What a Collector's Name Is Worth as Evidence

One family name, nine objects, three kinds of evidence

One family name appears in the provenance of nine records in this archive, across seven object classes. The objects have almost nothing else in common: a Flemish book of hours, a late Roman glass cup, a Vermeer, a fourteenth-century French manuscript, a reliquary, a silver ewer, a Boulle cabinet, a Sevres porcelain nef and a folio from a Persian Shahnameh. What the name is worth as evidence differs completely from one to the next, and the difference is not a matter of how famous the object is.

Read down the last column and the ranking has nothing to do with the collector. It has to do with whether anyone wrote the transaction down at the time, and whether the institution that ended up with the object preserved what was written.

The collection that catalogued itself

A collection that publishes its own catalogue converts a family name into a citable document, and the Vienna Rothschild collection did exactly that. The catalogue note for the Rothschild Prayerbook records that the manuscript was number 597 in Schestag's 1872 catalogue of Baron Anselm von Rothschild's art collection, that the London Hours now in the British Library was number 599 in the same catalogue, and that number 595, a prayerbook illuminated for Cardinal Albrecht von Brandenburg and now in the Getty, had been bought in 1868. That last date is what allows the acquisition of number 597 to be placed shortly after it, rather than guessed at.

The February 1906 inventory of Baron Nathaniel von Rothschild's estate goes further than an ownership claim. It records the Prayerbook as number 452, hanging in the Galerie of the palace in the Theresianumgasse, and valued at 150,000 Kronen against 80,000 for the Brandenburg book beside it, which fixes not only where the object was but what its owners thought it was worth relative to its neighbour.

Two later palace inventories left their numbers on the object itself: 927 on a label at the upper corner of the lower cover, and AR3390 in pencil on the front flyleaf. A provenance step that can be checked against a mark on the binding is a different class of evidence from a name in a list, and it is the reason this manuscript's ownership history survives the twentieth century intact while others in the same collection do not.

The name with no date attached

Cleveland's record for its Boulle cabinet names two consecutive owners and dates neither of them. The museum's own open access interface returns the whole provenance for accession 1949.539 as two entries: Baron Nathaniel de Rothschild, and Baron Alphose de Rothschild, Vienna (Rosenberg and Stiebel, Inc., New York). Both carry a null date field, neither carries a citation, and the second misspells the given name.

This is not a criticism of a museum that has published its provenance in machine-readable form, which most have not. It is an observation about what the record can and cannot support. A cabinet made around 1695 and acquired by Cleveland in 1949 has two and a half centuries of history compressed into two undated names, and the fullest published study of the piece, William Milliken's 1951 account in the museum's own Bulletin, begins at the same place. Nothing in the record distinguishes an ownership of forty years from one of four.

The consequence is specific and it is not academic. Which Nathaniel and which Alphonse are named decides whether this object was in Vienna in March 1938, and the record as published does not answer it. A researcher who wants to know cannot proceed from the museum's provenance; they have to reconstruct the family from outside it and then argue the identification, which is a different and far weaker kind of claim than reading a number off a binding.

The name the vendor supplied

The weakest form of the collector step is the one where the only authority for it is the person who was selling the object, and the Walters Art Museum says so on the page. Its record for the Sevres potpourri vase in the form of a ship, accession 48.559, opens the provenance with the collection of Sir Anthony Nathan de Rothschild, 1st Baronet, and marks it date and mode of acquisition unknown, with a footnote.

That footnote is the most useful sentence in the record. It states that the provenance is not confirmed and gives its source: the vendor Edwin Marriott Hodgkins's own 1927 sale catalogue of his Sevres porcelain, number 30, which described the vase as formerly in the collection of Sir Anthony de Rothschild. The museum then traces the documented chain it can stand behind, from Hodgkins in Paris by 1927, through A. Seligmann, Rey and Co. in New York by 1928, to Henry Walters's purchase in 1928 and his bequest in 1931.

A collector's name in a dealer's catalogue is a price argument before it is a historical claim, and it is worth remembering which direction the incentive runs. The Walters has done the one thing that keeps such a claim usable: it has published the assertion, named who made it, dated it, and declined to adopt it. A record that quietly promoted the same sentence into an unqualified provenance line would look stronger and be worth less.

The name that becomes the object's label

A collector's name can outlast the ownership it records and end up as the object's name in the scholarly literature. The British Museum's record for the Lycurgus Cup, registration 1958,1202.1, gives the acquisition as a purchase from Victor, 3rd Lord Rothschild, with a contribution from the body now called the Art Fund, and its acquisition note states that the cup was originally acquired by Baron Lionel Nathan de Rothschild, who died in 1879.

What the museum does not do is fill in the middle. It names the first owner and the seller nearly eighty years later, and leaves the holders in between unnamed, which is why this archive's own entry records that span as a documented family descent rather than as a list of owners it cannot support. The standard technical study of the object, published by Harden and Toynbee, is nonetheless titled after the family, so the name now travels with the cup through the bibliography while the ownership it describes remains only partly recorded.

That is a common and underrated hazard. A name repeated in citations acquires the look of established provenance without any new evidence entering the record, and the only defence is to go back to the acquisition note and see how much of the chain the holding institution actually claims.

What the rupture did to each tier

Three of these nine objects were taken from their owners between 1938 and 1940, and the time each took to come back tracked the paperwork rather than the loss. The Vienna palace holdings were appropriated in 1938. The Paris collections were seized from 1940 and processed by the Einsatzstab Reichsleiter Rosenberg, the confiscation agency that ran looted French collections through the Jeu de Paume.

The Louvre's own record for The Astronomer sets out the chain without softening it: seized in 1940 for the museum planned at Linz, recovered by the Allies in 1945, restituted to Baron Edouard in 1946, inherited by Guy de Rothschild, and acquired from him by dation en paiement of transfer duties in 1982 to 1983, which is how it came to the French State as inventory RF 1983 28 with no sale and therefore no price. The Hours of Jeanne d'Evreux was found at Neuschwanstein by an Allied Monuments, Fine Arts and Archives officer, restituted to Maurice de Rothschild in 1948, and reached the Metropolitan Museum in 1954 as accession 54.1.2.

The Vienna object is the outlier by more than half a century, and the reason is a legal mechanism rather than a lost file. Section 1 of the Austrian Kunstrueckgabegesetz, published as BGBl. I Nr. 181/1998 and in force from 5 December 1998, authorises the transfer back to original owners or their heirs of objects in federal museums and collections that had been restituted after 8 May 1945 and had then passed into federal ownership without payment in the course of a procedure under the 1918 law prohibiting the export of objects of historical, artistic or cultural significance. The statute names the mechanism it exists to undo. Once it was in force the Vienna holdings moved quickly: Christie's sold the collection of the Barons Nathaniel and Albert von Rothschild in London on 8 July 1999.