Code busting in Milan ... Da Vinci's Last Supper. Photograph: Carlo Ferraro/EPA
A somewhat eccentric scholar has just left my room. He wants to fit a scientific diagram into a particular painting (I don't want to be specific since that would be unfair on him and his unpublished research). He sets me thinking about what happens when a historian of science applies his or her way of thinking to the analysis of works of art.
There are now quite a number of instances of historians of science, or those minded to read science into paintings (including some art historians), adducing elaborate patterns of surface geometry in renaissance and baroque paintings.
Not infrequently the geometry imposed on the paintings has mystical connotations - the pentagram and that kind thing. Responsible historians of science look more soberly for geometry from the period and orbit of the artist, drawing lines on reproductions - all too often thick lines on greatly reduced images.
But there is an insuperable basic problem with this kind of analysis. It's not a question of whether the geometry looks apparently convincing - that is to say whether the key lines or points hit "significant" features in the painting. The problem is that there is no evidence (known to me at least) of intricate surface geometry being used to plan the composition of any renaissance or baroque painting.
We've hosts of compositional drawings. And now we can look to technical examination - microscopy, X-rays, infrared and so on. But there is not the slightest indication of a fretwork of surface lines being used to locate elements of any composition. There are a great deal of examples of perspectival geometry, to be sure, but no flat geometrical mapping of the distribution of forms across the surface.
"Ah, but it's secret geometry. I've found it in the painting. And because it's a secret code, the artist has not left any trace". That's the kind of argument I have encountered and, it seems, almost anything can be "proved" in this way. Perhaps Leonardo was a woman? (I'm not forgetting the theory that Mona Lisa is Leonardo in drag.)
The idea that paintings embody codes, geometrical or otherwise, much in circulation of late, falls before the same kind of argument as that I've used about surface geometry. There is no evidence from the renaissance and baroque that any painting was intended to be or was read as containing a code - taking the definition of a code as a concealed message that is quite different from what is immediately apparent. The more misleading the surface appearance, the better it is, if a code is to hide its meaning effectively. Paintings don't function like that.
Paintings are full of deeper meanings - symbolic and allegorical. But the surface and the depth are always related in a non-arbitrary way. Leonardo's Last Supper contains allusions to the Eucharist and Christ's sacrifice, in addition to the narrative of his betrayal, but these meanings are not "coded". They are not concealed by some kind of strategy to mislead.
So, please, no more surface geometry. And no more codes.
