The painting is smaller than a sheet of A3. Fifty-one and a half centimetres by thirty-eight and a half, oil on canvas, a Franciscan friar on his knees looking up at a light he cannot quite stand to look at. In 2018 it went through a Madrid saleroom as a Francisco Bayeu, which is to say as a competent devotional picture by a court painter whose name carries a respectable, unspectacular price. The Barcelona dealer Artur Ramon bought it, jointly with a colleague, and hung it in his gallery, and for two years that was the end of the story.
Then came 2020. Ramon moved the picture from the gallery to his flat, which is what dealers did that spring, and spent the lockdown looking at it across a room, every day, for months. Somewhere in that stretch the attribution stopped holding. The anatomy was wrong for Bayeu. The hands, particularly, were wrong for Bayeu. In September 2026 Spain’s culture ministry put it in the careful phrasing these things require: everything seems to indicate that this is an unknown work of Francisco de Goya from around 1780.
On 27 October 2026 the canvas goes on public view in Barcelona for the first time. In 2028 it is due at the Museo del Prado. What makes it worth more than the usual rediscovery paragraph is not the money, which nobody has quoted. It is the date.
1780: the worst year to be Goya and the best year to be Bayeu
Francisco Bayeu was not a rival. He was family, and he was the gate. Goya had married his sister Josefa in 1773, and Bayeu — Aragonese like him, a decade older, already established at court — was the man who brought him into the royal tapestry works and vouched for him in Madrid. In 1780 Bayeu’s name was the asset in that household. Goya’s was not.
That was the year it broke. On 5 July 1780 Goya was elected Academician of Merit at the Real Academia de San Fernando, submitting a Christ Crucified; in October he travelled to Zaragoza to paint the Regina Martyrum cupola in the Basilica of El Pilar, and presented his sketches to the cathedral chapter. The commission had been Bayeu’s to supervise — he was the project’s artistic director, too busy in Madrid and Toledo to paint it himself. Through the spring of 1781 the building committee and Bayeu picked at Goya’s handling: too loose, too free, not finished in the manner the chapel expected. His sketches for the pendentives were sent back. A priest, Padre Salcedo, had to mediate between the painter, his brother-in-law and the Junta de Fabrica. Goya finished the cupola on 28 May 1781 and left Zaragoza two days later, and never painted the second one he had been promised.
In a letter to his friend Martin Zapater that July, Goya wrote that he burned at the thought of Zaragoza and painting. He was forty-five days out of the city and still unable to say the word.
Six days after that letter, on 20 July 1781, the Count of Floridablanca named the painters for the altarpieces of San Francisco el Grande in Madrid. Goya was on the list. So was Bayeu. They would spend the next two years working on vast religious canvases — Goya’s Sermon of Saint Bernardine of Siena runs to 480 by 300 centimetres — in the same church, for the same patron, barely on speaking terms.
And somewhere in exactly that window, on a canvas a fiftieth the size of the Bernardine, Goya painted a kneeling friar. It then spent some part of the next two and a half centuries carrying his brother-in-law’s name.
Why the small religious pictures are the ones that go missing
There is a reason the gap in Goya’s catalogue is shaped like this. His famous work is enormous, public and documented — cupolas, altarpieces, royal portraits, the Caprichos, the Black Paintings. Those objects have paper. They were commissioned by chapters and ministers and kings, and the commissioning generated receipts, minutes, letters, payments.
A 51.5-centimetre devotional canvas generates none of that. It is a private object: a modello, a gift, a picture made for a chapel or a household, unsigned because signing it was not the convention and not the point. It leaves no minute in a Junta ledger. And when it eventually reaches a saleroom two centuries later, the cataloguer is looking at a small, dark, well-painted Spanish religious picture from the 1780s with no documentation at all — at which point the name that gets written on the label is the name that fits the school and moves the lot. In 1780, and for a long time afterwards, that name was Bayeu.
This is the quiet mechanism behind most rediscoveries, and it is almost never a conspiracy. It is the attribution of least resistance, hardening into fact through repetition. We have watched the same process run in reverse in Vienna, where a portrait sat in a storeroom under six successive attributions before a curator proposed Bernini, and in Madrid, where a Velazquez portrait of Olivares spent generations as a studio copy. The object does not change. The label is the asset, and the label is the thing that was always guesswork.
What has actually been established, and what has not
This part deserves more care than the headlines gave it.
What exists: the support of the Museo del Prado and of Manuela Mena, the museum’s former head of conservation for eighteenth-century painting and Goya, and the most consequential single voice on Goya attributions alive. Mena’s weight here comes from her refusals rather than her endorsements — Spanish press accounts put her lifetime tally at roughly a dozen paintings accepted out of hundreds examined. A yes from her is not a routine event. Both this canvas and its pendant, a San Francisco in another collection, have been declared Bien de Interes Cultural, which bars them from leaving Spain. The Art Newspaper reports the Prado’s Jose Manuel Matilla and the Wallace Collection’s Xavier Bray, a specialist in Goya’s religious work, among those involved.
