Blog

In the spring of 2002, in the Hoshiarpur district of Punjab in northern India, a team from the state’s Vigilance Bureau dusted nineteen banknotes with a white chemical powder and handed them to a man named Kulwant Singh. The powder was phenolphthalein. It is invisible on skin until you wash your hands in a solution of sodium carbonate, at which point the water turns a bright pink that is very hard to argue with. The notes were five hundred rupees each. Nineteen of them came to nine thousand five hundred rupees, which at the exchange rate of that April was worth a little under two hundred US dollars.

The man the trap was set for was a Patwari named Jaswinder Singh. Everything that followed, a criminal trial, a conviction, a prison sentence, an appeal to a High Court, a further appeal to the Supreme Court of India, and twenty four years of case files, came out of that sum. He did not live to see the end of it. He died with his final appeal still waiting its turn, and the people who finished the case were not him.

It is worth saying straight away what a Patwari is, because the whole story sits on it. In the village revenue system that much of South Asia inherited and still runs, the Patwari is the person who holds the land records. He keeps the register of who owns which field, he measures and maps, and when ownership changes he records the change. That recording is called a mutation. It is a clerical act performed by one of the lowest paid officials in the chain, and it is also the single document that turns a family’s claim to a piece of ground into something a bank, a buyer or a court will recognise. If your name is not entered, in practice you do not own the land, whatever you believe.

In this case, according to the Supreme Court’s own account of the prosecution story, Kulwant Singh’s sister in law wanted her late husband’s land entered in the names of his legal heirs. An ordinary piece of paperwork, the kind that happens after every death in every village anywhere. The allegation was that the Patwari asked for ten thousand rupees to move it along, and that the figure was then negotiated down to nine thousand five hundred. Kulwant Singh went to the Vigilance Bureau instead of paying, and the Bureau did what anti corruption units across the subcontinent do.

The method is standardised and, read cold, almost theatrical. The complainant’s own money, or money provided for the purpose, is noted down by serial number so that it can later be identified beyond dispute. Each note is treated with phenolphthalein. A raiding party is assembled, and crucially one member of it is designated as the shadow witness. The shadow witness has exactly one job. He is supposed to stay close enough to the complainant to see and hear the whole transaction, so that when the case reaches a courtroom years later there is an independent human being who can say, under oath, that he watched the official ask for the money and watched him take it. Everything else in the procedure is corroboration. The shadow witness is the evidence.

On the day of the raid the notes were recovered from the Patwari’s shirt pocket. His hands tested positive. In the ordinary course that looks like a complete case, and in February 2004 a special judge in Hoshiarpur agreed. Jaswinder Singh was convicted under two provisions of the Prevention of Corruption Act and sentenced to two years of rigorous imprisonment and a fine of ten thousand rupees, which is to say a fine slightly larger than the bribe he was found to have taken. He appealed to the Punjab and Haryana High Court, which took until November 2013 to decide and then affirmed the conviction. Eleven years had passed. He took it to the Supreme Court of India, which granted him leave to appeal in September 2014.

Then the file went quiet, in the way enormous court systems go quiet. The Supreme Court of India receives tens of thousands of matters a year. A criminal appeal from a two year sentence over a small sum is nobody’s emergency. And somewhere in that waiting, Jaswinder Singh died. The judgment does not say when, and it does not say how old he was, and this piece is not going to guess at either. What the record shows is that in October 2018 the Court permitted his legal representatives to be substituted for him, which means he had died at some point before that. From then on the case was titled with a bracket after his name noting that the appellant was dead and that the appeal was being carried by his legal representatives.

That phrase is where this stops being a story about a bribe and becomes a story about a decision somebody made.

Here is the mechanism, because it is genuinely not widely known, and it is the hinge of everything. Under Indian criminal procedure, most appeals simply die when the appellant dies. The reasoning is obvious enough. The point of a criminal appeal is to relieve a living person of a punishment, and you cannot un imprison a dead man. So the rule, broadly stated, is that an appeal finally abates on the death of the appellant, and the file is closed.

But the rule has two openings in it, and they matter. An appeal against a sentence of fine does not abate, because a fine attaches to the estate, and an estate survives its owner. And there is an express exception written in for appeals against a conviction carrying death or imprisonment. Where the appellant dies while such an appeal is pending, any of his near relatives may apply to the appellate court for leave to continue it. The application has to come within thirty days of the death. If leave is granted, the appeal does not abate. It goes on exactly as before, with the dead man still named as the appellant and somebody else standing in his place.

Read that provision twice and you see what the law is quietly conceding. It is accepting that a criminal conviction is not only a punishment. It is also a public statement about a person, entered in a public record, that outlives the body of the person it is about. The relatives who continue an appeal are not trying to get anyone out of prison. There is no prison left to get out of. They are going after the statement.

And there is a detail in that provision that is easy to read past and hard to forget once you have noticed it. The application has to be made within thirty days of the death. Whoever did this had one month, starting from the day they buried him, to decide that a decades old conviction over a sum worth less than two hundred dollars was worth continuing to fight, find out that such a thing was even legally possible, instruct someone who knew how to ask, and file. Thirty days is the window the law allows for that, and it runs through the worst month a family has.

There was also, to be fair, a concrete stake beyond the name. The fine attached to his estate, and a fine does not die with the person it was imposed on, which is precisely why the law treats a fine appeal differently. But a fine of ten thousand rupees, by 2018, was a trivial amount of money in absolute terms and nothing like enough to justify eight years of Supreme Court litigation. Anyone doing this purely as a financial calculation would have paid it and walked away in an afternoon.

Somebody in Jaswinder Singh’s family filed that application. We do not know who. The judgment records that legal representatives were permitted to step into the shoes of the appellant, and it names nobody, and neither does any reporting of the case, and neither does the case headnote. The courts of India are unusually open compared with most jurisdictions, and the full text of the judgment is free for anyone in the world to read, but it does not name the family, and so this article will not either. That is a real hole in the middle of the story and the honest thing is to mark it rather than fill it. What his widow or his children or his brother or his sister thought, how long they argued about it, whether anyone told them it was a waste of money, whether they could afford it, none of that is in the public record. All that is in the public record is that the application was made, and that it was allowed, and that the case continued.

It then continued for eight more years.

Think about what those eight years consisted of, in practical terms, for a family in a district town. A Supreme Court matter has to be maintained. There is counsel to instruct and pay, there are listings that come and go, there are adjournments, there are years when nothing at all happens and the only thing required of you is that you do not give up and withdraw. The person at the centre of it is dead and cannot be helped. The sentence can no longer be served or not served. The only thing still available, the entire prize, is the removal of a line in a register.

In September 2026 the Supreme Court of India removed it.

What the Court actually held matters enormously here, and it needs to be stated precisely, because the comfortable version is not the true one. The Court did not find that the trap was fabricated. It did not find that the money was planted, or that the Vigilance Bureau behaved improperly, or that nothing happened in Hoshiarpur in April 2002. It made no positive finding of innocence at all. What it found was that the prosecution had not proved its case, and the reason is the shadow witness.

Onkar Singh was the shadow witness. In the witness box he accepted that the money had not been handed over in his presence, because at the relevant moment he had stepped out of the office. The official witness, Harnandan Singh, said the same thing. Neither man saw the demand. Neither man saw the money change hands. The only person who claimed to have witnessed the transaction was the complainant himself.

That left the prosecution resting on the recovery of the notes from the Patwari’s pocket and the chemical test on his hands, and on this the Supreme Court applied a principle settled by a larger bench of the same Court in a case called Neeraj Dutta. Stripped of the Latin, it comes to this. To convict a public servant of taking a bribe, the state has to prove two separate things, that he asked for the money and that he accepted it. Both. Proof that money ended up in his possession is not the same as proof that he demanded it, and the law refuses to treat it as such, because the alternative is that physical possession of a note becomes conviction by itself, regardless of how it got there.

