not exist. What Malhotra's second book, In the Language of Remembering, revealed was that this negotiation happened inside families too — not as censorship but as love. Grandparents who refused to pass their trauma down. Parents who answered questions with silence, with deflection, with the particular Indian domestic art of changing the subject. What Malhotra found, interviewing descendant generations across India, Pakistan, and Bangladesh, was that the silence had its own inheritance. Children who grew up knowing something enormous had happened and had been told nothing. Grandchildren who felt, somehow, the weight of a grief they couldn't name or locate. Memory suppressed at the source doesn't disappear — it travels, altered, into the bodies and the emotional architecture of people who weren't there.

This is the particular cruelty of what Partition did to Indian storytelling. It created an event so large that writing about it felt presumptuous, and so painful that not writing about it felt like relief — and then punished both responses. The writers who looked directly at it were banned, censored, unable to find publishers, dragged into court. The families who looked away found the silence had settled into their grandchildren like sediment. There was no clean exit.

What survived is what was carried — against institutional resistance, against the body's own instinct to forget, against the border itself. A maang-tikka tied into the folds of a garment. A story written in a Lahore asylum. A family finally speaking, seventy years later, before it was too late.

This version had been circulating orally in these forests and plains long before anyone thought to write it down. It did not require official endorsement. It simply kept being told.

Teejan Bai heard it for the first time from her maternal grandfather, Brijlal Pardhi, in Ganiyari village, fourteen kilometres north of Bhilai. She was a child from a Pardhi Scheduled Tribe family — her parents made mats and brooms, their poverty ordinary and unexceptional. What was not ordinary was what happened when she heard the stories. She memorised them. Not as exercise but as instinct, the way some people absorb music before they can explain what music is. She trained informally under a local performer named Umed Singh Deshmukh. And at thirteen, she walked to the neighbouring village of Chandrakhuri and performed in public for the first time, for ten rupees.

She performed standing. This was the transgression.

Pandvani — the word means voice of the Pandavas, the oral singing tradition through which the Mahabharata had survived in Chhattisgarh for centuries — had two styles. Vedamati, in which the performer sits and narrates. Kapalik, in which the performer stands, enacts, inhabits the characters, becomes them — switches between Bhima and Draupadi and Krishna within a single performance, one body, one tambura, the story carried entirely by voice and gesture. Kapalik was the style with power, with full embodiment, with everything at stake. Kapalik was for men. Women were permitted Vedamati — the sitting style, the quieter style, the style that kept you at a certain manageable distance from the story you were telling.

Teejan Bai stood up.

Her community expelled her for it. She was twelve when she married, in the custom of her people. She was thirteen when her singing drove her out of the community entirely. She had to borrow utensils to cook her daily food. The people in her village called her characterless — a word that in the context of rural India in the 1960s carried a specific and devastating weight, meaning a woman who had chosen something other than invisibility. She kept performing. Her second husband beat her and forbade her from going on stage. She kept performing. Girls who appeared on stage were ridiculed and abused, she said in an interview decades later. She had faced enough of both for many lifetimes.

What Teejan Bai was doing, though she may not have framed it this way, was something older and more radical than defiance. She was insisting that the Mahabharata — the story that Chhattisgarh's tribal communities had carried in oral form across centuries, the version that centred Bhima and not Arjuna, the version that was never written down because it didn't need to be — belonged to her as much as it belonged to anyone. That the tradition of Pandvani was not a male tradition that women had been thoughtfully excluded from. It was a human tradition that women had been forcibly kept from. Those are not the same thing.

There is a strand of Indian cultural history — not often told — in which the oral traditions held by marginalised communities diverge significantly from the received, Brahminical, textual versions of the same stories. The Pandvani Mahabharata is one instance. The Gond communities who shaped it did not inherit the Sanskrit epic through scholarship or patronage. They took what they needed from it — Bhima, the strength, the earth — and made it theirs. Teejan Bai inherited this already-transformed tradition and transformed it further — brought the Kapalik style to it, brought her own guttural, unmistakable voice to it, brought herself to it in a way that the tradition had specifically tried to prevent.

The wider world noticed eventually. Habib Tanvir — the great Urdu playwright, himself from Chhattisgarh, one of the finest theatrical minds of twentieth century India — saw her perform in the 1970s and understood immediately what he was seeing. He arranged for her to perform for Prime Minister Indira Gandhi. That single performance changed the geography of her career. Invitations came from across the country, then from abroad — France, Switzerland, Germany, Bangladesh, stages she could not have imagined from the borrowed kitchen of her expulsion. She appeared in Shyam Benegal's Doordarshan series Bharat Ek Khoj in 1988, telling the Mahabharata to millions. The Padma Shri came in 1987. The Padma Bhushan in 2003. The Padma Vibhushan — India's second highest civilian honour — in 2019.

Teejan Bai performing in the Kapalik style

She performed in a wheelchair by then, her body finally giving up what her will had refused to. Offstage she would sit. Onstage, she stood.

