The name that appears in more than one chain

Only twenty-four names appear in three or more chains in this archive

We hold 291 objects and 2,033 recorded ownership links, and when we ask which personal or firm names appear in three or more separate chains, the answer is twenty-four. That is a small number, and it is small for a reason that is structural rather than accidental. Most people who own an important object own one important object. A provenance line is mostly made of names that occur exactly once in the whole corpus and can never be checked against anything else in it.

The names that do recur are dominated by the trade. Duveen appears in six of our records, Colnaghi in six, Knoedler in four, W. E. Hill in four, Wurlitzer in four, Yamanaka in three. The private collectors who recur are the very large ones, and they recur less: Pierpont Morgan in six, Henry Walters in four, Sir William Hamilton in three, Mariano Fortuny in three. A dealer passes hundreds of objects through their hands in a career and a collector passes dozens, so the trade is over-represented in any archive of ownership by simple arithmetic.

This has an immediate consequence that we think is under-appreciated. If you are checking a provenance line and you want to test any part of it against independent evidence, the dealer is very often the only link in it that appears anywhere else. Everything either side of that name is unique to the object.

A recurring name is the only cross-checkable node in a chain

A provenance line is a chain of assertions, and most of its links cannot be tested from outside the object's own file. A named collector who owned the thing between 1840 and 1871 either appears in the literature on that object or does not appear at all. There is nothing to triangulate with.

A dealer is different. A dealer has stockbooks, day books, photographic archives, correspondence and, in several well-known cases, an institutional archive that scholars have already indexed. If the same firm appears in the chain of an object you can consult, you have a way in: you can find out what their numbering looked like in that decade, what their labels looked like, whether they were dealing in that material at all at that date. We have written elsewhere about what a stockbook can prove and where it stops, and the short version is that it proves possession and rarely proves title. Even so, it is evidence that exists in a place other than the object's own dossier, and in provenance work that is rare enough to be valuable.

So the recurring name is the load-bearing part of a chain for diligence purposes, which is exactly why it deserves more suspicion than it usually gets.

In one chain here the same surname is two different men

The Warwick Vase is the case that should be pinned above the desk of anyone matching provenance records. Its chain reads, in consecutive links, that Gavin Hamilton found the vase in fragments while digging the silts of the Pantanello at Hadrian's Villa near Tivoli, and that Sir William Hamilton then purchased those fragments from him and commissioned the restoration, carried out between 1772 and 1774 by Grandjacquet to designs by Piranesi, using newly cut marble for the missing parts. Glasgow Museums dates the excavation to 1769 and 1770; the Glasgow Cultural History account puts the discovery in 1771. We have not been able to reconcile those, and the distinction does not affect the point below.

These are two different men, and the dates settle it. Gavin Hamilton, 1723 to 1798, was a Lanarkshire-born painter of history subjects who worked in Rome as a dealer and archaeologist. Sir William Hamilton, 1730 to 1803, was the British ambassador to the Kingdom of Naples from 1764 to 1800, an antiquarian and a volcanologist. The relationship between them in this transaction is the ordinary one between a dealer and a customer. Sir William then gave the restored vase, rather than selling it, to his nephew George Greville, Earl of Warwick, after failing to sell it to the British Museum, and it stood at Warwick Castle for nearly two centuries until Glasgow Museums bought it in 1979.

Collapse those two links on the surname and the chain acquires a fictitious continuity: one Hamilton who found the object and kept it for five years before passing it to his nephew. The dealer disappears, the sale disappears, and with them the whole question of what was paid and what was restored on whose instruction. It is a small error that removes the most informative event in the sequence.

The recurring dealers do not stay inside one material

We expected, when we started counting, that the recurring dealers would each sit inside a single category and the recurring collectors would range across several. The data says otherwise, and it is worth publishing the correction rather than the hypothesis.

Duveen crosses five object types in six records: a carpet, an Iznik dish, sculpture, two pieces of French furniture and a piece of goldsmith's work. Colnaghi crosses paintings, prints and photographs. So the great dealers were generalists, and the idea that the trade sorts itself neatly by material does not survive contact with the corpus.

The one clean exception is instruments. W. E. Hill and Wurlitzer appear four times each and never outside musical instruments, and they overlap on the same objects. That is a real feature of the violin trade rather than a coincidence: it is a small closed market with its own certification culture, in which a handful of firms have authenticated and sold the same instruments to each other for a century. If your object is a violin, the recurring name in its chain is more informative than usual, because the population of firms it could be is very small.

How to use a recurring name without being caught by it

Match on the person, not on the string. That sounds obvious and it is the failure mode we see most often, because provenance data is usually stored as free text and the obvious way to search free text is to search for a surname. Before you treat two occurrences as the same party, check the dates against that person's life, check the place, and check whether they dealt or collected in that material at all. The Warwick Vase would fail all three tests if you tried to make its two Hamiltons into one man.

Then be careful in the other direction, because a firm name is not stable over time either. A house that traded for a century under one name may have had four generations of proprietors, several changes of specialism and, in at least one famous instance in these pages, an ending that retrospectively colours everything the name touches. A provenance link reading Knoedler in 1910 and a provenance link reading Knoedler in 2005 are the same string and are not the same evidence, and treating them as equivalent in either direction is an error. The first is a serious old master dealership at its height; the second sits inside a period the firm did not survive.

What we would say to anyone building a chain is this. Note every name that appears more than once in your own holdings, because that is your only cross-checkable material, and record why you think two occurrences are the same party rather than just that they match. We do not always know. Where the sources conflict, as they do over what James Byres actually did to the Warwick Vase, the honest entry records the conflict and leaves the question open rather than picking the reading that makes the chain tidier.

Why this archive is free to read

We do this for everyone who loves art. The people who have studied it for years, and the people who just stopped to look. Iconotheca exists so that what we learn belongs to all of them.

Nobody pays us for this. No ads, no sponsors, no commercial interest. These works belong to all of us, and we would rather share them with as many people as we can than keep them to ourselves.

If it gave you something today, tell us to keep going. Follow us, leave a like, or write a positive comment. We read every one, and they are what keeps us going.