States: how a plate changes

Every time a plate is altered, impressions pulled after the alteration differ from those pulled before. Cataloguers call each distinguishable stage a state, and number them: state I, state II, and so on.

It sounds tidy. It is not.

Who decides what counts as a change

The artist does not declare states. A cataloguer studying a body of impressions notices that some have a line that others lack, groups them, and publishes the division. A later cataloguer with access to more impressions, better images or a sharper eye may notice a change the first one missed — and the state count goes up.

This is why the number of states for a plate is a property of the catalogue you are reading, not of the plate.

A worked example: this print

The plate at the centre of this project is a good illustration, because the catalogues disagree about it more than usual.

• Bartsch, 1797, recognises one state.
• Hollstein, the original series, recognises one.
• White & Boon, 1969, recognises one — the National Gallery of Art's record puts it flatly: “White/Boon 1969, no. 20, State only.”
• The New Hollstein volumes by Hinterding and Rutgers recognise four. The British Museum cites impressions as 170.II and 170.IV; the Rijksmuseum encodes its three impressions as 170-1(4), 170-2(4) and 170-4(4).
• And the National Gallery of Art's own bibliography cites the New Hollstein as “no. 170, state ii/iii” — a three-state reading of the same catalogue that two other institutions read as four.

So a description reading “state III of IV” is meaningful in exactly one of those frameworks, and unintelligible in three of them. A record that says “state III” without naming the catalogue has not told you anything checkable.

What distinguishes the states here

Where they are described at all, the differences are small and specific. The British Museum's description of the second state is “with jaw delineated” — the chin below the left ear is now clearly defined. The fourth state has “signature redrawn”, and belongs to an edition published in Paris around 1810, a century and a half after Rembrandt's death.

The third state is not described in any source this project could reach. That is not a gap in this website; it is a gap in what is publicly accessible about this plate, and it happens to sit exactly where it is least convenient.

The lesson for reading any print record

When you see a state cited, ask three questions. Which catalogue? What physically distinguishes that state from its neighbours? And did the person writing the record check, or copy it from the person before them?