The Signature : Finding the hand that made it

The Signature

Most commercial illustration was commissioned work, and crediting the artist was not standard practice. The name was frequently omitted, cropped during reproduction, or lost when the plate was trimmed.

When a signature does survive, it is often the only direct evidence of authorship the page carries. It is worth knowing where to look.

Where it sits

In the artwork, not the margin. Illustrators signed the drawing itself, so the signature is inside the printed image — along a lower edge, against a flat area of tone, or worked into a shadow. It moves with the composition rather than sitting in a fixed place.

At the boundary of a shape. A common habit is to run the signature along the edge of a dark mass, where the letters read clearly without interrupting anything. On a figure, that may be the line of a shoulder, a hat brim, or a coat.

Rotated with the image. Where the composition has a strong diagonal, the signature may follow it. On a page held upright it can be almost vertical, which makes it easy to read past.

Editorial illustration signs more openly than advertising. A magazine illustrating a story credited the artist as a matter of course, and the signature is often larger and more confidently placed. The same artist working on an advertisement in the same issue may not have signed at all.

Photographers are different

A photograph offers no surface to sign. Nothing can be written into the image the way it can be drawn into an illustration, so a photographer's credit, if it appears at all, is set as type — along the edge of the image, in the gutter, in a caption, or in a credits column at the front of the issue.

This makes photographic attribution far more dependent on the page staying with its issue. A signed illustration carries its evidence with it once separated. A photograph can lose its credit when the leaf is separated, unless the credit happens to fall within the retained portion.

What a signature is and is not

A signature is evidence of authorship. It is not proof of it.

Studios signed with a principal's name over work executed by assistants. Successful signatures were imitated. Some artists used more than one name across categories, particularly where advertising work sat below their editorial reputation. A monogram may be the only mark left — initials worked into a border, an object, or the corner of a shape.

The useful record therefore states two things separately: what the mark looks like, and what it is thought to mean.

How to read one that resists

Signatures were written at drawing scale and then reduced for reproduction, often heavily. A name that was legible on the original board can arrive on the page as four or five strokes.

Raking light and a loupe. Angle the page so light crosses the surface — the same technique used for [The Fine Print →] Scan at high resolution and enlarge. A signature that is illegible at print size frequently resolves at 1200 dpi, because the ink is still there — it is the eye that cannot separate it. Compare, do not decipher. Reading an unfamiliar signature cold is the hardest way to do it. Matching it against a known example from the same period is far more reliable, and this is why signed, dated, credited examples are worth keeping even when the page itself is unremarkable. Use the campaign history. Long-running campaigns were often handled by a small number of artists. Establishing the agency or the client narrows the field before you have read a letter. How we handle it

We record the signature as it appears before recording any interpretation of it. Those are two different claims and we keep them separate — a mark reading as five strokes is a mark reading as five strokes, whatever we later think it says.

Where a signature is present, we photograph it. Where we can support an attribution, we state what supports it. Where we cannot, we say the signature is present and unidentified rather than offering a name that would help the page sell.

What is observable is separated from what is inferred. A signature is observable. A name is usually an inference.