The Court held that mere recovery of tainted money, standing alone, could not establish the charges. With no witness to the demand and no witness to the acceptance, the two things the state was required to prove were simply absent from the evidence, whatever else the evidence showed. Both the trial court’s judgment of 2004 and the High Court’s affirmation of 2013 were set aside, and the conviction and sentence went with them.

So the honest summary is this. A Supreme Court of India acquittal is final, with nothing further to appeal and nothing left interlocutory about it. The conviction is gone, entirely and permanently. Whether Jaswinder Singh asked a widow’s family for ten thousand rupees in 2002 is a question no court will ever now answer, and the Court was careful not to pretend otherwise. What was decided is that he should never have been convicted of it on the evidence the state brought, and that is what his family was after, and that is what they got.

The arithmetic of the thing is what stays with you. The sum at the centre of the case was worth under two hundred dollars at the time. It was, for a global reader, roughly a decent week of groceries. For that, a man spent the last years of his life as a convicted person with a two year sentence hanging over him and a fine larger than the money he was said to have taken. The case outlasted his working life and then it outlasted him. By the time it ended, the trap was twenty four years in the past and the conviction had stood for twenty two.

And none of it was ever going to be given back. That is the part worth sitting with. His family did not win him anything. They could not restore a single year, or a single rupee of what the case had cost, or the ordinary dignity of not being a man who had been convicted. There was no compensation at the end of this and nobody suggested there would be. The only available outcome was a sentence in a judgment saying the conviction could not be sustained, and the family went after it anyway, for eight years, with no deadline and no guarantee and nobody left to thank them for it.

People talk about clearing a name as though it were a figure of speech. In this case it was the literal and only objective, pursued at real cost by people whose names we do not know, on behalf of someone who could not be told it had worked.

Anant Nag did not hold a farewell press conference. He did not post anything. He did not give the interview where an actor explains that it is time to let the younger ones have a turn. What he did was simpler and much stranger. He kept asking producers to send him the script first, and he kept reading, and he kept saying no. More than a hundred scripts over three years, across five different languages, from people who very much wanted him. He liked none of them. And somewhere in the middle of all that reading, without ever announcing it, a fifty year career ended.

Then, three years into a retirement nobody had been told about, the Government of India gave him the Dadasaheb Phalke Award, the highest honour the country has for a life in cinema, and the hall stood up.

That is a peculiar sequence, and it is worth sitting with for a moment. Most long careers end with a long goodbye. There is a last film, a tribute reel, a round of interviews about legacy. Nag’s ended the way a reader puts down a book they are not enjoying. He has been blunt about why. Seeing what was on offer, he stopped. He was being sent work, he says, from both the Hindi industry and the Kannada industry, and his answer every time was the same: send the script, and if I like it, I will do it. The scripts came. He read them. None of them were good enough to be worth the years he had left.

He was in his late seventies by then. He is seventy eight now, born in 1948, which means he was being asked to spend what is left of a working life on films he already knew were mediocre. Put that way, the decision stops looking like retirement at all. It looks like the same judgement he had been exercising for five decades, applied one more time, and arriving at zero.

Almost nobody outside India has heard of him, which is the part that is genuinely hard to convey. He has acted in more than three hundred films. More than two hundred and fifty of them are in Kannada, the language of the southern Indian state of Karnataka, and the rest are spread across Hindi, Telugu, Tamil, Marathi, Malayalam and English. If you are from Karnataka, this is not a biography you need explained. There are lines of his that people quote at each other without needing to say where they are from, in the way that certain films become a shared vocabulary rather than a thing you watched. If you are not from Karnataka, he is a stranger with three hundred credits.

That gap is worth explaining rather than glossing over, because it is not a measure of how big he is. Karnataka has a population in the tens of millions, and Kannada cinema is one of several large film industries operating inside India at the same time, each with its own stars, its own history and its own audience that does not necessarily follow the others. An actor can be a household name to tens of millions of people, carry films for fifty years, and still never appear in the export version of Indian cinema that reaches the rest of the world, which is overwhelmingly the Hindi one. Nag is not a regional figure who never made it. He is a very large figure in a place the rest of the world does not get told about.

Here is the shortest way to understand what he actually is. In the 1970s, Indian cinema split into two worlds that mostly did not talk to each other. On one side was the commercial industry, the songs, the fights, the stars and the money. On the other was a movement of serious, austere, politically angry films, made cheaply, about caste, land and power, which came to be called parallel cinema. Its most important director was Shyam Benegal, who died in late 2024 at the age of ninety. Nag was in Benegal’s films. He was in Ankur, which arrived in 1974 and is still taught as the film that opened the door for everything that followed. He was in Nishant the year after, and Manthan the year after that, and then Bhumika, Kondura and Kalyug. That body of work alone would be a career worth a lifetime honour.

And at the same time, in the same years, he was a genuine commercial star in Kannada cinema, in exactly the kind of popular films that the parallel movement was set up as a rebuke to. That is the unusual thing about him, and it is the thing that is hard to do. Actors who become credible in the serious films tend to stay there, partly by temperament and partly because the commercial industry stops believing in them. Actors who become stars rarely get asked back by the directors who are trying to make something difficult. Nag did both, for decades, without either side deciding he belonged to the other. He kept working in big Kannada films well into his seventies, which means there are audiences who know him only as a beloved presence from films made after the parallel cinema years were already history, and who would be surprised to learn what he was doing in 1974.

The route into all of it is more improbable than the career. He grew up in rural Karnataka, in a stretch of coast running roughly four hundred kilometres from Karwar in the north to Kasaragod in the south, and he was raised in an ashram and a math. A math, sometimes written as mutt or matha, is a monastic institution, a religious house with a teacher at its centre, and for a child it means being brought up inside an austere and ordered devotional routine rather than in an ordinary family home. He was schooled in Kannada up to the eighth standard, which is about as far as a village education went. Then, at twelve, he was put on a train and sent to Mumbai to be educated in English, at St Xavier’s High School.

Think about what that journey actually was in the early 1960s. A boy of twelve leaves a monastic house in a coastal village where he has been taught in his own language, and arrives in the largest, loudest, most commercially frantic city in the country, to be educated in a language he has not been using. Whatever else that does to a person, it gives them two registers. He has spent a career working in both.

What came next was theatre, not film. In Mumbai he acted in Kannada and Konkani plays, and the people he worked with in that world read now like a roll call of modern Indian drama: Satyadev Dubey, Girish Karnad, Amol Palekar. Stage work is where he learned the thing that is most obvious about him on screen, which is that he almost never appears to be acting. There is no visible effort in his performances. Fifty years of that is harder than it sounds, because the temptation in a long career is to develop a manner, a set of reliable gestures that an audience recognises and applauds, and he never did. He has the quality of a man who has walked into the frame from somewhere else and has somewhere else to be afterwards.

His screen career began in the early 1970s, and it is one of those filmographies where the sheer volume is the point. Three hundred films over fifty years is an average of six a year, every year, sustained through the entire adult life of a man who was also, at various points, doing other things entirely. He did not coast. The films he is loved for in Karnataka are spread across the whole span, not bunched at the start.

The other person in this story is his brother. Shankar Nag was six years younger, and he was a star in his own right and a director, and the two of them worked together. The piece of that relationship that Anant tells best is small and very revealing. An opportunity came from the north, a television series for Doordarshan, the state broadcaster, and Shankar was weighing it against continuing to direct films. Anant pushed him towards the television. It was a chance, he argued, for his brother to show what he could actually do as a director, and a serious opportunity coming south. Shankar took it, and the result was Malgudi Days, which became one of the most loved things Indian television has ever produced. Anant appeared in thirteen episodes of it.

