The Loan Register as Provenance Evidence

A loan is a dated, witnessed statement of possession

Provenance research is the reconstruction of a chain of custody from fragments, and the fragments are unequal. A dealer's stock book is strong. An auction catalogue is strong and dated. A family recollection is weak. The loan record sits near the top of this hierarchy and is used far less than it should be, because it does something no other document does routinely: it places a specific object, in a specific city, on a specific range of dates, at the request of an institution that had to identify the lender in order to insure the thing.

An exhibition catalogue formalises the same evidence. The catalogue entry names the lender, either directly or as a numbered anonymous collection, and the catalogue is printed, dated and deposited in libraries. Where a work appears in three exhibitions across forty years, the exhibition history alone establishes three anchor points that no later reconstruction can quietly move. Where a work is absent from an exhibition in which it should obviously have appeared, that absence is itself evidence, and it is the kind of evidence that has repeatedly turned out to matter.

The reason exhibition history is under-used is prosaic. Auction records have been aggregated, digitised and made searchable by commercial and scholarly projects. Exhibition catalogues have not. They sit on library shelves, one institution at a time, with lender information buried in a checklist at the back, and the researcher who wants them has to know which show to look for before they can look.

British law turned the loan register into a published provenance document

Part 6 of the Tribunals, Courts and Enforcement Act 2007 gives cultural objects lent from abroad to approved British museums immunity from seizure. Section 134 sets five conditions: the object is normally kept outside the United Kingdom, it is not owned by a UK resident, its import complies with UK law, it is displayed in a temporary exhibition at an approved institution, and the museum has published the required information about it. Protection runs for twelve months, extended where the object requires repair, conservation or restoration.

The publication condition is the part that produced a research resource. The Protection of Cultural Objects on Loan (Publication and Provision of Information) Regulations 2008 specify what must appear: the name and address of the lender; a description of the object with its type and materials; the artist; dimensions; date of creation; place of creation; identifying marks; a photograph where the object was made before 1946 and acquired after 1932; provenance; and a statement indicating whether or not the borrowing institution possesses a complete history of ownership from the beginning of 1933 to the end of 1945. That information must be published for at least four consecutive weeks ending on the day before the object enters the United Kingdom.

Section 136 gives the regime teeth. An approving authority must take account of an institution's procedures for establishing the provenance and ownership of objects, and may withdraw approval where those procedures are inadequate. In Scotland the application goes to Scottish Ministers at least eighteen weeks before the object arrives, and the institution must show its due diligence follows the 2005 guidance from the Department for Culture, Media and Sport. The purpose of the four-week publication window is to let a claimant see the object coming and object before it lands.

For a researcher, the consequence is that every immunity listing on a British museum's website is a dated, structured provenance statement made by an institution under a statutory scheme, including an explicit admission of whether the 1933 to 1945 record is complete. Very few provenance sources are that candid by law.

Portrait of Wally: the exhibition that produced the evidence

The case that made every registrar in the world read the immunity statutes began with a loan. Lea Bondi, a Viennese Jewish gallery owner, bought Egon Schiele's Portrait of Wally in 1925. After the Anschluss she sold her gallery to Friedrich Welz on 13 March 1938 and fled to England; the transfer of the Schiele, which hung in her apartment rather than the gallery, has always been the disputed link. American forces seized Welz's property in May 1945. In December 1947 the restitution division transferred fourteen paintings including the Schiele to the Austrian Federal Office for the Preservation of Historical Monuments, which delivered it in May 1950 to the heirs of a different collector, Dr Rieger, who had died at Theresienstadt. Late in 1950 the Rieger heirs sold to the Austrian Gallery Belvedere. In September 1954 Rudolf Leopold acquired the painting from the Belvedere without asking for ownership documentation. Bondi died in 1969 without recovering it. In August 1994 Leopold sold his collection to the newly created Leopold Museum, and in 1995 that museum published a catalogue setting out the provenance in full.

The painting travelled to the Museum of Modern Art in New York for an exhibition running from late 1997 into 1998. Three days after the show closed, the New York County District Attorney issued a subpoena; the New York Court of Appeals quashed it on the strength of state anti-seizure protection. The following day a federal magistrate issued a seizure warrant and US Customs seized the painting under the National Stolen Property Act, on the theory that it had been stolen from Bondi by Welz and unlawfully imported. Litigation ran for twelve years. On 19 July 2010 a stipulation and order settled it: the Leopold Museum paid 19 million dollars to the Bondi estate, lent the painting for three weeks to the Museum of Jewish Heritage in New York, and undertook to display permanent signage recording the court's 2009 conclusion that the work had been taken from Bondi in Vienna in the late 1930s.

The chain of documents that decided the case was almost entirely publication and loan material: a 1995 museum catalogue that published the provenance, a 1997 loan that brought the object into a jurisdiction where a claim could be filed, and an exhibition checklist that fixed the dates. The immunity statutes on both sides of the Atlantic were tightened in the years afterwards, and the price of that protection is exactly the disclosure the 2008 British regulations demand.

The market has been indexed; the exhibitions have not

The imbalance in provenance research tools is stark and it is worth stating in numbers. The Getty Research Institute's Provenance Index covers German sales of 1930 to 1945 through more than 1,600 auction catalogues yielding roughly 236,000 art-sale records, assembled with the Kunstbibliothek of the Staatliche Museen zu Berlin and the University Library at Heidelberg, which digitised close to 200,000 pages of catalogues and made them searchable through optical character recognition.

Confiscation records have had similar treatment. The Forschungsstelle Entartete Kunst at the Freie Universitat Berlin maintains a searchable inventory of works seized from German museums in the 1937 campaign, funded until 2015 by the Ferdinand Moller Foundation and since by the Federal Government Commissioner for Culture and the Media. Each record can be interrogated by artist, title, dating, object type and the fact of confiscation, which makes it a provenance instrument rather than a bibliography.

Nothing comparable exists for exhibitions. There is no single index that will tell you every show a given picture appeared in between 1880 and 1939, and the catalogues that would answer the question are scattered across national art libraries. The practical result is that an auction appearance is found in minutes and an exhibition appearance is found in weeks, which distorts what provenance researchers report and therefore what the market believes.

How to work an exhibition history properly

The method is unglamorous. Start with the exhibition list in the catalogue raisonne or the museum's own record, and treat it as a set of claims to verify rather than as a finding. Obtain the printed catalogue for each show and read the checklist, not the plates: the lender line is what you want, and it is often more specific than the summary in the raisonne. Note the exact dates of the exhibition, because those are the dates you can defend.

Then look for the transitions. A lender identified as a private collection in one show and by name in the next has usually not changed; someone has simply relaxed. A work that moves from a named lender to an anonymous one is worth a second look, particularly across 1933 to 1945. A gap in a run of appearances, where the object was shown in 1928 and 1937 but not in the obvious 1932 survey, is a question, and often the question is where it was.

Finally, cross-check against the loan disclosures. For any object that has entered the United Kingdom under the immunity scheme since 2008, a published statement exists recording the lender, the provenance and whether the 1933 to 1945 chain is complete. That statement was made by an institution to a regulator, and it is one of the very few provenance assertions in the field that carries an administrative consequence if it is wrong.