What does not exist, yet: the technical study. No published imaging, no pigment analysis, no cross-sections, no report. That work is expected during the run-up to the 2028 exhibition. Until then the case is connoisseurship — expert eyes on brushwork, anatomy and the rendering of hands — supported by the most authoritative eyes available, but not corroborated by the laboratory. The ministry’s everything seems to indicate is not diplomatic softening. It is an accurate description of where the evidence stands.
Two further things a reader should weigh. The first is incentive: the painting belongs to a dealer, and a reattribution from Bayeu to Goya moves an object between price categories that are not adjacent. The export ban blunts this considerably — a Bien de Interes Cultural cannot be sold abroad, which removes the deepest part of the market — but it does not erase it, and it is better said than left implied. The second is that connoisseurship has a real record of being right well before the instruments agree, and also a real record of being confidently wrong. Both happen. The honest position today is that this is a strongly supported proposal with an unusual weight of expert opinion behind it and an unfinished evidentiary file, and that 2028 is when the file closes. If it holds, the interesting question will not be whether a machine could have caught it sooner — a question we have put to algorithmic authentication before, with mixed answers — but why nobody looked for two hundred years.
The friar on the stone
The subject is worth a moment, because it is not incidental to the dating. Pascual Bailon was a sixteenth-century Aragonese shepherd who became a Franciscan lay brother, was canonised in 1690, and is the patron saint of the Eucharist — the saint of a man who kept seeing the host and could not explain it. The painting shows him at the moment of the vision, with the stone from which he was said to have drawn water for the poor.
An Aragonese saint, painted by an Aragonese painter, in the years he was being told by Aragon that his way of painting was unacceptable. It is tempting to read too much into that, and worth resisting: devotional subjects were assigned, not chosen, and the commission is undocumented. But the choice of Pascual Bailon places the picture squarely in the world of the Franciscan commissions Goya was taking in 1781, which is the same world as the Bernardine altarpiece, and that is a coherent fit rather than a coincidence.
What is striking is the register. The enormous public Goya of these years is theatre — crowds, columns, a king in the audience. This is a man alone in a small frame, in the dark, doing the one thing he is known for. The gap between the two is the gap the attribution has to cross, and it is also, as it happens, the direction Goya’s whole career was eventually going to travel.
Where to see it
The canvas is being unveiled in the centenary exhibition of Artur Ramon Art, the Barcelona firm founded in 1926 and now in its fifth generation, which opens on 27 October 2026 and runs to 5 February 2027. Thirty-seven works, with Picasso, Zurbaran, Ribera, Warhol and Barcelo alongside it — a dealer’s hundred-year highlight reel, with the Goya as its centrepiece.
After that, the Prado, in 2028, for the bicentenary of Goya’s death in 1828. The exhibition is being curated by Gudrun Maurer and is built around his religious subjects and his anticlerical ones, which is the right room for this picture to be tested in: hung among secure works of the same years, in front of the people most motivated to find the flaw. That is where attributions are made and unmade, and it is a better verdict than any press release. Provenance, in the end, is the only part of a painting that can be argued with.
One last thing to sit with. If the attribution stands, a picture by the most important Spanish painter since Velazquez spent two centuries catalogued as the work of the relative who told him his brushwork was too loose — and sold, as recently as 2018, on the strength of that relative’s name. There is no better illustration of how little of an artwork’s value lives in the artwork.
Frequently asked questions
What is the newly discovered Goya painting? El milagro de la aparicion de la eucaristia a san Pascual Bailon, oil on canvas, 51.5 by 38.5 centimetres, dated to around 1780. The Barcelona dealer Artur Ramon bought it at a Madrid auction in 2018 as a Francisco Bayeu and reconsidered it during the 2020 lockdown. The Prado and the Goya specialist Manuela Mena support the reattribution; Spain’s culture ministry says everything seems to indicate it is an unknown Goya of about 1780.
Where can you see the rediscovered Goya? At Artur Ramon Art in Barcelona from 27 October 2026 to 5 February 2027, in a 37-work centenary exhibition that also includes Picasso, Zurbaran, Ribera, Warhol and Barcelo. It is then due at the Museo del Prado for the 2028 show marking the bicentenary of Goya’s death, curated by Gudrun Maurer around his religious and anticlerical subjects.
Has the Goya attribution been proved? Not technically. It currently rests on connoisseurship and on the Prado’s and Manuela Mena’s support. The full technical investigation — imaging and pigment analysis — has not been completed and is expected while the 2028 exhibition is prepared. The culture ministry’s wording is deliberately provisional.