When science takes over art history
September 10 2012
Picture: Sunday Telegraph/Rubenshuis
A consequence of the decline in connoisseurship is the disproportionate weight placed on science when attributing paintings. In the world of science, an answer is usually binary - something either is or it isn't. When scientists get involved in art history, therefore, they often attempt to bring too much certainty to a situation.
Here's an example: 'This zippy new machine and Professor X say this painting is by Van Dyck. Therefore it must be by Van Dyck.' And today's art historians, increasingly unsure of how to assess paintings because their teachers taught them that connoisseurship is a Bad Thing, are often only too happy to go along. Of course, Professor X may have looked at as many Van Dycks in his life as his zippy new machine has - ie, none. But because 'science' is involved, the decision is final. Nobody ever asks how much the new machine costs, and whether the scientists are overly keen to market their new invention.
Yesterday we had the Sunday Telegraph reporting that a a new 3D X-ray 'Minidome' machine had proved that a painting thought to be by Rubens of Van Dyck was in fact a self-portrait by Van Dyck. The key fact was this:
By looking at the brush strokes on the surface of the painting, the researchers found that it had been built up in layers, continually revised during its creation – a technique associated with van Dyck, who lived from 1599 to 1641 and is best known for his portraits of the pre-civil war royal court. Art historians know that he continually rethought his composition and technique as he went along.
Rubens, by contrast, did not; he would paint according to a pre-conceived plan, not revising his earlier work, meaning that his brushstrokes would show up entirely differently under 3D examination.
This is such a load of old phooey that I can scarcely believe it to be true. (If you don't believe me, make a quick trip to the National Gallery and look closely at any of the Rubens' on display there. You'll soon find evidence of Rubens changing his mind - moving the leg of a cow here, or the finger of a model there - or as we call them in art history jargon, pentimenti.)
But let's leave aside the question of whether that's the only evidence for changing the attribution from Rubens to Van Dyck, and hope that the facts will be better presented in a proper article, which we can return to. The main point here is that before we go around changing attributions based on a zippy new x-ray machine, we surely need that machine to make many, many scans of other paintings by Rubens and Van Dyck (for example). Then we will not only have a proper body of comparable information from which to draw our conclusions, but we can also be sure that we are interpreting these new images correctly.
I'm certainly all in favour of using science to help attribute paintings. We will feature three examples of how just how important it is in our new series of 'Fake or Fortune?' But we must be careful how much emphasis we place on science, which by and large can only tell you what a painting is not, not what it is. If we rely too heavily on new technology, we may find to our cost that old fashioned art history is discarded entirely. And as the poorly written article in The Sunday Telegraph shows us, it's already happening.