He also told his brother, at one point, to stop producing films, because they were losing money on them. That is not a glamorous thing to say to a sibling who is a star. It is the kind of thing you say to someone you are actually looking after. Reading it back now, next to the retirement, a pattern shows up: this is a man with an unusually clear eye for when something is not working, and an unusual willingness to say so out loud and then act on it.

Which brings us to what the Phalke actually is, and why the timing of it matters.

The Dadasaheb Phalke Award is named for the man generally regarded as the father of Indian cinema, and it is given by the Government of India for lifetime contribution. It is not a popular vote and it is not a trade prize. It sits inside the National Film Awards, which are run by the Ministry of Information and Broadcasting, and it is handed over in person by the President of India. Nag’s was the award for 2024, presented at the 72nd National Film Awards. The ceremony was held last month at Ekta Nagar, in the Narmada district of the state of Gujarat, and President Droupadi Murmu gave him a shawl and a certificate.

He had already received the Padma Bhushan, one of the country’s highest civilian honours, the year before. So the state had not overlooked him. But the Phalke is a different kind of statement, because there is only one of them each year and it is explicitly about a whole life.

What makes it land is the order of events. He had already gone. He had made his own assessment of the work being offered to him, decided it was not worth doing, and stopped, without telling anyone, three years before the award existed as a possibility. The honour did not arrive in time to persuade him to stay, and it was not a consolation for being pushed out, because nobody pushed him. It arrived after the decision, on top of it, as if to say that the judgement had been correct all along.

By his own account he found the occasion slightly overwhelming. He said he was trying to take in the scale of the thing, which from a man who has been on film sets since the early 1970s is a more interesting reaction than gratitude. And when he spoke about the award itself, he did not keep it. He said it did not belong to him alone, that it belonged to everyone, and he dedicated it to Karnataka, the state whose language he had been schooled in to the eighth standard before anybody sent him anywhere.

There is one loose thread, and he left it loose himself, so it should stay that way here. Asked directly whether he has retired, he said yes, he has. Asked whether that is final, he said that if something interesting came along he would consider it, and then added, flatly, that it does not come along, and that he means across all the languages he could work in. Both of those things are true at the same time. He is not sitting by the phone. He is also not going to pretend that he would refuse a good script out of principle, because refusing good scripts was never what he was doing. The door is not locked. It is simply that nothing has walked through it in three years, and he has stopped expecting anything to.

That is a better ending than a farewell tour, and it is a far more honest one. A farewell tour is a performance about stopping. This was just stopping, privately, for a reason, with no audience and no reel. The reel came three years later, from the state, at a ceremony he had not campaigned for, for a body of work he had already closed.

It is worth separating the two honours, because they are not the same gesture. A high civilian decoration is given across every field in the country, dozens at a time, to scientists and civil servants and musicians and surgeons. The Phalke is one award, once a year, for one person, and it is only about cinema. Getting the first says the state has noticed you. Getting the second says the state has decided where you sit in the history of the form. For a man who had spent the previous three years quietly establishing that he no longer wanted anything the industry was offering, being placed in that history by the country rather than by the industry has a certain symmetry to it.

He read a hundred scripts looking for one worth doing and did not find it. That is not a sad story about a man the industry forgot. It is a story about someone who was still, at seventy eight, applying exactly the same standard he had applied at twenty five, and who turned out to be the only person in the room willing to enforce it.

The hearing in Suffolk Superior Court in Boston took a few minutes. The prosecution filed a motion stating that it was unable to sustain its burden of proof at trial, the judge dismissed the charges, and Stephen Pina, fifty eight years old, stopped being a man convicted of first degree murder. His mother Josephine was in the room. So were two other men who had each spent decades in prison for killings they said they did not commit.

Pina had already been out of prison for four and a half years when that happened. That is the part of this worth sitting with, because most people assume the walking out is the ending.

Before anything else, one thing needs saying plainly, because readers collapse it almost automatically and it is not fair to anybody to let that stand. Charges being dropped is not a court finding that nobody did it. No judge has declared Stephen Pina innocent. What the courts found was that his trial was unsound and that the state could not lawfully convict him on what it had. Those are different findings, and the distance between them matters.

It matters most to one family. Keith Robinson was twenty six when he was shot and killed outside the Mission Hill housing complex in Boston in the winter of 1993. Thirty three years later nobody has been held to account for his death, and with both of the eyewitnesses who testified now dead, that is almost certainly how it will stay. There is no version of this story in which his family gets an answer. Everything that follows is about what happened to another man inside the machinery of the case, and none of it undoes that.

The trial itself was thin, and that is not a characterisation, it is what the state’s highest court said in writing. There was no forensic evidence, no video, no photograph, nothing physical connecting Pina to the shooting. The case was two people who said they saw it.

The first was Timothy Hall. He picked Pina out of a photographic array and put his own confidence at roughly six out of ten. By the time he was on the stand he told the jury he had always been a hundred percent certain, and explained the gap by saying he had been having second thoughts earlier. His descriptions of the shooter had shifted over time on height, on weight, on complexion.

The second was Debra Annas, and she was the one the case rested on. She said she had known Pina for a year and a half or two years and had bought drugs from him most nights. She testified that she watched a man get out of a maroon car, fetch a gun and shoot Robinson. But when police first showed her eight photographs weeks after the killing, she identified nobody. She picked Pina out of a different array that September, six months after the shooting. She testified that she had been high on crack cocaine when she viewed at least one of the arrays. Her account and Hall’s did not match on significant points.

And there was something about Debra Annas the jury was never told, and neither was the defence.

When she testified in January 1996, she was a psychiatric inpatient. She was hospitalised at Taunton State Hospital at the time she took the stand in the case that convicted Stephen Pina of murder. The prosecution had in its file a detective’s note containing the contact details of her psychologist. Nobody told Pina’s lawyers. Her records also reflected diagnoses that bore directly on perception and memory. More than a decade passed before the defence learned any of it existed.

It is worth being precise about why that is a serious failure rather than a procedural annoyance, because the phrase withheld evidence gets used loosely.

A criminal defendant is allowed to test a witness’s credibility, and one of the legitimate ways to do that is to ask whether the witness was in a condition to perceive events accurately, to remember them accurately, and to describe them accurately. That is not an attack on a person’s character. It is the core function of a trial. The prosecution carries a constitutional obligation to turn over evidence favourable to the defence, including material that undermines its own witnesses, and it carries that obligation whether or not the defence knows to ask for it. The state’s highest court said in its decision that the Commonwealth had violated that obligation.

When the Supreme Judicial Court of Massachusetts ruled on the case, Justice Frank Gaziano writing for the court, it put the point about as directly as a court does. Given the evidence linking what was in those treatment records to Annas’s capacity to perceive, remember and articulate the events she described, the records had impeachment value. In other words, the defence had been deprived of a real and material line of attack on the only witness who truly mattered. The court added that this was not a case where there was a mountain of other evidence to fall back on. Annas was the key to the prosecution’s case, and there was nothing hard underneath her.

It was not only the psychiatric records, and the piece would be misleading if it suggested otherwise. The motion judge, Peter Krupp, granted a new trial on the totality of what had been withheld, newly discovered or newly available, finding that justice may not have been done. Three strands did the work. The first was Annas’s hospitalisation and treatment history. The second was evidence pointing at other people: two men, Brian Johnson and Anthony Woods, who had access to the murder weapon and were stopped in a stolen maroon Nissan in the weeks after the killing, with Johnson carrying injuries consistent with a robbery that had happened hours before Robinson died, and Woods resembling Pina. That material had not been shared either, and the judge found it supplied the missing links that turned an alternative theory from speculation into something a jury could have been asked to consider. The third was DNA. Testing done in 2015 and again in 2022 excluded Pina as a contributor to samples taken from the weapon and from the victim’s jacket, and it is worth noting honestly that the court treated the DNA as marginal support rather than the centrepiece. Krupp also observed that there was no evidence Pina knew Robinson or had any motive to shoot him.

