Authentikon

Why Nothing by Cimabue Is Signed, and What That Means for Attribution

An artist at the origin of the Western canon left no signature and one documented work. Everything else is reconstruction.

Iconotheca Desk2 August 20268 min read

In short

No work by Cimabue is signed or securely dated.
Exactly one work is documented in his lifetime, and it is a mosaic.
The accepted corpus was assembled by scholarly consensus over 150 years.
An attribution to Cimabue is therefore an attribution to a reconstructed hand.

One document

Cimabue died in Pisa in 1302. The only surviving work documented as his during his lifetime is the mosaic of St John the Evangelist in the apse of Pisa Cathedral, made in 1301 and 1302. Every panel painting given to him rests on attribution rather than record.

This is not unusual for the thirteenth century. What makes it consequential is the position the artist occupies: Vasari begins the Lives with Cimabue, which means the written history of Western art opens with a figure whose corpus is almost entirely inferred.

What a reconstructed corpus is

The forty or so works given to Cimabue, with widely varying confidence, were assembled over the nineteenth and twentieth centuries by comparing pictures to one another and to the documented mosaic. Stricter counts run to about a dozen.

The logical structure is worth stating plainly. There is no autograph benchmark. The corpus was built by treating certain works as central and measuring others against them, which means the standard against which candidates are judged is itself an argument.

Why the panel from Compiegne is unusually well founded

Against that background, the small Mocking of Christ found in a French kitchen in 2019 rests on stronger evidence than most. Infrared examination revealed underdrawing consistent with the accepted works, which is the ordinary kind of argument.

The unusual part is physical. The panel shares woodworm tunnels with two other pictures, the Flagellation in the Frick and the Virgin and Child in the National Gallery. Insects bored through the wood while the panels were joined. That is evidence about objects, not about style, and no amount of connoisseurial disagreement can dissolve it.

The honest formulation

The correct way to state an attribution in these conditions is not 'by Cimabue' but 'by the hand that painted the works we have agreed to call Cimabue'. That sounds like pedantry and is not.

It changes what a challenge would have to show. Disputing this panel does not mean proving a negative about a documented artist; it means arguing that the reconstructed group should be divided differently. That is a live scholarly question for several thirteenth-century figures.

An artist with no signature, one document and a corpus built by argument is a useful reminder of what attribution actually is. Not the identification of a person, but the reconstruction of a hand from the objects it is agreed to have left behind.

Questions

Did artists sign works in the thirteenth century?

Sometimes, and some contemporaries did. The absence is normal rather than remarkable, but it removes the simplest form of evidence.

How many works are accepted?

Roughly forty with varying confidence; stricter counts around a dozen. The number depends entirely on how much doubt is tolerated.

Could the corpus be split between two artists?

It has been proposed for comparable figures of the period. For Cimabue the group is held together partly by documented association with Assisi and Pisa.