In 1972, a documentary film-maker named William Greaves talked his way into Duke Ellington’s townhouse in Harlem and filled it with people who had built a movement and then scattered. The painter Aaron Douglas was there. So was the photographer James Van Der Zee, who had made the portraits that are still how most of the world pictures Harlem in the 1920s. Eubie Blake and Noble Sissle, who had put a Black musical on Broadway when that was not supposed to be possible. The writer Arna Bontemps. Richard Bruce Nugent. Louise Thompson Patterson. The historian John Henrik Clarke. Romare Bearden. Ida Mae Cullen, widow of the poet Countee Cullen.
Most of them had not seen each other in decades. Greaves got roughly four hours of them talking. Then he put the film away and never finished it.
He did not lose it. It was not stolen, it did not burn, and nobody found it in an attic. It sat in his own possession for the rest of his working life, and he worked for another three decades. He died in 2014, at 87, having produced something in the region of two hundred films and directed or written more than half of them, and this was still sitting there uncut.
One of the people who had been in that townhouse holding a camera in 1972 was his son David. Fifty-odd years later, David Greaves is the one who finally decided what the film was. It is called Once Upon a Time in Harlem, it runs a hundred minutes, it is credited to both of them, and it opens in cinemas in the United States on 16 October. Between its premiere and that release it has collected a run of festival slots that documentaries of this kind almost never get, including the Directors’ Fortnight at Cannes, Telluride, Toronto and a main-slate berth at the New York Film Festival, and it currently sits at a hundred percent on Rotten Tomatoes across twenty-nine reviews.
The obvious question, the one the film cannot answer and the coverage keeps politely stepping around, is why the father never cut it.
Start with who he was, because it matters. William Greaves was born in Harlem in 1926, close enough to the tail of the Renaissance to have grown up inside its afterglow. He started out acting, got tired of the parts available to him, and went into documentary instead. He co-hosted and produced Black Journal, the public television programme that was, unusually and deliberately, made by Black journalists about Black American life, and won an Emmy for it. He made Symbiopsychotaxiplasm: Take One, an experimental film about the making of itself, shot in Central Park with three camera crews, which confused almost everyone in 1968 and is now in the National Film Registry. He taught acting for over a decade. He spent ten years on a film about the diplomat Ralph Bunche and then had to compress seventeen hours of material down to two.
That last detail is worth holding onto, because it tells you something about how he worked. Greaves shot enormous amounts and then had to face the cutting. And 1972 in particular was not a quiet year for him. It is the year of Nationtime, his film about the National Black Political Convention in Gary, narrated by Sidney Poitier. He was busy with a piece of live political history while also convening a room full of elders whose movement was already half a century in the past.
Nobody who has written about the Harlem footage has produced a document in which Greaves explains himself. The honest answer is that we do not know, and the film does not pretend to. What we do know is the shape of the problem he left behind, and anyone who has ever tried to cut a long conversation will recognise it immediately.
Think about what he actually had. Not a story. Not a plot, not an event, not a crisis that resolves. Four hours of very old people being pleased to see each other. Conversations in twos, conversations in small knots around a room, and then one long sprawling exchange that pulls nearly everybody in at once, which the reviews single out as the best thing in it. There is no argument to build and no ending to arrive at, because the only ending available is that everyone in the room is going to die, which is not an ending a film-maker in 1972 could use.
There is also the plain difficulty that this material is only legible if you already know who these people are. A viewer who cannot tell Aaron Douglas from the man sitting next to him is watching pensioners chat. To make it work you have to teach the audience the Harlem Renaissance while they watch, without turning the whole thing into a lecture with clips. That is a genuinely hard editorial problem, and it was harder in 1972, when the Renaissance had not yet been canonised in schools and museums the way it has since.
And then the reason that is least romantic and most likely. He had commissions. Documentary film-making at that level is a business of deadlines and funders, and a personal project with no broadcaster attached is the thing that gets moved to next year. Then next year again. Greaves worked until 2001. There were something like thirty years of next year.
Which brings us to the part that is actually unusual about this story, and it is not the footage. It is the inheritance.
David Greaves was in the room. He is credited as one of the cinematographers on the 1972 shoot, alongside Doug Harris, David Hoffman and Jerry Pantzer. So when he came back to the material after his father’s death, he was not a stranger opening a stranger’s boxes. He was looking at images he had helped make, of an evening he had physically been present for, shot by a young man who was himself and who no longer exists in any practical sense. That is a strange position to be in, and it cuts both ways. He knows things about that night nobody else alive knows. He also has to accept that his memory of it is now fifty years old and less reliable than the film.