Pina had been convicted in January 1996, at a second trial, the first having ended in a mistrial the year before. The sentence for first degree murder in Massachusetts is mandatory life imprisonment without the possibility of parole. His appeal was rejected by the same court that eventually ruled in his favour, back in 1999. His lawyers filed the motion for a new trial in 2020.

Now the strange part.

In March 2022, Judge Krupp ordered Pina released from prison while the court went on considering that motion. He did not vacate the conviction. He said the likelihood that Pina would ultimately prevail looked fairly high, scheduled a further hearing to examine the evidence more closely, and let him go home to his mother’s house in Dorchester that night. Pina had served twenty eight years.

So he was out. And in the eyes of the law and of every records system in the state, he was a man convicted of murder in the first degree, serving a life sentence, who happened to be at liberty.

Think about what that actually involves day to day. He was on GPS monitoring, which means an ankle device and a log of where you are at all times. He had probation requirements, including weekly check ins with the court. His conviction was not under review in any way a form could reflect: it simply stood. A first degree murder conviction is the single heaviest thing a background check can produce, and the ordinary machinery of adult life runs on background checks. Rental applications. Employment. Licences of almost any kind. Anything that asks whether you have ever been convicted of a felony, where the honest answer was yes and the explanation was a sentence nobody has time to hear.

There is also the part that no form captures. A man in that position cannot plan. He cannot take a job that requires a year of training, or sign a lease with confidence, or make a promise to anybody about where he will be, because the honest answer to any question about his future is that it depends on a ruling that has not come yet. Everyone around him treats his release as the end of the story, and he is the only one who knows it is not. He has to keep explaining that the thing everybody is congratulating him on has not actually happened.

And underneath all of it, the state was actively trying to put him back. That is not a metaphor. The prosecution was litigating to keep the conviction standing. For a man who had already lost twenty eight years, the appeal was not an abstraction about legal principle. It was a live attempt to return him to a cell for the rest of his life, and he had to get up every morning and live beside it without knowing which way it would go.

The common assumption about these cases is that a judge opening the door means you have won, and it is wrong in a way that is hard to grasp until you look at the dates. Krupp did not vacate Pina’s conviction in 2022. He vacated it on the third of February 2025, almost three years after sending him home. For those three years Pina was free and the conviction was entirely intact, with no ruling in his favour on the merits at all, waiting on a decision that might simply have gone the other way and sent him back inside to die there.

Then, when the vacatur finally came, it did not end anything, because the Commonwealth appealed it. So the vacatur was provisional too, and he waited another eighteen months for the Supreme Judicial Court to say whether Krupp had been right.

There is an institutional story folded into that delay, and it should be attributed rather than asserted, because it is a characterisation of two officials’ positions as reported. The withheld material was reportedly uncovered by an integrity review unit inside the district attorney’s own office under Rachael Rollins. Her successor in the post, Kevin Hayden, opposed the new trial and took the appeal. The same office that found the problem fought the remedy, four years apart, under two different names on the door. Whatever you make of either decision, it is a large part of the reason a man spent four and a half years free and convicted at the same time.

The Supreme Judicial Court affirmed Krupp on the twenty eighth of July 2026, after hearing argument in the spring. Just over a fortnight later, in the late summer, the Commonwealth filed the motion that ended it and the charges were dismissed. By then the arithmetic of a retrial had become impossible: both eyewitnesses were dead, their earlier testimony was unlikely to be admissible, and there had never been any physical evidence to fall back on. The state did not concede that Pina was innocent. It stated that it could not carry its burden. That is a narrower thing, and it is also the only thing the state was ever going to say.

In the room were Jill Tessier of the Committee for Public Counsel Services, who had represented him, and Josephine Pina, who had outlasted the whole case. Also present were Sean Ellis and FuQuan McGee, each of whom had served more than two decades before their own convictions came apart. When Pina had walked out of that same courthouse four years earlier, it was Ellis and another exonerated man, Darrell Jones, that he turned and held on to in the corridor. There is a small and specific community of people who understand what the four years in the middle feel like, and they tend to show up for each other.

The National Registry of Exonerations lists him now: Massachusetts, murder, convicted 1996, exonerated 2026, roughly thirty years lost. The contributing factors recorded against the case are mistaken witness identification, perjury or false accusation, official misconduct and inadequate legal defence. Two different measures of lost time sit in that entry and both are true: twenty eight years in custody, and about thirty years carrying the conviction. No compensation is recorded, and whether he intends to seek any has not been established.

Later that same month he sat down with Sean Ellis on Boston Public Radio and talked about it, as two men who had moved from accused to convicted to imprisoned to free while maintaining throughout that they had not done it.

It would be easy to write Stephen Pina as a saint and the twenty eight years as a journey, and both would be a kind of lie. He is a fifty eight year old man who lost most of his adult life to a case built on two shaky identifications and a file the defence never saw, and who then spent another four and a half years at home with an ankle monitor and a murder conviction, because getting out of prison and getting the conviction off you turn out to be two separate processes with years of silence in between.

What finally ended it was not a dramatic exoneration or a confession or a new suspect in the dock. It was the Commonwealth of Massachusetts putting on paper, after thirty three years and an appeal it chose to bring and lost, that it could not prove its own case. Thirty three years to write it. A few minutes to read it out. And his mother was alive to hear it.

Sometime in the early 1950s, a young Swedish conductor was in conversation with the Gävle Symphony Orchestra about a job. The orchestra was interested. Then the question of his availability came up, he explained that he would not work from sunset on Friday until nightfall on Saturday, and that was the end of it. They turned him down because of it. The Norrköping Symphony Orchestra, a little further south, decided to make an exception for him, and in 1954 he took the post there and kept it for eight years.

Seventy-two years after that rejection, in September, the Japan Art Association announced that the music laureate of its 2026 Praemium Imperiale was Herbert Blomstedt. He is 99 years old. The other four laureates were the artist Jenny Holzer, the sculptor Andy Goldsworthy, the architect Daniel Libeskind and the actor Cate Blanchett. The citation praised the depth and balance of his readings across an unusually wide repertoire, and said his authority comes from meticulous preparation and complete command of the score.

Which is a slightly extraordinary thing to say about a man who spends one day in seven refusing to prepare.

Blomstedt is a Seventh-day Adventist, and the Sabbath observance follows from that. Adventists keep the biblical Sabbath on the seventh day rather than Sunday, counting it from sunset on Friday to sunset on Saturday, and treat it as a day set apart from ordinary work. In Blomstedt’s case the practice is more specific than most people assume, and the detail is the whole point. He does conduct concerts on Saturdays. What stops at Friday sunset is the preparation. He does not rehearse with an orchestra, and he does not sit down with a score and study it. The work ends; the performance does not.

That distinction is not a loophole he found. It is the result of a conversation he has described having, as a young man, with Johannes Norrby, who ran the Stockholm Philharmonic and was a believing Christian himself. By Blomstedt’s own account, given to a church publication when he turned 90, Norrby said he respected the position on rehearsals but did not understand the rest of it. If you say you make music to glorify God, Norrby asked him, why would you not want to glorify God on the Sabbath?

Blomstedt went away and thought about it, and read the Gospels, and noticed how often and how deliberately Jesus is described as doing things on the Sabbath. He settled on a model drawn straight from his own father, who was an Adventist pastor: a minister spends his week writing the sermon and then stands up and delivers it on the Sabbath, and the delivery is not work in the sense that the writing is. So Blomstedt rehearses all week and studies scores all week, and when the sun goes down on Friday he stops studying, and whatever happens on the podium on Saturday happens on what the orchestra and he have already built together. He has described it in almost exactly those terms: on the Sabbath they do not practise any more, they play what they rehearsed.

