Fabricated Provenance: The Forger's Real Craft

The picture must survive the second look; the paperwork must survive the first

A forged painting is examined by a specialist who is looking at it because he already believes the story. A forged provenance is what produced that belief. This is why the highest-yield forgery operations of the last four decades have been documentary operations with a painter attached, rather than the reverse.

The clearest demonstration is the case tried in London in 1998 and 1999. John Myatt, a painter and former art teacher, produced roughly two hundred works in the manner of Chagall, Dubuffet, Bissiere, de Stael, Le Corbusier, Giacometti, Sutherland and Nicholson. John Drewe supplied their histories. Drewe obtained at least one million pounds from sales, with later reporting putting his takings at about 1.8 million; Myatt, who painted them, received something in the region of 100,000 pounds.

That split is the whole argument in one line. The market was not paying for the pictures. It was paying for the histories, and the man who made the histories took the money. Myatt was sentenced on 13 February 1999 to one year and served four months. Drewe received six years. A third defendant, Daniel Stoakes, was acquitted of conspiracy.

About sixty of the two hundred works were recovered. The rest remain unaccounted for, and they carry documentation that was designed to be verified.

Drewe did not forge documents; he forged archives

The ordinary forger produces a certificate and hopes it is believed. Drewe produced the record against which a certificate would be checked. From 1989 he obtained access to the archives of the Institute of Contemporary Arts, and subsequently to the Tate archive, to the National Art Library at the Victoria and Albert Museum, and to the records of commercial galleries including Hanover and Ohana. He inserted photographs of Myatt's paintings into gallery records. He replaced pages in old exhibition catalogues and added new ones so that the forgeries appeared in the historical literature.

Access was bought. He made a donation of 20,000 pounds to the Tate, and his name duly appeared in the gallery's annual report for 1992 to 1994 among those thanked for their gifts. The donation is the most instructive detail in the case, because it was not a bribe and nobody at the institution did anything improper. It was a purchase of the standing that comes with being a benefactor, and standing is what opens a reading room.

The consequence is that a diligent buyer performing exactly the right check would confirm the fake. The provenance was not plausible; it was documented, in the institution a specialist would consult, in a catalogue with the correct binding and the correct paper.

Drewe was caught not through the archives but through a person. His former partner Bat-Sheva Goudsmid found incriminating material after he left in 1995 and took it to the police and to the Tate. The trial ran for five months from September 1998. The institutions then faced a second problem with no clean solution: identifying and removing insertions from archives that researchers had already used and cited.

The 1933 to 1945 gap is the forger's preferred address

Every provenance needs an explanation for silence. The strongest available explanation in twentieth-century European art is the Nazi period, because in that period silence is the norm: collections were sold under duress, dispersed, looted, hidden, destroyed, and their owners scattered or murdered. A picture with no exhibition history and no photographs is, in that frame, unremarkable.

Wolfgang Beltracchi built his operation on exactly this. The invented collections attributed to Werner Jaegers and to Johann Wilhelm Knops, both in fact grandfathers of members of the group, were said to have been formed by purchase from the Jewish dealer Alfred Flechtheim at low prices in 1933, shortly before Flechtheim fled Germany. Labels bearing a caricature of Flechtheim and the words for the Flechtheim collection were manufactured, stained with tea and coffee to build a convincing patina, and applied to the backs of the works. A photograph was staged using an old roll-film camera and pre-war photographic paper, showing Helene Beltracchi dressed as her own grandmother in a room hung with the forgeries.

Each element is checkable and none of it was checked. Flechtheim's own records exist. The Jaegers family history was traceable. The photographic paper could have been examined. The reason none of that happened is that the story was emotionally coherent, and coherence reliably substitutes for verification.

This is the practical warning for anyone buying German or Austrian modernism, and it runs against instinct. A wartime loss narrative should raise the standard of documentary proof required, not lower it, precisely because it is the story that most conveniently explains why proof is missing.

The document betrays itself before the picture does

Fabricated papers fail on internal consistency, and they fail in a characteristic way: the forger controls the content but not the incidental apparatus of the period he is imitating.

The J. Paul Getty Museum's kouros is the standard teaching case. It was offered by the Basel dealer Gianfranco Becchina, loaned for study in 1983, and bought in 1985; the price is given as 9.5 million dollars in one scholarly database and as 10 million in tertiary accounts, and the discrepancy is unresolved. The supporting documents traced it to the Geneva collection of Dr Jean Lauffenberger, who was said to have acquired it from a Greek dealer in 1930, with no archaeological findspot recorded.

The papers were subsequently shown to be fabrications, and the demonstrations were arithmetical rather than art-historical. A letter attributed to the classical archaeologist Ernst Langlotz and dated 1952 carries a postal code that did not come into use until 1972. A letter dated 1955 refers to a bank account that was not opened until 1963. Neither finding required an opinion about Greek sculpture.

The object's own status has never been resolved. A colloquium of nineteen invited specialists in Athens in 1992 reached no consensus. Stanley Margolis argued that the de-dolomitisation of the marble surface could only have occurred naturally over centuries; Miriam Kastner subsequently produced the same effect in a laboratory. Jeffrey Spier had published in 1990 on a torso known to be a forgery and carved from the same dolomitic marble. The Getty label eventually read "Greek, about 530 B.C., or modern forgery", and the director Timothy Potts confirmed that the work was taken off display in 2018.

The kouros is instructive because the documents were settled long before the object was. Two clerical anachronisms disposed of the provenance permanently. Four decades of scientific and stylistic argument have still not disposed of the sculpture.

The hardest case: when the certificate is genuine

Every technique above assumes the paperwork is false. Ely Sakhai built a business on the opposite premise. Operating a New York gallery, he bought genuine but modest Impressionist and Post-Impressionist works, with authentic certificates, and had copies painted on the upper floor of his premises. The copies were sold with the original documentation, largely to buyers in Asia who relied on the papers rather than commissioning independent European expertise. The genuine originals were then re-certified and sold separately.

The scheme collapsed on a coincidence of scheduling. In May 2000 both Christie's and Sotheby's offered a work titled Vase de Fleurs, also known as Lilas, given to Gauguin. Sylvie Crussard at the Wildenstein Institute established that the version at Christie's was the copy. Sakhai received 310,000 dollars from the sale at Sotheby's, and the Federal Bureau of Investigation traced it. He was charged in 2004, pleaded guilty, and in July 2005 was sentenced to 41 months in prison, fined 12.5 million dollars and required to forfeit eleven works.

No examination of the paperwork would have caught this, because the paperwork was real. The only defence is the physical link between document and object: the certificate must describe dimensions, condition, inscriptions and marks specific enough that it cannot be transferred to a second canvas, and a buyer must confirm that the object in front of him is the object the certificate describes rather than an object of the same subject.

The same principle answers a more recent case. The pictures seized by the FBI from the Orlando Museum of Art in June 2022, exhibited as works by Jean-Michel Basquiat said to have come from a storage locker, were the subject of a plea agreement in which the Los Angeles auctioneer Michael Barzman admitted that he and an associate identified only as J.F. had made them, on cardboard, in as little as five and at most thirty minutes each, then aged them outdoors. He was sentenced on 18 August 2023 to 500 hours of community service, three years of probation and a fine of 500 dollars. The museum's director, Aaron De Groft, had been dismissed in 2022. The provenance in that case was a single unverifiable story about a dead man's storage unit, and it carried a museum exhibition.