The experiments, gathered by the question they ask rather than the medium they use.
These are the experiments. Not everything I've made — only the work that went looking for something it couldn't specify in advance, gathered here by the questions each one turned out to be asking rather than by when it happened or what it was made with. The questions weren't all planned. As is so often the case, they gained focus after the event.
A word on the "I" that follows. I instigated each of these and directed them, but only one — People We Love — was close to something I made alone. The rest were made with other people, and couldn't have been made any other way. The first person here means the thread of the thinking, not a claim on every hand that touched the work.
It starts twelve years ago. In 2014 a group of us made The Knife That Killed Me, the first feature in the UK shot entirely against green screen, every set a construction, nothing behind the actors but a colour. The conceit is that the protagonist is dead before the film begins, and the world you watch is the one he leaves behind: built from his drawings, his writing and his imaginings, blended with multiple composited perspectives so that almost every frame fractures the single fixed lens that conventional cinema relies on. The point was never to fool you. It was to break the one-eyed perspective of the camera and draw your eye through the frame — to make you choose where to look, the way you do in a room rather than at a screen, and so to pull you from spectator toward participant.
Four years later I returned to the same question with the artifice fully foregrounded. The greenscreen Macbeth (2018) used an evolution of the same techniques — actors shot against green, a world built around them in post, but where Knife drew the eye through the frame, Macbeth made the frame itself the subject: Shakespeare put on film in the most theatrical way film could absorb. A cross-sectional doll's house from which the camera pulls back to find the action playing simultaneously in adjoining rooms, its edges left unfinished, walls and stairwells sketched rather than closed. In one sequence the camera pans down from Macbeth to the murder of Lady Macduff in the room below, binding killer to victim by architecture rather than cut. The point of all this was to take Shakespeare's most interior tragedy — a play that happens almost entirely inside one man's head, and give the interiority a space to live in: characters set by psychological distance rather than geography and Macbeth's isolation, at the end, rendered as a withdrawal toward the empty sky at the top of the structure, which is finally revealed to be an enormous sphere — the whole world disclosed, in the last moment, as a made and bounded thing. The construction stayed visible. The visibility was the invitation.
So that was the first question, the one underneath both films: can a made thing admit that it's made and lose nothing — can the seam stay in view, the single lens be broken, and the spell hold anyway? Every piece after this turned out to be the same question, asked from somewhere new.
People We Love asked the same question at the opposite scale, and it is the first of these works that is mine rather than ours. It opened in York Minster in November 2020: five life-size screens, each showing a slow cycle of faces, every person on screen gazing at a photograph of someone they love — a photograph you never see and which was never recorded. From across the room they read as a crowd, a wall of strangers. Then a narrow spotlight invites you to a single screen, and you are left alone with one steady, silent gaze, doing the only work the piece keeps for you: imagining the absent face that they can see and you cannot. My own part was the smallest part — I watched as the images were colour-graded, and arranged them myself. The work is made by whoever is standing in the spotlight, and it belongs to them. The second lockdown closed that first edition after only sixty-nine people had been filmed, and the enforced stillness — those faces gazing out into a near-empty Minster — suited the work better than any crowd would have. It was the piece that taught me the contract didn't need a public square. It only needed one person willing to do half.
Which raised the next question almost by itself. If the work lives in what the viewer brings, what happens when you stop putting it on a screen in front of them and set it down in their own room? Don Quixote is the answer we've just completed — a thirty-minute retelling of the novel performed on your tabletop by miniature figures, the character built digitally from the ground up, the whole thing borrowing its language from theatre and asking, again, that you supply the imagination it deliberately withholds. It shares its architecture with Macbeth — the doll's house, the world disclosed as a made and bounded object, but where the film held you outside the sphere, this puts the sphere on the table in front of you and lets you walk around it. The performance is linear; it can never be seen the same way twice, because where you stand and where you look changes what the story is. Knife asked your eye to dance across the frame. Here you move your whole body and the frame is gone — lean in, walk round, crouch to its level, and the story rearranges itself around where you chose to stand.
And now the work the others were, in retrospect, walking toward. The Darkroom — a feature in production now — is a film about a confession: a dying woman's account of something she did sixty years ago, and a young woman who listens, doesn't believe her, and reconstructs it from the evidence in the room, and finally imagines the parts neither memory nor evidence can reach. The film is made in three visual languages, because memory, evidence, and imagination do not look the same and shouldn't be made to: one register lit warm for the version she's needed to live inside, one forensic and suspicious of its own composition, one an open act of invention that knows it is inventing. The same villa, the same garden, rendered three ways that quietly disagree. A singular representation would have to choose which one was true — and the film's whole argument is that the choice is the lie. So the construction here isn't just kept in view the way Knife kept it; it's the moral substance of the thing. Every experiment on this page turned a tool built to hide its workings the other way: this is the one where the seam became the subject. It isn't finished. That's rather the point of putting it here.
None of these questions started with me, and they didn't start on this site. They run back through fifteen years of work made with Tom Wexler for public space -- Flock, which turned Trafalgar Square into a theatre with no performers but the crowd; Congregation, which attracted ten thousand strangers to a Pittsburgh square and showed them, from above, the shape they were making together. The spotlit stranger in People We Love is the same figure as the chalked body those installations kept returning to, which was itself borrowed from a Rembrandt. The thinking is slow. It changes medium, changes scale, loses a collaborator, changes its tools, but it has been one idea the whole way along. That earlier work has its own home, kept as a record rather than a shop window.
The KMA work, 2005–2020, is archived at kma.co.uk