Now consider what that costs in this particular profession.

A symphony orchestra is a weekend business. Its revenue comes from the nights when the hall is full, and the hall is fullest on Friday and Saturday. The working week of a guest conductor is built backwards from those nights: a series of rehearsals across the week, tightening as it goes, with the last and most important one commonly falling on the Friday, sometimes a public general rehearsal, sometimes a final sitting the same day as the first concert. That is the session where everything that has gone wrong gets fixed. It is the one a conductor would least willingly give up.

Blomstedt gives it up, every single week, in winter especially. In northern Europe in December the sun sets in the middle of the afternoon. In Stockholm at the turn of the year it is dark before four o’clock, which means a Friday is not a working day with an inconvenient ending; it is most of a working day gone. Any orchestra engaging him has to shift rehearsals to earlier in the week, compress them, or lose one. It has to do that around the schedules of eighty or a hundred players, a hall booked months ahead, soloists with their own commitments, and in some cases a broadcast slot. Doing it once is a courtesy. Doing it for every engagement of a seventy-year career is a structural accommodation, and somebody at every one of those orchestras had to decide it was worth it.

In the 1950s nobody had decided that yet, which is why this is a hard-choice story rather than an anecdote about an eccentric maestro. A conductor in his twenties applying for a provincial orchestra job has no leverage of any kind. He is not a brand, he is not a draw, and there are more qualified young conductors than there are posts. The only thing he can offer is total availability, and that is precisely the thing he was not offering. Gävle’s decision was not unreasonable from where Gävle was standing. It is simply what the choice cost, in the only currency a young musician has, which is the chance to be in front of an orchestra at all.

Most stories about a costly decision are about one decision. This one is the same decision, made again, every seven days, for the length of a working life. Nobody renegotiated it as he got more famous and could have. If anything the stakes rose, because the bigger the orchestra the more expensive its rehearsal time.

And the career did get big. After Norrköping he went to the Oslo Philharmonic in 1962, then the Danish Radio Symphony, then the Swedish Radio Symphony, which established him as one of Scandinavia’s leading conductors before he was fifty. His international reputation turned on his decade as chief conductor of the Staatskapelle Dresden, from 1975 to 1985, a partnership with one of the oldest orchestras in Europe that produced recordings people still argue about favourably. Worth noting, and worth noting carefully: Dresden in those years was in East Germany, a state that was officially atheist, and nothing I could find suggests his observance was obstructed there. From 1985 to 1995 he was music director of the San Francisco Symphony, where he is generally credited with lifting the orchestra into the front rank of American ensembles through sheer rehearsal discipline. Then Hamburg, and then from 1998 to 2005 he was Kapellmeister of the Gewandhaus Orchestra in Leipzig, which is about as serious a chair as the art form offers.

His Beethoven is admired for rhythmic drive and architectural clarity. His Bruckner is admired for patience, which is the hardest thing in Bruckner. He has done as much as anyone alive to put Carl Nielsen, Jean Sibelius and the almost forgotten Franz Berwald in front of audiences who would otherwise never hear them. He conducts from memory routinely. His reputation is built explicitly on serving the composer rather than himself, which in a profession full of personalities is a position as much as a method.

The honours accumulated accordingly: two Grammy awards in the 1990s, Sweden’s Rolf Schock Prize, Denmark’s Léonie Sonning Music Prize, and in 2022, in Dresden, the Great Cross of Merit with Star of the German Order of Merit, presented by the state premier of Saxony, who remarked on his vitality and mental elasticity. He already held the Great Cross from nearly twenty years earlier. The star was the upgrade.

He also kept working into his nineties in a way that has very few parallels. After a fall put him in hospital in late 2023 and forced a string of cancellations, he was back on a podium within months, conducting the Leipzig Gewandhaus in Schubert and Berwald, and at that point had engagements booked more than a year ahead.

This past year was harder, and this is the part that has to be reported precisely, because the temptation to round it off in either direction is strong and both roundings would be wrong.

In the late spring he conducted Mahler’s Ninth Symphony with the Detroit Symphony, then travelled to San Francisco to conduct the same piece with the orchestra he once ran. He led the Friday concert. During it he became visibly tired, and afterwards he was taken to hospital, which a San Francisco Symphony spokesperson described as a decision made out of an abundance of caution. He withdrew from the Saturday and Sunday performances and David Robertson took them over. A week later he cancelled two Mahler Ninths with the Swedish Radio Symphony in Stockholm, where he is honorary conductor; the orchestra said he had been through a period of health problems, was on the road to recovery, and had been prescribed rest. Alan Gilbert stepped in at very short notice.

At that stage he was still intending to conduct a special concert in Bamberg in July, Beethoven and Brahms, programmed in honour of his own ninety-ninth birthday, with the Bamberger Symphoniker, whose honorary conductor he also is. According to a statement from that orchestra reported at the time, he was unable to travel for health reasons and did not conduct it. I was not able to reach the orchestra’s own notice directly, so that one rests on a secondhand transcription of an official statement rather than on the document itself. He is also reported to have withdrawn from concerts in September, in a trade report I could not open.

What I can establish is this, and no more than this. He has cancelled engagements this year for reasons of health. No orchestra, agency or family statement announcing a retirement exists that I could find, and his management in Munich was still publishing news about him in September. Around the time of the award, reports described him as scheduled to return to conducting in April and July of next year, around his hundredth birthday. The San Francisco Symphony’s current season carries three concerts in January under the title Celebrating Herbert Blomstedt, and when he withdrew in the spring the orchestra said he was expected back at that hall in the new year. Whether he will stand in front of an orchestra again is not something anyone outside his family is in a position to say, and nobody who respects a 99-year-old man should pretend otherwise. What is documented is that the diary is not empty and nobody has closed it.

Into the middle of that came the telephone call from Tokyo.

The Praemium Imperiale is not widely understood outside Japan, which is a shame, because it is one of the largest arts prizes in the world. It was established in 1988 by the Japan Art Association, the country’s oldest cultural foundation, in memory of Prince Takamatsu, who had been its honorary patron, and first awarded the following year. The structure is deliberate. There are five fields, painting, sculpture, architecture, music, and theatre or film, chosen precisely because they are the areas of human achievement that the Nobel prizes do not touch. Each year an international panel of advisers nominates candidates and the association selects one laureate per field, for a lifetime body of work rather than a single piece. Previous music laureates stretch across composers, performers and conductors from every major tradition.

Each laureate receives 15 million Japanese yen, which at the exchange rate prevailing when the 2026 announcement was made comes to roughly 96,000 US dollars, along with a medal and a testimonial letter. The presentation takes place in Tokyo at the end of October, at the Meiji Kinenkan, where Prince Hitachi, the association’s honorary patron, hands the medals over in person. This was the thirty-seventh edition.

It is a prize for a life, in other words, awarded by a committee that had the whole of a life to look at, and what it chose to praise in Blomstedt was his preparation.

That is worth sitting with for a second. The quality the award singled out, complete command of the score, is the direct product of the week he spends working, and it exists alongside a rule that shuts that work down for roughly a seventh of his existence and has done since before most of the people in the hall in Tokyo were born. The constraint did not dilute the thing he was eventually honoured for. On his own account, the two are not in tension at all: the week is for building, and the Sabbath is for whatever you have built.

None of which required anybody to agree with him. The interesting thing here is not the theology, it is the arithmetic. He decided something in his twenties that made him more expensive and less convenient than every one of his competitors, in a business with no shortage of conductors, and he never once traded it away for a better job. He lost at least one post over it that we know about. He almost certainly lost others nobody recorded, because orchestras that decline to hire you do not usually explain why.

