What the Word Signed Hides

The word signed covers at least five different kinds of evidence

A catalogue entry that says an object is signed is making one of several completely different claims, and the entry rarely says which. The mark under discussion may have been cut by the artist, stamped by a guild inspector, struck by a state assay office, printed by the press along with the image, or added by a firm after the named person had died. These carry incompatible evidentiary weight, and only the first is a statement about a hand.

Reading across the marks recorded in this archive, a pattern emerges that runs against the way the market ranks them. The marks applied by institutions are the most reliable and the least informative about authorship. The mark applied by the artist is the most informative about authorship and the least reliable, because it is the one nobody regulated, the easiest thing on the object to add, and the hardest to date.

So the useful question is not whether a thing is signed. It is who applied the mark, under what obligation, and what exactly the mark was designed to fix.

The hallmark dates an object to a year and does not name its maker

A full traditional British hallmark comprises five marks, of which three are compulsory: the sponsor's mark, the millesimal fineness mark and the assay office mark. The traditional fineness symbol and the date letter are optional, although the Goldsmiths' Company Assay Office applies both as standard.

Two things about that system matter for research and are widely misunderstood. The first is what the sponsor's mark means. The London Assay Office states plainly that although this mark is sometimes informally called a maker's mark, its purpose has always been to establish responsibility rather than identify the manufacturer, and that it is applied on behalf of manufacturers, commissioners of goods or third parties and can be applied to items made by others. The Paul de Lamerie mark on the Hardwicke tureen therefore establishes that de Lamerie's workshop was answerable for submitting the piece to the assay office. It does not establish that de Lamerie raised it, chased it or engraved it, and in a busy Georgian workshop he very probably did not do all three.

The second is the date letter. It changes annually on 1 January, and the font, the case and the shape of the surrounding shield all change with it, so that each letter can indicate only one specific year. Every date punch is destroyed at the end of the year it stands for. There is no other mark discussed in this article whose evidence rests on a physical object having been deliberately eliminated, and that destruction is the reason a piece of London silver can be dated to a twelve-month period with an assurance no painting in this archive comes close to.

A guild stamp can outrank a museum's own reading of an object

The pair of commodes at the Getty, accession 71.DA.96.1 and .2, is the clearest case in this archive of a mark overriding connoisseurship. The Getty's published catalogue records that the gilt-bronze mounts on these commodes are of designs quite unlike any others used by Van Risenburgh, that their attachment was ill considered, that the corner mounts obscure the profile of the corners and that the clumsiness of one hanging mount is not found elsewhere on this master's usually carefully conceived pieces.

Having said all that, the catalogue then states that if it were not for the presence of his stamp on these pieces and on the commodes in Dresden, they would not be attributed to Van Risenburgh on the basis of their design and quality. The museum is telling the reader that its own eye would reach a different conclusion and that the mark decides. That is an admirably honest position and it is also worth noticing what it implies: for stamped French furniture, the mark is not one input to the attribution, it is the attribution.

There is a further twist. The B.V.R.B. stamp was in use through the eighteenth century and required by the guild, and yet nobody knew whose initials they were until Jean-Pierre Baroli published the identification in 1957. For roughly two hundred years the market bought, sold and catalogued objects under a mark whose owner was unknown. The stamp was doing full attributional work long before anyone could say what it attributed them to.

A printed monogram is a property of the plate and identifies no sheet

Toulouse-Lautrec's monogram on the 1891 poster Moulin Rouge: La Goulue is incorporated into the lithographic design itself, so it is printed along with the image on every impression that came off the stones. A commercial poster of this type carries no separate hand signature.

That has a consequence that is obvious once stated and constantly forgotten in practice. A mark that reproduces cannot distinguish between reproductions. The Metropolitan Museum's sheet, accession 32.88.12, and the Art Institute of Chicago's sheet, accession 1954.1193, are two separate physical objects of the same design and both carry the monogram identically. If you are trying to establish which impression passed through a particular collection, the monogram tells you nothing at all.

The same applies to Hokusai's signature on the Great Wave, which reads in translation from the brush of Iitsu, formerly Hokusai. It is cut into the block. It is on every impression, and it is genuinely informative, but about the date of the design rather than about any sheet: Hokusai changed art-name at intervals, so the form of the signature dates the block. Once again the mark is doing real work, and the work is not the work people assume.

One mark in this archive was designed to say not by him

Emile Galle died on 23 September 1904, and the firm that carried his name went on producing glass under that name until 1936. His widow settled on adding an asterisk or star beside the existing Galle signature to mark the new era, and that star is the standard indicator of posthumous production in the Galle market.

It is a strange object as evidence. It is a mark that reproduces the signature of a man who is dead and adds one character meaning he did not make this. It is applied by a company for commercial rather than scholarly reasons, and its dating turns out to be shaky: Samuel Provost, who has worked through the chronology of the firm's signatures in detail, shows that the usual 1904 to 1906 dating rests on the oral testimony of a former Galle employee, Rene Dezavelle, and that the testimony was mistaken.

The hand at the Musee d'Orsay, given by the museum as 1904 and described by it as very probably the last work of crafted glass produced by the master from Nancy, is engraved in relief at the base with the name and no star. So the object's date rests partly on the absence of a mark: on a firm's convention, uncertainly dated, read negatively. This is the weakest link in the chain of marks in this archive, and it is doing work in the market every day.

The autograph signature is the one mark nobody regulated

Against the hallmark, the guild stamp and the printed monogram, all of which are applied by somebody other than the artist under some kind of obligation, the painter's signature stands alone. Nothing compels it, nothing records it, no register lists it, and anybody with a brush can add one at any point in an object's life. It is the mark the market pays most for and the only one with no institutional backing whatsoever.

The one thing that can be established about a painted signature is where it sits in the layer structure, and doing that properly is expensive. Girl with a Pearl Earring is probably the most examined painting in the world, and the Girl in the Spotlight project imaged its signature using macroscopic X-ray fluorescence, optical coherence tomography and 3D digital microscopy at 1.1 micrometres per pixel, which is 140 times magnification. The signature had degraded into the darkened background and was barely visible to the naked eye.

The published conclusion is worth quoting in shape rather than in full: the images appear to support the theory that the signature is on top of the glaze, individual white and black pigment particles appear to sit on the surface in one thin diagonal stroke of the M, and where cracks are present within that brushstroke the greenish glaze appears to lie beneath. That is careful, honest, expensive science, and it lands on appears. A date letter on a spoon, applied by a clerk with a punch that was then destroyed, lands on a year.

Five marks, and what each one actually fixes

Setting the cases side by side makes the inversion plain: reliability and informativeness run in opposite directions across this list. The marks that can be trusted absolutely say least about authorship, and the mark that speaks directly to authorship can be trusted least.

A researcher can use this directly. When a catalogue says signed, find out which row it is in, and calibrate accordingly.

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.