Culture · Essay
The Fayum portraits: ordinary people who still look like themselves
About a thousand painted panels survive from Roman Egypt, bound over the faces of the mummified dead. They are the oldest portraits in which the sitter looks like a specific person, and almost none of them are of anyone famous.Written to last.
By Confinity Editorial · 2026-08-21 · 6-minute readQuiet tools, kept out of the way.
Most ancient portraiture is of rulers, and most of it is a type rather than a face. A pharaoh looks like a pharaoh. An emperor looks like the version of himself the mint has agreed on.
Then, in Egypt under Roman rule, for about two hundred years, somebody starts painting people who look like they could be identified from the painting. Not idealised, not typed: a specific nose, a slight cast in one eye, an unconvincing beard on a young man, a woman whose expression suggests she was tired of sitting.
There are around a thousand of them. Almost none is of anyone we can name from any other source.
They are funerary. Each panel was painted on thin wood, cut to shape, and bound into the wrappings of a mummified body over the face, so that the wrapped body carried a portrait of its occupant.
The practice is the meeting point of two systems. Egyptian mummification, which was thousands of years old by this point and had always used a mask so that the soul could recognise its body. And Graeco-Roman portraiture, which had developed an interest in individual likeness that Egyptian funerary art had not needed.
The result is a Roman face on an Egyptian body, prepared by Egyptian methods, often with a Greek name recorded and an Egyptian one alongside it. The people in these paintings lived in a genuinely mixed province and the objects show it in every detail: the hairstyles follow Roman court fashion closely enough to date the panels, the jewellery is local, the gods invoked on the wrappings are Egyptian.
Most are painted in encaustic: pigment suspended in hot beeswax, applied while molten and worked with a heated blade.
It is a difficult medium and an unusually durable one. Wax does not fade like a bound pigment and does not crack like tempera, and the dry conditions that preserved the Chinchorro dead preserved these too. The colours in a good Fayum panel are close to what left the workshop eighteen hundred years ago, which is why they are so unsettling to stand in front of. Nothing about them looks retrieved.
The brushwork is loose in places and startlingly modern. Highlights sit on the panel as separate strokes of hot wax. Twentieth-century painters noticed and said so.
The detail that changes how the whole group reads is the age distribution.
A great many of the portraits are of the young: children, adolescents, adults in their twenties and thirties. Where a panel can be matched to the body it was bound to, the age in the painting usually matches the age at death.
These are not the portraits of long lives. They are what a family commissioned when somebody died early, in a province where dying early was ordinary, and the painting exists because that person's face was going to be gone otherwise.
Some panels show signs of having hung in a house before burial, which would mean the family lived with the portrait for a while first. The evidence for that is debated. It is easy to want it to be true.
The Fayum portraits are also a case study in how not to recover things.
They came out of the ground in quantity in the late nineteenth century, in a period when the panel was worth money and the body was not. Portraits were cut away from mummies and sold; the mummies were frequently discarded. Findspots went unrecorded. The result is that most surviving panels are faces with no body, no name, no grave and no context, sitting in museums across Europe and America.
A few dozen were excavated properly, and those few carry most of what is actually known about the practice. The rest are beautiful and mute.
There is a bitter symmetry in it. An object made so that a specific person would not become anonymous, separated from that person by the people who saved it.
Set them beside post-mortem photography, which did the same job for Victorian families eighteen centuries later, and the continuity is not subtle. A face is what goes first. Names survive on stone, dates survive in registers, and the face survives only if somebody made an image of it while it was there to copy.
The Fayum panels work because they were made from life, or close to it, by somebody sitting in front of a person. That is the whole difference between a portrait and a memorial.
Most families now have thousands of photographs and no equivalent object, because a photograph nobody has labelled is not a portrait of anyone. In two generations the pictures on a phone will be exactly what a Fayum panel is in a museum: an undeniable face, and no way to say whose.
The remedy is unglamorous and it has not changed in eighteen hundred years. Put the name with the face, and the story with the name, while the person who knows both is still there to be asked. Confinity is somewhere to put them.