The Gävle Symphony Orchestra, for what it is worth, is still going. It plays in a hall on the Gävle river, it has had a long line of chief conductors since the 1950s, and it is a perfectly respectable regional orchestra in a country full of them. It is the job Herbert Blomstedt did not get. He went on to Dresden, San Francisco and Leipzig instead, and at 99, with his diary half empty and half full, he was handed a prize in Tokyo for the thoroughness of his preparation by people who had no idea they were describing six days of the week.

Town Brook runs about a mile and a half through the middle of Plymouth, Massachusetts, from a freshwater pond down to the sea, and for roughly two centuries a fish trying to swim up it had to get past six dams. Five of them are now gone. Before the removals began, something like 25,000 river herring came up the brook in a year. Recent runs have come in above 200,000, and in one recent season the camera that counts them logged more than 276,000 fish. The man most responsible for that is David Gould, who runs the town’s Department of Energy and Environment, and who has quietly pulled out so many barriers that a retired federal biologist told the writer Ben Goldfarb that Gould has probably removed more dams single-handedly than anyone east of the Mississippi.

Up the coast, in Durham, New Hampshire, somebody on a lawn above a different river had put out an all-capitals sign. It read HELP SAVE MILL POND DAM.

That sign is the whole problem in four words, and it is a much better story than the usual one about rivers.

The famous dam removals are the enormous western ones. The Klamath, where four hydroelectric dams came down and opened more than 400 miles of spawning water in a single project. The Elwha in Washington State before that. Those are the ones that get documentaries. But the great majority of dam removals in the United States happen in the east, where the barriers are small, old, and almost completely unknown outside the towns that own them. By the accounting of the conservation group American Rivers, more than 1,100 eastern demolitions have reopened over 20,000 miles of river. The same organisation counts about 2,350 documented removals across the whole country since 1912, with a hundred of them in a single recent year, reconnecting nearly 4,900 miles of water, the most ever recorded in twelve months.

And then there is what is left. One regional inventory puts at least 35,000 dams still sitting across eastern fish runs. More than 3,000 of them are in Massachusetts alone, a figure the state’s own restoration programme confirms, along with the detail that roughly 541 of those Massachusetts dams, about eighteen percent, are in poor or unsafe condition. Connecticut has something like 5,000. New York has around 9,000. Almost none of these are the concrete monsters of the Colorado basin. Most are low walls of granite block or mortared fieldstone, built when the only thing that could turn a wheel was falling water.

To understand why there are so many, you have to picture the economy that made them. From the middle of the 1600s, colonists in New England dammed essentially every stream capable of turning a wheel. Gristmills to grind corn. Sawmills to cut boards. Fulling and textile mills. Paper mills. Tanneries. Anchor forges. By the mid 1800s, by one calculation, more than 65,000 mills had been stapled across eastern creeks. Town Brook’s six dams, in that mile and a half, had powered grain, cloth, leather shoes, and anchors.

The fish paid for it. The great migratory runs of the northeast were alewives and bluebacks, collectively called river herring, silvery things about a foot long that spawn in freshwater and feed at sea. They came in volumes it is hard to credit now. One historian called them the passenger pigeons of the sea, and scientists have estimated that New England’s dams have cost the region’s ecosystems more than 20 billion pounds of alewives over four centuries, something close to the weight of 300,000 humpback whales. Shad went the same way, along with Atlantic salmon, sturgeon, sea lamprey, tomcod and eels.

One of the earliest people to get properly angry about this was Henry David Thoreau. In 1859 a committee of farmers in Billerica, Massachusetts, hired him to survey the Concord River, because they believed a dam owned by an industrialist named Charles Talbot was flooding their hay fields and they wanted a court to order it destroyed. Thoreau found the work tedious and said so in his journals. But he took the job partly because he already hated that dam. A decade earlier he had written about it blocking the spawning runs, and asked, in a line that has outlived every mill it was aimed at, where the poor shad was to find redress. His surveys proved the farmers right. The state let the dam stand anyway. Thoreau predicted gloomily that it might take a few thousand years for nature to level the Billerica dam, if the fish could be patient.

He was wrong, which is the good news. It has taken about 167 years rather than a few thousand, and the Talbot Mills dam may finally come down. But Thoreau was also right about something else, which is that dams make intractable opponents. He just had the wrong idea about why.

Gould’s education in this began in the early 2000s, when Plymouth went looking for federal help to repair a bad fish ladder on the Billington Street Dam. The ladders on Town Brook were a disgrace. They were too long, badly sited, and hard for fish to find in turbulent water, and Gould has described watching thousands of herring wash back downstream bruised and cut. The request landed on the desk of Eric Hutchins, then a restoration biologist with the National Oceanic and Atmospheric Administration, who suggested something nobody in the town had seriously proposed: do not fix the ladder, take out the dam.

In 2002 they did. An Army Reserve engineering battalion pulled the Billington Street Dam apart with excavators in exchange for food, fuel and lodging, which is the kind of arrangement that only happens once. It became the first coastal dam removed in Massachusetts. Then Gould went after the others, one at a time, with whatever money he could assemble: Off Billington Street in 2013, the Plymco dam in 2015, the Holmes Playground dam on Newfield Street in 2018, which also restored 2,000 feet of channel.

What he discovered along the way is that in a town with four centuries of industry layered on top of itself, dam removal is mostly not dam removal. Asbestos shingles had to come off mill buildings. Rights of way had to be legally extinguished. Contaminated sediment that had settled behind the walls for generations had to be dug out and landfilled. The brook itself had to be rebuilt afterwards, because the bottom of a drained reservoir is not a river, and boulders, bends and streamside plants all had to be put back by hand. And the permitting was enormous. Gould’s own summary of it is that you are working through the same regulatory framework that would permit a large retail store.

The masonry of a couple of the old Town Brook dams had crests that doubled as town streets, and those were converted into bridges, so that the road still crosses where it always did and the fish now run underneath. Walk the trail upstream today and Gould will point at a field, a park, a patch of oak and pine, and tell you it used to be underwater. The evidence is subtle once you know to look: a rotting dock skeleton, a high water stain on a glacial boulder.

So that is the success. The complication is that nobody is entirely sure how big it is.

Nathan Furey is an associate professor in the biological sciences department at the University of New Hampshire, where he studies fish ecology and animal movement. He came to the state in 2018 having done most of his doctoral research on relatively undeveloped Pacific salmon rivers in British Columbia, and found himself cast in a role he had not trained for. When you are the fish person at a state university in New England, people expect you to tell them what happens after a dam comes out.

The honest answer is that very often nobody knows. Furey works the Great Bay watershed, a tidal estuary fed by seven rivers. On one of them, the Bellamy, two old dams at a defunct woollen mill funnelled the whole river through a narrow pipe until a coalition of agencies, nonprofits and the mill’s current owner demolished them a few years ago. After that, the state tried to count returning alewives by visual survey, which means a person standing by the water tallying fish. Furey’s assessment of what that tells you is blunt. They have no idea how many fish come back there.

Alewives are not salmon. As Furey puts it, they cannot jump, they are not powerful swimmers like sturgeon, and they are not eels that can climb and crawl. Taking a wall out does not automatically mean a foot-long herring can get through the rapids left behind. So Furey and his students catch them as they enter the river and implant rice-grain microchips called passive integrated transponders, then chain a battery box to a streamside tree with a cable running down into a riffle, so that every tagged fish that wriggles past gets logged by serial number. It is a very old migration being surveilled by very new equipment, and it exists because the celebratory numbers everyone quotes have mostly never been properly measured.

Which brings us back to the sign on the lawn in Durham.

The Mill Pond Dam on the Oyster River was built in 1913, about 140 feet across, originally for water power it had not generated in decades. It was one in a series of dams that had stoppered that river since the 1600s. Water sheeted over its lip and frothed at its base and the pond above it was, by any ordinary standard, lovely. Families skated on it in winter and paddled it in summer. People fed the ducks.

