How to Read a Provenance Line

The line is a compression, not a story

A provenance line in a catalogue entry looks like a list of names and dates. It is in fact a compressed legal history of an object, written in a notation that assumes the reader already knows the conventions. Almost nobody is taught those conventions directly. They are absorbed slowly, from reading entries until the pattern emerges, which is why a first-year student and a department head can look at the same six lines and take away completely different amounts of information.

The notation is worth learning explicitly, because once it is visible you cannot stop seeing it. Most of what a provenance line has to tell you is in its punctuation, its hedges and its gaps rather than in its names.

Punctuation carries the transactions

In the standard convention, a semicolon separates owners: the object moved from one to the next. A comma does not, and typically continues a description of the same owner, their city, or the manner of transfer. So a line reading 'Paris, private collection; sale, Christie's London, 3 July 1998, lot 44; acquired by the present owner' describes two changes of hands, not three, and tells you that the last transfer happened at that sale.

Square brackets usually mark an editorial interpolation, something the cataloguer has supplied rather than found. Parentheses often carry a dealer through whose hands the object passed without the dealer being the beneficial owner. The distinction sounds pedantic until you are trying to establish who had title in a particular year, at which point it is the whole question.

The hedges are graded, and they are the content

Cataloguers hedge deliberately and the hedges form a rough scale. 'Probably' generally means there is documentary support that falls short of proof. 'Possibly' means the identification is plausible but unevidenced. 'Said to be' means someone asserted it and the cataloguer could not confirm it, and is frequently a polite way of recording a claim the house does not itself believe.

Treat every one of these as a flag rather than as prose. A line that says a picture was 'possibly' in a famous seventeenth-century sale is making an argument, and arguments can be checked. Our own record for Vermeer's Girl with a Pearl Earring carries exactly this problem: the painting is often identified with an unnamed lot in the 1696 Dissius sale, and that identification has been repeated so often that it can read as settled. It is not settled, and the record says so.

Read the gaps before you read the names

The most valuable part of a provenance line is usually the part that is not there. A gap is not automatically sinister; most objects have them, and two centuries of silence on a Dutch cabinet picture is unremarkable. But gaps are graded too, and one span matters more than all the others.

An object of European origin with no documented owner between 1933 and 1945 requires an explanation, and its absence is the reason provenance research became a professional discipline rather than a cataloguing chore. Museums, auction houses and insurers all treat that window as a distinct question, and a catalogue entry that passes over it in silence is telling you something about the cataloguer as well as about the object.

The second most informative gap is the one immediately before a work appears on the market. An object that surfaces with a first documented owner in the year it was consigned has, in effect, no provenance at all, however respectable the consignor.

A clean line is not the same as a sourced line

The habit worth building is to ask, for every link in a chain, what document would prove it and whether anyone has cited that document. Wills, inventories, insurance schedules, shipping manifests, exhibition catalogues and dealer stockbooks are the primary material; a name and a date with none of them behind it is an assertion.

This is why an unbroken-looking provenance with no citations should worry you more than a chain with an honest hole in it and a footnote at every other link. The first has been written to look finished. The second has been researched.