Tag: Painting

A data science firm claims that a painting rescued from a garage sale bin in Minnesota for less than $50 could be a lost work by Vincent van Gogh—potentially worth $15 million. Newly attributed to the famed Post-Impressionist painter by art tech firm LMI Group, the artwork, titled Elimar, has undergone extensive scientific and stylistic analysis. The discovery, if confirmed, would mark a major addition to Van Gogh’s known oeuvre and reignite debates about the role of technology in art authentication. The Van Gogh Museum, however, refutes the claim.

In 2018, the buyer of Elimar hoped they might have a priceless masterpiece on their hands and submitted an inquiry to the Van Gogh Museum in Amsterdam. The museum replied in 2019, stating that, having examined the information provided, “based on stylistic features” the work could not be attributed to the artist. That same year, the owner sold the painting for an undisclosed sum to the New York-based data science company LMI Group, which was established the previous year.

Now, the company claims the “orphaned” painting Elimar was indeed made by Van Gogh in 1889, towards the end of his life while he was living at the Saint-Paul asylum in the south of France. The word “Elimar” scrawled in the lower-right hand side of the canvas refers to the name of a character from Hans Christian Andersen’s 1848 novel The Two Baronesses. However, LMI believes the painting is the artist’s interpretation of a similar painting, Portrait of Niels Gaihede, by the Danish artist Michael Ancher.

“The analysis conducted on this distinctive painting provides fresh insight into the oeuvre of Van Gogh, particularly as it relates to his practice of reinterpreting works by other artists,” said Maxwell L. Anderson, an art historian and the chief operating officer of LMI Group.

The principal authority on the authenticity of paintings attributed to Van Gogh is the Van Gogh Museum in Amsterdam. A spokesperson for the institution said that, after considering the new information in the LMI Group’s report, “we maintain our view that this is not an authentic painting by Vincent van Gogh.”

Proof in the Data?

New scientific approaches to art authentication have gained popularity over the past decade but have yet to become mainstream within the trade. LMI Group says it teamed up with various experts, from chemists to curators, to devise a “data-based” approach to analyzing the artwork. Costing over $30,000, according to the Wall Street Journal, this approach merged traditional connoisseurship—which considers formal analysis, provenance, and historical context—with newer scientific methods. A lengthy report detailing the findings along with extensive art historical context intended to situate the painting within Van Gogh’s oeuvre has been published online.

In one traditional visual analysis, the canvas was compared to known authentic Van Gogh paintings, with which it apparently shares significant similarities, most notably the decision to show the fisherman in a three-quarter view, matching Van Gogh’s four self-portraits from 1889. LMI says that in the final year of his life, a hospitalized Van Gogh stopped painting with his trademark vibrant colors and opted instead for a more muted palette, like that used for Elimar. In this same period he also did several “translations” of other artists’ works found in books.

Another test, this time of the pigment used in the painting, dated it to the 19th-century. It also found that the surface had been given a temporary glaze of egg-white, which Van Gogh was known to apply to protect rolled canvases.

A human hair found embedded in the surface of the painting was submitted to a DNA analysis. It was found to have belonged to a man, with the investigating scientists “observing” that it appeared to be red in color, according to the report.

Finally, a more experimental analysis method involved the close “mathematical” comparison of the letters in “Elimar” in the bottom-right hand corner with other words that have appeared on known Van Gogh paintings. In particular, the word was compared to the word “Emile Zola” on Van Gogh’s Still Life With Bible (1885). This apparently showed significant similarities in certain characteristics of the letters, including the stroke length and width, the bounding size, and the angle. In some cases, this resemblance was measured to be 94 percent. It is worth noting that Van Gogh didn’t sign all his canvases, excusing the lack of signature on Elimar.

The company has also suggested that Elimar could be interpreted as being autobiographical. A press release announcing the discovery describes how the artist “reimagines himself as an older, wiser man depicted against the serene palette-knife-sculptured sky and smooth expanse of the water, evoking Van Gogh’s lifelong personal interest with life at sea.” It said the composition evokes The Poor Fisherman (1881) by Pierre Puvis de Chavannes, which Van Gogh allegedly admired.

“Through Elimar, Van Gogh creates a form of spiritual self-portrait, allowing viewers to see the painter as he wished to be remembered,” Anderson said.

Authentication Debates

Previously the director of the Whitney Museum of American Art between 1998 and 2003, Anderson founded LMI Group with former lawyer and LMI Group president Lawrence M. Shindell and the company’s CFO, ex-Wall Street investor Steven P. Novak. The company is inviting Van Gogh scholars and dealers to book a private appointment to view the work in person.

“LMI Group’s data-based approach to verifying authorship of this painting represents a new standard of confidence for bringing to light unknown or forgotten works by important artists,” said Shindell. “By integrating science and technology with traditional tools of connoisseurship, historical context, formal analysis, and provenance research, we aim both to expand and tailor the resources available for art authentication based on the unique properties of the works under our care.”

The firm did not provide information on how Elimar ended up at a garage sale in Minnesota, but alleges in the report that Van Gogh lost many works by giving them away to friends or being neglectful in his lifetime. It did not provide any provenance prior to 2016, which would indicate that the alleged Van Gogh painting was completely unknown to scholars until now since catalogues raisonnés usually list an artist’s lost works, whether destroyed or missing.

The Van Gogh Museum has rejected independently authenticated works before. In 2016, the Van Gogh expert Bogomila Welsh-Ovcharov authenticated drawings that she said were from one of the artist’s 1888 sketchbooks, publishing the works in her book Vincent van Gogh: The Lost Arles Sketchbook. Their authenticity was later denounced by the Van Gogh Museum.

Like many artists’ foundations, the museum has previously been threatened with legal action over past disputes and now no longer accepts authentication requests from individual members of the public. Hopefuls must already have the backing of serious specialists before their claim will be considered, reducing the number of annual submissions from several hundred to around 40.

If the painting is real, it is sure to drum up excitement among collectors and everyday art fans alike. A recent exhibition of works from the end of Van Gogh’s life at the National Gallery in London recently broke the museum’s previous attendance records by attracting 334,589 visitors.

The use of new, data-driven methods to authenticate art has been adopted by several start-up tech companies in recent years. Late last year, the Swiss company Art Recognition used A.I. tools to authenticate three artworks offered for sale by Germann Auction House in Zurich. Most notably, it was the only certification listed for an untitled, undated watercolor attributed to the German Expressionist artist Marianne von Werefkin, which sold for CHF 15,000 ($17,000) over a high estimate of $9,300 on November 25.

This story was updated on Friday, January 31, at 6:24 a.m. ET.