It is very easy to write about those people as an obstacle. They are not. They are the reason this story is interesting, because what they are defending is real.

A millpond is not a mistake to someone who learned to swim in it. It is a view from a kitchen window that has not changed in three generations, a place where a grandmother watched her children and then her grandchildren, a fixed point in a town that has lost most of its other fixed points. Dam enthusiasts are sometimes labelled NOOMBYs, for Not Out of My Backyard, which is a clever enough joke and also a way of not taking the argument seriously. In Billerica, officials objecting to the demolition of Thoreau’s old enemy argued that the Talbot Mills dam had long served as a symbol of the Industrial Revolution and remained one of the major focal points in the area. That is a historical claim, and it is true. Eric Hutchins once found himself arguing with a man whose objection was that his golden retriever played fetch in the impoundment. Hutchins countered that the dog was stirring up toxic sediment and probably poisoning itself, which is also true, and which is not the point the man was making.

Here is the asymmetry that actually decides these fights, and Furey named it exactly. On the west coast, salmon runs collapsed inside living memory, so people remember what the river used to be. In New England the rivers have been plugged for so long that almost nobody alive has seen one run silver with herring. The pond is in living memory. The fish are not. Two biologists, John Waldman and Karin Limburg, gave the resulting feedback loop a name: eco-social anomie. The fewer fish there are, the less anyone cares, and the less anyone cares the fewer there will be.

So the millpond defenders are not being sentimental against the facts. They are being loyal to the only version of the river they have ever been shown.

Durham is what it looks like when that changes. In 2002 the state declared the Mill Pond Dam structurally deficient and told the town to repair it, and engineers worked out that repairing it would cost considerably more than demolishing it. Residents protested, and the town dropped the idea. The logic did not go away. Opponents of removal gathered over a thousand petition signatures and forced the question onto a ballot, and in 2022 the town voted 1,706 to 596 to take the dam out, close to three to one. The project, about 3.1 million dollars with most of it covered by grants, began last summer and is due to finish next year with native planting around the drained pond bed. The town administrator’s summary was that the Oyster will run free for the first time since colonial settlers began damming it.

The pond is gone. People who loved it have lost something, and saying so is not a concession to nostalgia; it is just accurate.

Not every dam is an easy call either. Hundreds of northeastern dams are twentieth-century structures still feeding electricity into regional grids, some owned by towns, some by utilities, some by investors holding them inside renewable energy portfolios. A few power only a handful of houses. But with energy prices rising and data centre construction expanding, it is genuinely hard to argue for switching any generation off, and the people who say so are not being unreasonable. What is moving those dams toward removal is less often fish than money: as small hydroelectric plants age, repair costs and the flood risk of a failure can turn them into liabilities their owners would rather be rid of.

The money for all of this is unsettled. A great deal of federal infrastructure funding was directed at fish passage, and by one independent tally about 2.2 billion dollars of it was earmarked specifically for removing blockages in waterways, of which roughly 1.78 billion has been awarded and around 420 million was still waiting to go out. That same tally found none of the projects had been targeted for cancellation. Goldfarb’s reporting, though, describes a programme that put a billion dollars toward culverts and barriers falling into administrative limbo, with applications unreviewed and money unspent, alongside staffing and budget cuts at the federal agency that has driven dozens of New England removals. Andrew Fisk, who runs American Rivers’ northeast region, told him the reductions in funding and the pressure on partner agencies have made the work no easier. American Rivers has set a target of 30,000 dams removed by the middle of the century. At the current rate, hitting it would require going roughly ten times faster.

Fisk also said the thing that explains why Plymouth matters more than any funding table. Proximity to abundance, he said, is transformational. You cannot argue someone out of loving a pond. You can show them a river with a quarter of a million fish in it.

Which is what Plymouth now does every spring at its Herring Festival, where parents and children walk up the path beside the brook, someone has chalked LOOK FISH on the asphalt, and the alewives stack up thicker and thicker the higher you climb. Malissa Costa, a Mashpee Wampanoag cultural leader, keeps a kiosk there. Massachusetts has banned alewife harvest since 2005 and the Wampanoag are exempt; Costa takes around twenty each spring, sometimes with her bare hands, salts and smokes them, and gives them away to elders and tribal members who need them. She does not especially like the taste. She goes anyway, every year, because she feels she owes it. Her family has been meeting these fish at this brook since the end of the last ice age.

And here is how Plymouth is finishing the job, which is the part that nobody expected.

One dam is left on Town Brook, at the old grist mill, and the town is not going to remove it. The mill is a working museum and the pond is part of the town. Instead, with a ten million dollar federal habitat grant, Plymouth is building what engineers call a nature-like fishway, a constructed channel that routes the river around the dam, along with dredging the pond and repairing the dam itself. When it is done, herring, eels and the rest will get access to another 1.6 miles of brook and 270 acres of spawning water in the pond at the top of the watershed. The mill stays. The millpond stays. The fish get through.

It took twenty-odd years, five dams, an Army battalion, thirteen partner agencies and an unreasonable number of permits, and it ends not with a wall coming down but with a river being given a way around the last one. Gould sometimes pushes a kayak out onto the spawning pond to watch the fish swirl in the willows, by himself, which is a strange hobby for a municipal department head and a reasonable one for a man who has spent his career on this. He has said that to him the single most historic thing in Plymouth is this brook. In a town that has built an entire identity out of its history, that is a provocative thing to claim, and he is probably right. The mills ran for two centuries and stopped. The herring have been coming since the ice left, they are still coming, and nobody has yet managed to count them properly.

Inside the stadium, the whole thing was visible. The giant screens carried Christine Sun Kim in real time as she performed the American national anthem and America the Beautiful in American Sign Language, in front of a stadium full of people and alongside the singers who were doing the audible half of the job.

Everyone else got something else. In her own words, the people watching on their televisions, computers and phones got a seriously truncated version. One estimate of her total screen time across both songs is about ten seconds. The network cut between angles repeatedly and showed a couple of snippets of her. There was even a dedicated feed on the broadcaster’s own website, set up for exactly this, and that cut away too. Two days later the National Association of the Deaf, which she was there representing, posted the full performance on Facebook so that anyone who wanted to watch it finally could.

Kim wrote about it afterwards in a newspaper op ed with a headline that functioned as a shrug and an accusation at the same time: she had performed at the Super Bowl, and you might have missed her. The question she put in that piece is the one nobody at the network appears to have asked beforehand. Why stage a sign language performance that is not accessible to the people who would like to see it.

If you only know one thing about Christine Sun Kim, that is probably it, and it is the least interesting thing about her. What makes the episode worth more than a complaint is that she had already spent fifteen years making work about precisely this: who gets to decide what counts as communication, what hearing people assume sound is, and the specific texture of being left out of something while being thanked for taking part in it.

Last week the MacArthur Foundation named her one of twenty fellows for 2026 and attached 800,000 dollars to it, paid over five years, to spend however she likes. The citation does not mention the Super Bowl. It says the fellowship is for exploring the political, cultural and multisensory dimensions of sound, spoken language and communication, which is a dry way of describing a body of work that is frequently very funny.

Here is what that work actually looks like, because it is nothing like what people expect when they hear that a Deaf artist makes art about sound.

Kim draws, mostly in charcoal, and what she draws looks like the kind of thing a management consultant would put on a slide. Pie charts. Bar graphs. Musical notation. Flow diagrams with labels. She takes the visual grammar of information, the stuff we all accept as neutral and objective, and uses it to deliver observations that are personal and pointed. One series is called Deaf Rage, and it is exactly that: her frustrations with living in a world organised around hearing, plotted as angles on a diagram. Degrees of rage, measured and sorted.