The Chhattisgarhi Mahabharata — the one that chose Bhima, the one that belonged to communities the official tradition had no particular interest in, the one that a thirteen-year-old girl from a mat-weaving family heard and refused to stop telling — is still being performed today, by women Teejan Bai trained herself. Usha Barle, Ritu Verma, others. The tradition she was expelled for perpetuating is now taught formally. A girl who was called characterless for standing up to sing is now the name the art form is known by.

She said once that according to Kapalik's origin, the style might be as old as the Mahabharata itself — because very few people could read in those times and that is how they passed their stories down, generation after generation.

The journey involved trials, divine encounters, heaven itself. She reached Lord Shiva. She wore her ghoongat before him — the veil, the gesture of modesty — and made her case. She came back with her husband alive.

This is the story that the women of Bhagalpur's Anga region painted, for centuries, on paper boxes made of jute sticks. Panel by panel, sequential, narrative — the boat, the body, the serpents coiling through borders of flowers, Bihula's journey rendered in bold lines and natural dyes, ochre and indigo and the deep green of forest vegetation. It is, by most accounts, India's oldest surviving sequential narrative art form. Not decorative. Not abstract. It tells a story, panel by panel, the way a manuscript does — except the manuscript is a box, and the box is an offering, and the offering is made once a year during Bishahari Puja to honour the snake goddess and ward off the very serpents whose image covers every surface.

For centuries this was how the story survived. Painted, offered, renewed each year. No institution required. No patronage necessary. Just the women of two castes — the Kumbhakar and the Malakar — who made the pots and the boxes and painted the story on them, season after season, because the story needed to be told and they were the ones who knew how.

Then it nearly stopped.

The decline was not dramatic. It was ordinary — the particular ordinariness of things that disappear not because anyone decides to destroy them but because the conditions that sustained them quietly dissolve. Urbanisation. The dominance of more commercially visible art forms — Madhubani, to the west, had found an international market; Manjusha, confined to its ritual function and its specific geography, had not. Fewer families passing the knowledge down. The story still told at Bishahari Puja, but the painted boxes becoming rarer, the hands that knew how to make them fewer each decade. By the time anyone looked up and assessed what was happening, there were only two families left who fully practised the art.

In 1940, a British ICS officer named W.G. Archer had encountered Manjusha during his posting in Bihar, recognised immediately that he was looking at something extraordinary, collected pieces and organised an exhibition at the India Office Library in London. The art received international attention — briefly, under colonial conditions that made it impossible for the artisans themselves to benefit, and then lapsed back into its Bhagalpur obscurity while Archer's collection sat in London becoming the Archer Collection.

The first serious attempt at revival came in 1984, when the Bihar government's Jansampoorna Vibhag sent teams to the villages of Bhagalpur with slideshows — actual slideshows, images of the art shown back to the communities that had made it, in the hope that seeing it reflected this way might rekindle something. It was a strange and touching intervention, slightly absurd, completely sincere. And it worked, partly. The woman who stepped forward from those sessions and gave the revival its first real momentum was Chakravarty Devi — from the Malakar caste, one of the two families, the keeper of one of the last intact lines of knowledge about how Manjusha was actually made and painted. She worked tirelessly — a phrase that in this context means something specific: she kept going when there was no market, no recognition, no institutional structure to support her, just the conviction that the art could not be allowed to end with her.

It almost did. Near the end of her life, Chakravarty Devi said: Lagta hai meri maut ke saath Manjusha kala bhi khatm ho jaayegi. It seems Manjusha will also die with my death.

It didn't. Nirmala Devi — another woman, no institutional support, the same conviction — was working in parallel, painting Manjusha on plates and utilitarian objects to sell, selling her jewellery during the lean years when her family's finances collapsed, keeping the practice alive through sheer domestic stubbornness. Her son, Manoj Pandit, grew up watching this. He earned a fine arts degree from Chandigarh, went to Delhi, tried Mumbai, considered acting — and then came home to Bhagalpur and started painting Manjusha on the mud walls of his house. He is now known as Manjusha Kala Guru. In 2014 the Ministry of Culture gave him that title formally.

In 2021, Manjusha art received a Geographical Indication tag — legal recognition, finally, that this art belongs to Bhagalpur and can only be called Manjusha if it comes from there. More than a thousand rural women have been trained in the tradition since the revival gathered momentum. The art now appears on tussar silk, on canvas, on sarees and dupattas and cushions. It has e-commerce listings. It has a future that Chakravarty Devi, in her final years, could not see.

What she could see was what she had. A story painted on paper boxes, made by women for a festival honouring a goddess, about a woman who refused to accept a death and journeyed to the heavens to reverse it.

The earliest known incident of a film reel catching fire was Dadasaheb Phalke's Raja Harishchandra, made in 1913. A few years after the film was made, while it was being transported by bullock cart between cinema tents, the friction and heat caused it to combust. Phalke re-shot the entire film — shot for shot, from memory — to produce the version that exists today, incomplete, a reconstruction of a reconstruction. He was a meticulous man. He understood that if he did not act, the film would simply be gone. Most people did not act. Most people did not think to.

