The Forgery That Shocked the Art World

As art buyers flipped through art dealer David Croal Thomson’s 1925 catalog, lot number 24 may have stopped them in their tracks.

There, in a grainy black-and-white photo, was a preparatory oil study (an early version of the final painting) for Jean-François Millet’s world-famous Man with a Hoe.

“This is the most magistral and monumental of all the works of the great artist,” the catalog breathlessly reminded any readers who had somehow forgotten Millet’s seminal work, a dignified portrayal of an exhausted farmer working a rocky field with a heavy-duty hoe.

This early version of the composition would have been considered an incredible (and valuable) find. Millet’s reputation had grown after his death in 1875, and by the turn of the century, his sensitive paintings of laboring French peasants had found a global audience.

Thomson, a Scotsman based in London, was an influential promoter of Millet’s work in the UK and North America, and this purported study for Man with a Hoe soon entered a private collection in Montreal.

Just five years after Thomson published the painting, however, a shocking revelation stunned the art world: the “study” was in fact a forgery, along with many other paintings and drawings supposedly by Millet. The masterminds behind the scheme: Millet’s own grandson, Jean-Charles Millet, and his associate Paul Cazot, a small-time painter.

The fake is on display in the new Getty Center exhibition Reckoning with Millet’s “Man with a Hoe,” which charts the tumultuous public life of Millet’s iconic work. The story of how Jean-Charles capitalized on his grandfather’s name to helm one of the biggest art fraud schemes of the 20th century made for sensational headlines in the 1930s, but few today know about this remarkable chapter in art history. Even Millet scholars are largely silent about it.

“By the mid-1920s, David Croal Thomson had been a recognized expert in Barbizon painting for decades, so it’s not at all surprising that collectors bought fakes from him. They would have implicitly trusted him, especially since the source in this case was the artist’s own family. What better provenance could there be?” said Scott Allan, the exhibition’s curator. “We have the unique opportunity of presenting a historic fake with the original Man with a Hoe. It’s not been done before and should make for a really fascinating comparison.”

Creating a Forgery Factory

Born 17 years after Millet’s death, Jean-Charles shared his grandfather’s interest in art and even spent time in Barbizon, the small village near the forest of Fontainebleau that Millet had helped to make famous. But Jean-Charles’s own Barbizon-inspired paintings were highly derivative, and he never found true success as an artist. To make extra cash, in the early 1920s he sold a few original Millet drawings to a Monsieur Douhin, a businessman who bought Millet’s old house in Barbizon and turned it into a museum and tourist attraction celebrating the artist.

The seeds of the scheme began to sprout. With Millet’s real drawings gone, Jean-Charles began producing imitations of them and passing them off as originals. Making the job even easier: he owned the studio stamp that was used to mark the works remaining in Millet’s studio after his death, prior to their dispersal in the artist’s estate sale. “So he’s able to authenticate fake drawings quite easily—it’s got the studio stamp on it, so it must be original,” Allan said.

The scheme escalated when Jean-Charles teamed up with Cazot. After seeing some of his work in a shop window in Paris and noting a resemblance to Millet’s style, Jean-Charles recruited him to paint forgeries. Cazot would buy old canvases at flea markets, scrape off the paint, and produce new “Millets,” often masquerading as studies for or variants of his famous compositions.

Jean-Charles then applied a fake signature or the studio stamp on the new paintings and, to help market them, concocted authenticating documents, sometimes even using old family stationery. Over the course of the 1920s he and Cazot reportedly created some 40 fake Millet paintings and around 200 fake Millet drawings.

Jean-Charles sold several of these fakes to Douhin for his Millet Museum (including another version of Man with a Hoe, which Douhin persisted in claiming was authentic), and he passed off many more to art dealers, particularly Thomson, whose customers were primarily British, American, and Canadian. It’s impossible to know if Thomson was aware that he was selling forgeries, since he died in 1930, before the Millet-Cazot case went to trial. “Thomson may have been an unwitting accomplice who was duped, but he nevertheless profited from moving the fakes,” Allan said.

“The English Like Fakes Better”

Jean-Charles and Cazot’s scheme came crashing down when a buyer became wary of the authenticity of a Millet painting he had purchased and began to dig deeper, eventually taking it to an art expert, who declared it a fake. In addition, “the art dealers began to grow a trifle suspicious,” a 1930 newspaper article reported, due to the “veritable flood” of works by famous artists on the market. When police tracked the duo down, they both quickly confessed.

