There is a promise being made in museums right now, and it is a generous one. Bring us a skull, it says, and we will give you back a person. Not a diagram of a person. Not a shelf of labeled bone. A face at eye level, close enough to count the lashes.
This summer a museum in Budapest is keeping that promise sixteen times over, in a show called Once We Were Like You. Sixteen people who died in the Roman town that stood where the city stands now. Six of them are built in silicone, finished with real human hair, eyebrows, eyelashes, and artificial eyes that hold your gaze. You can stand face to face with them, walk around them, study them from any side. That is the design.
Here is what I think we should refuse. The more lifelike one of these faces becomes, the less it is a window onto the past. A hyperreal face is not access to a dead person. It is an argument, made in silicone, that you have already met her. The better the face works, the more completely it takes a guess and hands it back to you as a memory.
Start with the woman the show crowns with a wreath.
In 1962 a man digging a water pipe on a hillside in the third district of Budapest struck a sealed stone coffin. It took four workers and a winch to raise the lid. Inside, wrapped in dense cloth, lay a body still whole after more than a thousand years, and it began to come apart the moment the air reached it. They saved what they could with what they had. They laid her on a sheet of styrofoam, fixed her with preservative, set a glass dome over her, and for more than sixty years that was that.
One evening in February 2026, outside patient care hours, she was brought into a hospital imaging center and scanned under its hygiene rules. Hungary's first CT of a Roman mummy. She went in on the same styrofoam and its wooden frame, one ear already missing its small bones, removed for DNA. The scanner cut her skull into slices six tenths of a millimeter thick, then two tenths. It found a little sediment inside the cranium and beads still lying under the shroud. The skull it mapped had lost its face a long time ago. The wreath was still on it.
By summer she had a face again: skin, brows, lashes, a steady pair of eyes, and the wreath lifted onto a head that finally had one. She has no name. There was no epitaph on the coffin, and Roman coffins were routinely used more than once. Nobody knows what she died of. Nobody knows what she was called. Nearly seventeen hundred years after her death, she is easy to recognize and impossible to identify.
Now the strange part. The museum is honest about all of this. It says so itself, in its own words: this is "a meeting of scientifically verifiable findings and historical imagination." The names and life stories attached to the faces are, it says plainly, "products of the imagination." Its plainest sentence: "We know almost everything about them, yet they remain unknown to us."
So the confession is right there, in the museum's own words. And the face works anyway. You can know exactly what the museum admits and still feel, standing in front of the silicone, that you have met her. That feeling is the thing to explain.
The easy explanation is that the science simply got good enough to earn it. It did not. In one blind test, practitioners started from just four skulls and built sixteen faces; viewers trying to match those faces to the right photograph, choosing from ten, got it wrong four hundred and three times out of five hundred and ninety-two. Only one of the sixteen faces did better than chance. A face can be made with real skill and real care and still not be a likeness of anyone. Other tests land better; built from a clean scan and shown in a small lineup, a reconstructed face can beat chance and then some. The point was never that these faces fail.
The next easy explanation is that visitors are just careless, that they never take in the admission and fool themselves. But the admission does not lose to inattention. It loses to the face. You can hold every word of it in your head and still feel you have met her. The museum did not hide the seam. It named the seam, and the face closed over it anyway.
What is left is the uncomfortable one. The lifelikeness itself is doing the persuading, and it does its strongest work in exactly the places where the evidence is thinnest. The problem is not that the face is imagined. The problem is that the imagined parts are the ones that make it feel most true.
Look at what the bone can actually carry. The skull gives you the rough width and height of the face, and even then the ears and the tip of the nose are the first things to go wrong. The soft tissue over it is not a skeletal fact; it comes from tables of living donors, and body weight moves those numbers more than age or sex do. The eyes are a guess; some methods just reach for a standard eyeball, a borrowed default that belonged to no one in particular. The pigment is a probability, and the tools are strongest at the far ends of the spectrum and weakest in the pale and intermediate range that much of Roman Pannonia would have lived in; one validation study called pale skin correctly less than one percent of the time. The hair, the lashes, the beard, the fullness of the cheeks, the small crookednesses by which you actually know a living face: none of that is in the skull. It is added. The name is invented. The story is invented. A woman buried in a children's cemetery is read as a nurse or a healer. Two others were laid face down with their legs bound, and the exhibition ties one to witchcraft and one to poisons, readings the coverage flags as interpretation, not anything the burials prove.
Every added lash is a small claim. The more you add, the harder it becomes for anyone to see the guess.
Every one of those choices is defensible on its own. Stacked, cast in silicone, and set at eye level where you can circle them, they stop looking like choices. They look like someone looking back.
This is not a Budapest problem. It is what hyperreal reconstruction does everywhere. A Bronze Age woman in the far north of Scotland was given a face with light skin and blue eyes; when the DNA finally came back it said brown eyes, black hair, skin a little darker than a Scot's today, and the face had to be remade. A figure from a sunken warship was first reconstructed as a man, then remade as a woman after a test found no Y chromosome. The artist who built a famous ice mummy's face said he wanted her to look "both scared and proud." The skull and the scan are evidence. "Scared and proud" is not. Both reach you in the same face, and the face gives no sign of where one ends and the other begins. A responsible reconstruction can be flatly wrong and never once look unsure. The old face did not look uncertain. It just looked like a person.
There is a real case on the other side. For most of history the record kept the emperors and threw the rest away. It saved the men with monuments and lost the women, the workers, the children, the migrants, the sick. Many of the skeletons in this show carry the marks of hard labor, long hunger, chronic inflammation: ordinary people whose names were never going to survive. Giving one of them a face is not vanity. It may be the first plain courtesy anyone has paid them in seventeen centuries. One reconstruction specialist puts it simply: when you see a face, you stop seeing bones in a case and start seeing a person. That is not a cheap effect. It is the whole moral argument for doing this at all, and it is a good one.
It is good, and it turns on a hinge. The argument holds only as long as the face keeps admitting what it does not know. It breaks the instant the face looks more certain than the evidence underneath it. And a hyperreal face, by construction, always looks certain. So the better the craft gets, the more the honest version of this work quietly works against itself. Every added lash is a small claim. The more you add, the harder it becomes for anyone to see the guess.
So stand face to face with her, if you want to. Walk around the model, study it from every side. But hold the two things the show puts under one roof: the skulls in their glass cases, and the finished faces built to stand where those skulls cannot. Keep the museum's own sentence louder than the eyes. We know almost everything about them, and they remain unknown. Do not let the wreath talk you out of the gap. The reports leave her a spindle, a wreath, cork sandals said to be cut like birds, a painted male portrait near her feet, and no name. She still has no name. The face gives you someone to miss. It was never going to give you her.