The attitude of the Indian film industry to cinema has always been, as the Film Heritage Foundation puts it with careful understatement, purely commercial. Producers and distributors treated films as liabilities once their theatre runs were done. Nitrate reels — the film stock used until 1951, chemically similar to gun cotton, capable of spontaneous combustion in vaults, projectors, warehouses and studio floors — were expensive to store and dangerous to keep. When the industry shifted to safety film stock and a silver crisis drove up costs, original negatives were sold by the kilogram to recover the silver from the emulsion. Entire filmographies, decades of work, traded by weight at flea markets. The economics of the thing are almost beautiful in their indifference — no malice required, just the ordinary workings of an industry that had not yet been told it was sitting on top of history.

By 1950, before the National Film Archive of India existed, before anyone had formally decided that Indian films were worth keeping, seventy to eighty percent of everything made was already gone. Alam Ara — India's first talkie, the film whose premiere on 14 March 1931 was one of the most significant dates in Indian cultural history, the night the country heard dialogue and song from a screen for the first time — no print survives anywhere in the world. We know what the night was like from written accounts and photographs. We cannot watch it. We can only read about watching it.

The NFAI was established in 1964 in Pune, at the old Prabhat Studio complex — a drier climate, better for storage than the humid coastal cities that had destroyed so much already. It was, by then, fifty years too late for most of what had been made. The archive began acquiring, preserving, cataloguing — and on 8 January 2003, a fire in its oldest nitrate vaults destroyed 607 films in 5,097 reels in a single night. The Information and Broadcasting Ministry's immediate response was to underplay the disaster — a counterclaim was issued that little had been lost because there were copies of almost everything. This was, to put it plainly, not true. By 2019, a Comptroller and Auditor General audit of the NFAI's records found that 31,000 reels were reported lost or destroyed. The number is almost too large to hold as a single fact.

What survives of India's film heritage survives largely because of the stubbornness of individuals who decided, without official mandate or sufficient resources, that it could not be allowed to disappear. Shivendra Singh Dungarpur founded the Film Heritage Foundation in 2014 — the only non-governmental organisation in India working specifically in film preservation. His description of the work is precise: it is as much detective story as preservation. For the restoration of Sholay, there was no original camera negative to work from. For Shyam Benegal's Bhumika, a single print survives. Pakeezah is currently being restored — frame by painstaking frame. Do Bigha Zamin, Bimal Roy's 1953 masterpiece, was restored in 4K and premiered at Venice, then went to Criterion — which is how an Indian film made seventy years ago found its way to the collection where it will now be properly kept, possibly indefinitely.

The international dimension of this survival is one of the more quietly humiliating facts of Indian film history. A print of Rustom Dotiwala's 1919 silent film Bilwamangal was found not in India but at the Cinémathèque Française in Paris, which made a digital version and sent it home. Ghatak's Titas Ekti Nadir Naam was first properly released in the UK by the BFI, then restored by the World Cinema Foundation. Regional cinema has fared worse than Hindi cinema across every metric — Manipuri, Odia, Assamese, the films of the Northeast — because regional cinema had even less commercial clout, even less institutional protection, and was stored in conditions that guaranteed decay. The entire silent era of Madras cinema — 124 films and 38 documentaries — yielded exactly one surviving title: Marthanda Varma from 1931. One.

There is a particular kind of loss that is harder to grieve than simple destruction, and that is the loss of things you did not know you had. A film that burns in a fire is a tragedy with a date and a cause. A film that decays across thirty years in an uncatalogued godown in an unventilated warehouse is a quieter disappearance — slow, administrative, entirely preventable, discovered only when someone finally goes looking and finds nothing. India's archival catastrophe is mostly the second kind. Not dramatic fires (though there have been those too) but the accumulation of a thousand smaller decisions — to not allocate the budget, to not build the storage, to not catalogue the collection, to not think of the films as anything other than old product.

What the Film Heritage Foundation and the NFAI's current restoration programme represent is a belated but genuine reckoning with that accumulation. The National Film Heritage Mission, launched in 2016 by the Ministry of Information and Broadcasting, has been digitising, preserving, restoring. Ghatak's Subarnarekha received a 4K restoration, its colour grading supervised frame by frame. Regional films from Manipur, Odisha, Assam are being brought back. It is slow work, expensive work, work that outpaces the decay only some of the time.

Dungarpur describes the enterprise with the precision of someone who has spent years reckoning with what is already gone. Seventy percent of films made before 1950, he has said, are gone forever. He is trying to save the rest.

This is, in the end, what an archive is. Not a building. Not a database. Not a government programme, though all of these help. An archive is the decision, made by specific people at specific moments, that something is worth keeping. India made that decision late, made it inconsistently, and is still making it now — urgently, expensively, against a clock that has been running since 1913 when a bullock cart carried a reel of nitrate film through the summer heat and the friction was enough.

Twenty-nine films. Of 1,700.