On trial in 1935, Jean-Charles specifically admitted to forging versions of Man with a Hoe and other Millets, seemingly relishing how he had swindled unsophisticated (read: non-French) buyers and dealers and netted profits his grandfather had never seen during his life.

“I am half English and I know how gullible the British are,” Jean-Charles bragged in an interview. “I repeatedly went to England with originals and copies and invariably returned with the originals—the English like fakes better.”

For Jean-Charles, some of the appeal in his scheme appeared to lie in sticking it to the art world’s elitist gatekeepers and self-styled experts.

“Besides the purely venal interest in making money, it was a way to exact revenge for his grandfather, who struggled financially for most of his life before others came along and reaped fortunes from his work; at least that’s how Jean-Charles justified himself,” Allan said. “But as a struggling artist, he must have also felt the oppressive weight of his grandfather’s legacy, and the bold forgery scheme, however shameful, was one way of slipping out from under Millet’s shadow.”

Jean-Charles and Cazot were both sentenced to short prison terms and ordered to pay fines. As for the fakes they created, Allan suspects many have been destroyed, while others must still be sitting in private collections and probably museum storerooms as well.

Comparing the Real and the Fake

While preparing for the Getty exhibition, Allan resolved to track down the forgery purchased by the Canadian collector, which turned out to be relatively simple, as the painting was still with the collector’s family. The current owner, who was aware of the painting’s dubious attribution to Millet, agreed to lend it for the exhibition. Once it arrived at Getty, Allan and paintings conservator Devi Ormond had it x-rayed and examined under infrared so they could compare it to the original painting and its related compositional drawing (both in Getty’s permanent collection). This helped them dig deeper into the differences between the real and fake versions.

Besides the fact that Jean-Charles identified the Thomson version as a fake during the trial, there are a number of clues in the painting itself that cast suspicion on its attribution to Millet and its supposed function as a preparatory study.

The first red flag: the mere existence of a painted study. Millet generally did not make preparatory oil sketches for his major Salon paintings at this point in his career. You’d expect to find some preparatory drawings by Millet, but that’s about it.

Second: the figure’s pose. We know from the compositional drawing and the x-ray of the Getty painting that Millet made changes to the figure’s pose while he was working on the final canvas (for example, he adjusted the angle of the man’s left leg and widened the space between his clogs). In the Cazot version, the figure’s firmly delineated pose is already completely resolved, as can clearly be seen in the infrared image.

“To me this strongly suggests that the figure was drawn after the figure in the Getty painting. This is definitely not a painting in which the artist is still figuring things out,” Allan said.

Similarly, the fake version carefully follows minor details, like some of the shadowy folds in the figure’s pants and shirt, something you wouldn’t necessarily expect to see in a roughly executed preparatory sketch.

The compositional format is also strange. The forgery was produced on a panel, vertical in orientation, while the Getty support is canvas, obviously horizontal. That wouldn’t necessarily be strange if the “study” were focused more closely on the central figure or if it were a variant composition with a different setting (and Millet did in fact sometimes make such variants). But here the artist has simply taken the background details of the final painting and awkwardly compressed them into the narrower format.

In addition to these visual and technical clues, the Thomson version has zero documentation prior to his 1925 catalog, which appeared more than 60 years after the painting was supposedly created. Given Millet’s prominence in the late 19th century, one could reasonably assume that an autographed oil study for Man with a Hoe would have left more of trace—appearing, for instance, at the artist’s 1875 studio sale or some other auction in the years following.

Allan hopes that the exhibition, in presenting both the original Man with a Hoe and a fake, will encourage visitors to look more closely for themselves at the two works and develop a deeper appreciation for Millet’s qualities as an artist, while also gaining a better understanding of his tremendous popularity and high stature in the late 1800s and early 1900s.

“Forgery scandals certainly make for sensational stories and fun reading,” Allan conceded. “But the forgeries themselves are also unique and valuable tools for connoisseurship, teaching us how to see the originals better.”

Reckoning with Millet’s Man with a Hoe is on view September 12–December 10, 2023, at the Getty Center.