What Is Under the Paint

When the Ghent Altarpiece came into the studio, the team expected to clean it. What they found was that around seventy per cent of the outer panels was not van Eyck at all. It was sixteenth-century overpaint, applied over the original, and much of it had not shown up on the technical imaging taken before work began. Some of it was discovered only once the surface was being worked on.

That is worth pausing on. The most famous Northern European painting of the fifteenth century had been mostly covered in later paint for four hundred years, and the imaging available before treatment did not reveal the extent of it.

In 2019 the team removed the overpaint from the central lower panel, and the Lamb of God underneath was not the animal anybody expected. The familiar version has a mild, sideways, sheep-like head. The original is frontal, wide-eyed, with a face described by the project head as cartoonish, looking directly out at the viewer with far more intensity. Hyperspectral imaging, which captures many narrow bands of light rather than three, had helped establish what was original before anything came off, and the same analysis showed the later hand had quietly reshaped the animal's body too.

The internet decided it had been ruined. The face became a meme. A great many people who had never previously held a view about fifteenth-century Flemish painting were confident the restorers had botched it.

They had not. What they had done was remove four hundred years of somebody else's taste, and reveal that van Eyck's lamb was stranger and more confrontational than the polite version that had replaced it. The reason the original looks wrong is that we grew up with the overpaint.

Now set that beside the other case everybody knows.

In 2012, in Borja in Spain, a parishioner named Cecilia Giménez repainted a deteriorating fresco of Christ by Elías García Martínez. She had no training. The result was unrecognisable: a blurred, brown, hooded figure that bears almost no relation to the original. It became an international joke and then, improbably, a tourist attraction that has made the town money.

Both of these were called ruinous in the press. One was.

The difference is not skill, and it is not whether the result is attractive. It is authority and record. In Ghent, a documented institutional team with imaging, analysis and a published methodology removed a later layer and could demonstrate, in detail, what was original and what was not. In Borja, somebody with good intentions and no documentation painted over a work, and no analysis exists that supports what she produced because she was not producing a hypothesis about anything.

Which gives a rule the public instinct will not supply. The reaction that a restoration looks wrong because it changed what you were used to is unreliable in both directions: it defended four centuries of overpaint in Ghent and it correctly condemned the Borja fresco, for the same reason in both cases, which is that it was not tracking evidence at all.

The question to ask is always the dull one. Can anybody show you what was there before.