In October 1940, five months after the German invasion of the Netherlands, ‘Woman with a Fan’, a Cubist painting by Jean Metzinger from 1913, was auctioned in Amsterdam. It had belonged to Emanuel Lewenstein, a Jewish collector who ran a sewing machine factory. No one today knows who offered it for sale or under what conditions. In 1959, it arrived at the Art Institute of Chicago as a gift from a collector couple. On 22 September 2026, the museum announced that it will return the work to Lewenstein’s heirs.
The museum’s reasoning goes beyond the case at hand. Its investigation did not clarify what occurred in 1940. The museum resolved the uncertainty through a criterion formulated by its head of provenance, Jacques Schuhmacher: in the absence of evidence to the contrary, it is appropriate to treat the sale as a consequence of Nazi persecution, whether as a confiscation or a sale under duress. Without knowing the circumstances, the museum decides that the uncertainty cannot count against the heirs of a persecuted family.
That same month, in two other places, the same lack of paperwork met with different answers.
On 2 September, the Toledo Museum of Art (Ohio) announced a partnership spanning more than twenty years with the Ethiopian Heritage Authority. The starting point was a collector, Bob McCarthy, who donated Ethiopian artefacts for repatriation only to discover they lacked provenance documentation. There was no proof of either illicit or legitimate export. Adam Levine, the museum’s director, described the issue as a gap in the timeline: millions of works acquired in good faith between the 1970 UNESCO Convention and the point in 2008 when that criterion was incorporated into the professional practices of the American museum sector, without meeting a standard that arrived later. The response was to suspend judgement: to investigate the pieces alongside Ethiopian specialists and decide their fate between the two institutions. Of the first five repatriations, only two fifteenth-century manuscripts have a definitive conclusion: they were recorded in Ethiopia after 1970 and probably left the country later. The other three are being returned without that determination.
The third response came from the market. Inlibris, a dealer in antiquarian books based in Vienna, included in its offering for Abu Dhabi (between 13 and 18 September) an 18th-century Ethiopian gospel linked to Emperor Tewodros II, with a guide price of 250,000 euros. Following protests, on 12 September, it offered Ethiopian institutions the opportunity to purchase it at a preferential price, provided the transaction was treated as a commercial purchase and not as an admission of any defect in its title or of any restitution claim. The Ethiopian Heritage Authority rejected those conditions on 15 September and demanded the suspension of the sale and the unconditional return of the manuscript. Two days later, the founder of Inlibris stated that he was not attempting to sell it at the fair; its withdrawal was subsequently reported. There is no record of it having been sold.
Three responses to a single absence. Chicago presumes in favour of the persecuted family and resolves the doubt against the museum itself. Toledo splits the doubt: it neither presumes looting nor rules it out, and shares custody while it investigates. Inlibris resolves it in its own favour: it acknowledges that the object matters to Ethiopia, but maintains that its title remains intact and that any return would require payment. In all three cases, the document that would settle the decisive point is missing. What changes is who bears the cost of that absence.
It appears in almost every case I have written about this year. In Daniel Druet’s lawsuit against Maurizio Cattelan’s gallery and publisher, authorship was presumed in favour of the signatory, and it fell to Druet to rebut that presumption. In the case of Robert Kneschke against LAION, the photographer had to prove that his reservation of rights was machine-readable in 2021. In Joe Morford’s case against Cattelan, he had to prove that Cattelan had seen his work. Rules of evidence are not a technical detail that comes after justice: they decide in advance who loses when the document is missing. And the document is almost always missing, because the history of works is written on papers that no one thought would ever need to be produced.
Chicago did not wait for evidence that may no longer be possible to reconstruct. It decided that its absence, in Amsterdam in 1940, could not be held against a persecuted family. This is not a general museum rule or a court ruling; it is a criterion for this specific case. Yet it demonstrates that the burden of doubt is a matter of choice.
Another work from the same family, Kandinsky’s ‘Das bunte Leben’, returned to its heirs in 2023, after a German advisory commission recommended its restitution and the bank that held it agreed. On 14 October, it will be auctioned at Christie’s, London. This is a legitimate outcome: restitution restores ownership, and what is done with the work is a decision for its owners. However, it closes the circle in a different way. Once the consequence of that doubt is resolved, the work returns to circulation, this time with documented provenance.
On the open conversation
This text compares three responses from September 2026 to the lack of provenance documentation: the restitution of a Metzinger by the Art Institute of Chicago, the partnership of the Toledo Museum of Art with the Ethiopian Heritage Authority, and the offer from Inlibris regarding an Ethiopian gospel. I do not claim that a specific confiscation has been proven in Chicago or an illicit export for all the Toledo pieces. In the case of the gospel, I report an ongoing dispute: the Ethiopian authority demands its restitution and the dealer defends the validity of its title. I am interested in how each institution decides who bears the burden of doubt when the document that would resolve the decisive point is missing, and I relate this to the rules of evidence in other cases I have written about. Should anyone wish to contribute from the fields of provenance research, heritage law, or the art market, this notebook remains open.
Sources
The Art Newspaper, ‘Art Institute of Chicago restitutes Metzinger painting sold in Nazi-occupied Amsterdam’, 22 September 2026
Art Institute of Chicago, Jean Metzinger, ‘Woman with a Fan’, 1913, inv. 1959.10
The Art Newspaper, ‘Toledo Museum of Art partners with Ethiopian Heritage Authority to find solutions for “objects in limbo”’, 4 September 2026. The institutional announcement is dated 2 September.
The Voice, ‘Antique dealer defends upcoming sale of looted ancient Bible belonging to former Emperor of Ethiopia’, 12 September 2026
Capital Ethiopia, ‘Ethiopia Rejects Commercial Buyback Offer for Looted 18th-Century Royal Gospel’, 15 September 2026
Al Jazeera, ‘Rare Ethiopian manuscript stolen by British forces triggers restitution row’, 17 September 2026
Beratende Kommission, recommendation regarding ‘Das bunte Leben’, agreed on 16 May and published on 13 June 2023
Christie’s, ‘Das bunte Leben (1907): a monumental turning point on Kandinsky’s road to abstraction’, 24 September 2026
Christie’s, ‘20th/21st Century: London Evening Sale’, 14 October 2026
Notes
Chicago: Schuhmacher’s criterion was formulated for this sale, whose consignor and circumstances are unknown; it does not appear as a published general museum policy. The investigation does not identify a specific confiscation.
Toledo: objects without documentation are treated as ‘in limbo’, with no evidence of illicit status or legitimate transfer. Only the two manuscripts have a specific conclusion regarding their departure after 1970.
Inlibris: the Ethiopian Heritage Authority maintains that the manuscript was looted; Inlibris defended its title. There is no record that the sale took place.
Kandinsky: the commission recommended restitution, which the owning bank accepted. The Christie’s catalogue does not identify the consignor.