The harder part is the authorship. Every edit is a decision about intention, and normally you can ask. In an ordinary edit suite, the question of whether a scene belongs in the film gets settled by the director saying what the film is about. Here the director is dead and the answer has to be inferred from the only evidence available, which is the footage itself and a lifetime of other films that show how the man thought. You can watch Symbiopsychotaxiplasm and learn that William Greaves was comfortable with mess, with people talking over each other, with letting the process show. You can watch the Bunche film and learn that he could be patient and conventional when the subject deserved it. Neither tells you what he wanted here.
So David Greaves had to make a decision that is not really an editing decision at all. He had to decide whether he was finishing his father’s film or making his own. The credit, both names, father and son as co-directors, is the answer he arrived at, and it is a more honest answer than either alternative would have been. Crediting only William would be a fiction, because someone made a hundred minutes out of four hours and it was not him. Crediting only David would erase the fact that the casting of that room, which is the real act of genius here, was entirely his father’s. Getting those specific people into one specific house on one specific night, and having the nerve to think it was worth doing, is the whole film. Everything after that is carpentry.
There is a technical dimension to this that is easy to miss. Material shot in 1972 on film, with sound recorded separately, does not simply open on a laptop. Picture and sound have to be brought back into relationship with each other, and every frame has to be transferred before anyone can even look at it properly, let alone start making choices. With four hours of coverage from multiple cameras running at a live event, much of it is people crossing in front of each other, backs to the lens, audio smeared by a crowded room. A modern editor working on modern footage takes decent sync and clean dialogue for granted. Here, a lot of the labour is just getting the material into a state where the editorial question can be asked at all.
There is also the question of what you are allowed to do to it. You can cut a conversation to make a point the speaker was not making. You can put a reaction from one part of the evening next to a remark from another and invent an exchange that never happened. That kind of thing is ordinary craft in documentary and it is also the exact place where a son finishing a dead father’s film about dead subjects has nobody to stop him. The restraint the reviews keep describing, letting one long sprawling exchange simply run, reads like a deliberate refusal of that power.
He did not work alone. Liani Greaves and Anne de Mare produced, de Mare and Lynn True cut it, and the composer Tamar-kali wrote the score. That is a real film crew doing real work on material older than most of them, and it took years.
It is worth being clear-eyed about what has been recovered and what has not. Nothing about this was rediscovered. The footage was never missing, and the family knew exactly what it had. What was missing was a film, and a film is not footage. It is a set of choices about footage. For fifty years those choices had not been made, and the reason the thing exists now is not luck or archaeology but somebody sitting down and doing the work that his father, for whatever combination of reasons, did not.
What that work produced, by the accounts of people who have seen it, is not a history lesson. It is a room. Reviewers describe friendship, needling, a gentle competitiveness between people who had been young together and were now in their seventies and eighties, comparing what they had made of themselves. Eubie Blake, who was around eighty-five that year and would live to be ninety-six, in the same space as Van Der Zee, who had photographed him decades earlier. Douglas, whose murals are now the visual shorthand for the whole movement, being talked over at a party. Nugent, who had written openly about desire between men in the 1920s and paid for it in obscurity, still there, still talking.
None of them lived long enough to see it. Sissle died within a few years of that night. Douglas, Van Der Zee and Blake within a decade or so. William Greaves outlasted every one of his subjects by decades and still did not finish the film about them, and his son finished it long after that. There is no way to write that as uplifting and it should not be. It is simply what happened.
What is not sad is what the material now is. A four-hour conversation from 1972 between people who made the Harlem Renaissance is not an interesting artefact. It is the only one of its kind. There was never going to be a second night like that, because by the time anyone thought to organise one, the guest list would have been empty. Greaves seems to have understood, in 1972, that he was on a clock and that the gathering had to happen whether or not he knew what to do with it afterwards. That is the decision the film rests on, and he got it right.
The commercial afterlife has been improbable. A hundred-minute documentary consisting largely of old footage of conversations premiered at Sundance in January and set off a bidding war. Neon, the distributor behind an unusual run of crossover successes, won it, put out a teaser in the summer, and is giving it a proper theatrical release rather than dropping it onto a streaming service, which for a film of this shape is close to unheard of. Along the way it picked up a nomination for the Cannes documentary prize. It is, on the numbers, one of the best reviewed films of the year.
David Greaves is in his seventies now himself. He spent part of his life as a young man working on his father’s crew, and part of his later life finishing his father’s last unmade thing, and the outcome is a film that has his name on it next to a man who died twelve years ago.
The temptation is to call that a gift passed down, and that is not quite it. A gift is something you are given. This was a job left undone, handed over without instructions, by someone who could no longer be asked what he meant. David Greaves took it, worked out an answer he could defend, and in October it goes into cinemas with both their names on it. The father found the room. The son decided what happened in it.