Has AI just solved the mystery of one of El Greco’s most enigmatic paintings?

For years, art historians believed that The Baptism of Christ was likely painted by El Greco with the assistance of other artists. But new research suggests otherwise.

“The Baptism of Christ” is one of the most mysterious works by the Spanish Renaissance painter El Greco. The oil painting depicts an imposing John the Baptist pouring water over the head of an even larger, almost resplendent Jesus; In the background, God, angels and cherubs watch from heaven in an ecstatic frenzy.

It’s a vibrant and captivating image. Art historians believe that the work it was unfinished when El Greco died in 1614, and which was completed by the painter’s son, Jorge Manuelwith the help of other apprentices from El Greco’s studio.

However, a new research suggests otherwise. Using Artificial Intelligence, researchers analyzed The Baptism of Christ at the microscopic level, looking for patterns in the texture of the paint with resolution equivalent to that of a single brush bristle.

The results, published on Friday in the journal Science Advancesindicates that El Greco will have painted most of it of “The Baptism of Christ — but some experts warn that more research is needed.

Although not conclusive, the study comes to “shuffle things” with regard to the hypothesis of multiple painters and raises questions about the work, which justify a more in-depth analysis, he states Andrew Van Hornresearcher at Purdue University and lead author of the article, cited by .

During the Renaissance, great master painters habitually employed apprentices to work alongside them while they learned the trade.

These assistants could grind pigments, stretch canvases and even help fill certain details in the paintings. Without detailed work records, determining exactly who painted what in a given work from a master’s studio can be a challenge. real challengeexplica Van Horn.

Over the decades, studies by art historians have been based on brushstroke styles and other visual cues and texture to identify whether a painting can be entirely attributed to a single artist or whether it should be attributed to your studio in a more general way.

But this detective work It is sometimes problematic and can lead to priceless works being misattributed, ignored or disputed. This is where Artificial intelligence could make a difference.

Van Horn and his colleagues created a model of machine learning trained with 25 paintings by nine art students. Next, the researchers showed the model two paintings by El Greco: “The Baptism of Christ” and “Christ on the Cross with Landscape”, which, unlike the first, was believed to have been entirely by El Greco.

As expected, artificial intelligence determined that “Christ on the Cross” was the work of a single artist. But when he analyzed The Baptismthe AI ​​detected an underlying link between segments of the painting that were believed to have been executed by different painters, says Van Horn. In other words, at the microscopic level, painting was more uniform than previously determined.

“What helps us is the fact that we can observe on a very fine scalewhich allows us to see things that you might not be able to see with the naked eye,” says Van Horn.

The conclusions, obtained in collaboration with art historians, suggest that The Baptism may have been largely created by El Grecoperhaps using different brushes than usual — or with hands affected by aging.

“Creating an artificial intelligence system capable of detecting the authorship of a painting is an incredibly complex problem,” he says. Mark Hamiltona researcher at MIT’s Computer Science and Artificial Intelligence Laboratory, who was not involved in the study.

“Artists can change styles while paintingcollaborating artists may attempt to imitate a master’s style, and conservation interventions and physical damage may affect measurements,” he adds.

“The work represents a good first step to establish El Greco’s authorship, but there is still a lot to be done”, he says Richard Taylorprofessor of physics, psychology and art at the University of Oregon, who was also not involved in the study.

“Although the analysis reveals a tightly interconnected set of communities within the painting, cannot definitively rule out none of the hypotheses”, he adds, highlighting that the work of students can be more varied than that of artists trained in the same studio and that the study has a small sample.

“A more exhaustive and complete analysis of many more paintings would be necessary to be able to say anything definitive.”

If confirmed, however, the results could change the way art historians view the end of El Greco’s life, as well as the value of the work he created during this period.

Van Horn is now waiting for the AI ​​tool to analyze different paintings from the same studio, looking for the styles of lesser-known individuals as a way of identifying and following the trajectory of these artists. But he makes a point of highlighting that the AI was not designed to replace art historians.

“Being able to track people, or a kind of ghost figurethrough different ateliers to his own studio, would be extraordinary”, he states.