She has been explicit about why she works that way. The best way to reach a wide audience that is not Deaf, she has said, is to use a format people already understand instantly. Nobody needs a wall text to read a pie chart. The joke lands before the viewer has a chance to get defensive about it.

She also likes charcoal specifically because it is messy, and because that mess sits oddly in the very clean white rooms where her work ends up hanging. That is not an accident either.

Her relationship with musical notation is the part that tends to make people stop and think. Notation is a system invented to put sound on paper for people who can read it. Kim takes that system and runs it in a direction it was never designed to go, borrowing its symbols for volume and duration and rest to describe things like social attention, politeness, interruption and the way a conversation gets handed around a room. A rest in music is a measured absence. Used her way, it describes a person being left out of a group chat.

American Sign Language is her first language, and a great deal of her work involves translating its movement and its grammar into something fixed. ASL happens in three dimensions and in time. It uses space, direction, facial grammar, the body’s position relative to other people. Putting that on a flat surface means losing almost all of it, so instead of trying to transcribe signs she traces what they do. One of her large murals, Prolonged Echo, treats the gallery floor as a flat open palm and runs the walls as the rebounding path of the sign for echo. The sign has a shape in the air. She gave the shape a building.

At the MIT List Visual Arts Center, in a show called Off The Charts that opened weeks after the Super Bowl, she put up drawings that broke the ASL alphabet into three separate points of view, which she labelled speller, bird and lurker. The speller is the person whose hands are doing it. The bird is the view from above, which is to say the view nobody in a conversation ever has. The lurker is the person watching from the edge of a group, trying to follow a language they half know. Anyone who has sat at a table in a country whose language they are still learning recognises the lurker instantly, which is the point. She was asked what she wanted visitors to take away and declined to offer a lesson, saying only that she wanted it to sit at the back of their minds and stay with them in terms of respecting Deafness and sign language.

Another drawing is titled sorry zero, sorry not sorry, which tells you most of what you need to know about the tone. The work is not polite. It is precise, and precision about social behaviour tends to read as rudeness.

It is worth saying plainly what a job like the Super Bowl one actually involves, because the public assumption is that signing is a kind of decorative accompaniment. It is not. Performing song lyrics in ASL is a translation problem with no good answers, done live, in front of an audience, with the prosody and rhythm and emotional register of the music all having to be carried by a body rather than a voice. It takes preparation and it takes judgement. It is work, and it was commissioned as work.

Her scale varies wildly and deliberately. There are small charcoal drawings. There are site specific murals that take over an entire wall. There are sculptures, videos and installations. One piece made with Thomas Mader, called ATTENTION, puts two oversized red inflatable arms through opposite walls of a white room, their index fingers pointing at each other and meeting over a rough stone sitting on the floor. It is an enormous physical pun about getting someone’s notice, which, if your access to a room depends on people looking at you, is not an abstract subject.

And then there is a work that answers the broadcast directly. In 2020 she made a piece called The Star Spangled Banner. She did not write a letter to the network. She put the thing into her own medium, where nobody can cut away from it.

There is another strand of her practice that gets less attention and is arguably the most radical thing in it. She has made audio works. One Week of Lullabies for Roux came out of her becoming a parent: she wrote a score, then invited seven other artists who were also parents to make songs from it, and required each of them to write a description of their song so that the piece would be conceptually accessible to Deaf audiences. Think about what that inverts. The audio is not the work with a translation bolted on afterwards for the people who cannot hear it. The description is part of the composition, commissioned at the same time as the music, from the same person. Access is not an accommodation added at the end. It is a material the piece is made of.

Kim has described what shifted her thinking, and it happened after she moved to Europe. She found that the Deaf community knew an enormous amount about sound. Not about hearing, about sound: how it moves, who controls it, what it does to a room, who is permitted to make it and how much. She has talked about sound as something auditory, visual and spatial all at once, which is a definition that includes vibration through a floor, the behaviour of a crowd, the way a loud room makes people stand, and the social fact of being shushed.

Once you have that definition, the Super Bowl stops being an insult and becomes a case study. A performance was commissioned in a visual language. It was transmitted on a medium that is almost entirely visual. And the people controlling the pictures spent the available seconds on other pictures. Not out of malice, as far as anyone can establish. The network does not appear to have offered a public explanation at the time, and nobody should fill that silence with invented motive. What happened is more ordinary and more useful than malice: a room full of professionals making rapid decisions about what was worth showing, and a sign language performance never making the cut, over and over, in real time.

Others who watched described it as feeling tokenised, the camera pulling away mid song on both songs. That is the right word. The performance was booked. It was announced. It was presented as inclusion. The part that would have made it inclusion was the part that did not get aired.

Her career, to be clear, is not a story of someone being overlooked. The institutional record is substantial and it predates all of this. Her work has been shown at the Museum of Modern Art and the Whitney Museum of American Art in New York, the Art Institute of Chicago, the Walker Art Center in Minneapolis, Secession in Vienna, the Contemporary Art Gallery in Vancouver and the National Museum of Modern and Contemporary Art in Seoul. She holds a bachelor of science from the Rochester Institute of Technology, taken in 2002, and two master of fine arts degrees, one from the School of Visual Arts in 2006 and one from Bard College in 2013.

The gap between that record and ten seconds of airtime is the actual story. Somebody whose work hangs in several of the most heavily visited museums on two continents was, for one afternoon, a logistical problem for a camera operator.

Last year she had her first major museum survey, which is the point in an artist’s life where the work stops being a series of shows and becomes a body. All Day All Night opened at the Whitney in New York and ran through to late September, co organised with the Walker Art Center, curated by Jennie Goldstein at the Whitney and Pavel Pys at the Walker with the independent curator Tom Finkelpearl, and assisted by Rose Pallone and Brandon Eng. It pulled together work from 2011 onwards: drawings, site specific murals, paintings, video installations and sculpture. It then travelled to Minneapolis, where it ran from the spring of this year through to the end of the summer, closing only weeks before the fellowship was announced.

That survey is the context for the prize, and it matters, because a fellowship given to someone in the middle of a career can look like a response to a moment. This one is not. The Foundation’s description of her is of an artist conveying the experiences of Deaf people and finding new ways to understand sound through drawings, large scale murals, video and interactive work. That is a description of twenty years of practice, not of an incident.

The money is the part that is difficult to explain to anyone who has not been a working artist. Grants normally come with a shape. There is a project, a budget, a timeline, a report at the end, a funder with a view about what the work should be for. The MacArthur has none of that. It is paid out across five years, there is no application, there is nothing to deliver, and the Foundation does not get a say in what she does next. It is money given to a person on the theory that the person is the thing worth funding.

For an artist whose subject is access, there is something quietly appropriate about being handed a sum with no conditions on it. Access, in her work, is rarely about generosity. It is about control: who holds the frame, who decides what gets transmitted, who gets to be in the room and who gets described to afterwards. An unconditional fellowship takes the frame out of anyone else’s hands for five years.

Kim is forty six and lives and works in Berlin. She was born in California, grew up in a Korean American family, trained in New York, and has spent well over a decade in Germany, which is where that realisation about the Deaf community and sound arrived. The MacArthur announcement lists her as a Berlin artist and that is accurate in the plain sense, though the work does not belong to anywhere in particular. Pie charts are universal. So is being talked over.

She has not, as far as anything published shows, made a public statement about the fellowship yet, which is in character for someone whose response to being cut from a broadcast was to make an artwork rather than issue a press release.

What is on the record is this. Six years ago she was given the largest audience available to any performer on the planet, and the apparatus in charge of that audience looked somewhere else for almost all of it. This year an institution read the entire body of work she had built about exactly that problem, and gave her the one thing no camera operator can take back, which is the next five years, to make whatever she wants, with nobody at a desk deciding how much of it the rest of us get to see.