… [Here] is a reproduction of Ingre’s portrait drawing of Madame Louis-François Godinot from 1829 …
… You will come across my term ‘eyeballing’ throughout this book. By this, I mean the way an artist sits down in front of a sitter and draws or paints a portrait by using his hand and eye alone and nothing else, looking at the figure and then trying to re-create the likeness on the paper or canvas. By doing this, he ‘gropes’ for the form he sees before him.
But the swiftness of Ingres’s lines in this drawing does not look groped for. The form is so precise and accurate. I think he did the head first by looking at the lady through a camera lucida and making a few notations on the paper, fixing the position of her hair, her eyes, her nostrils and the edges of her mouth (notice the deep shadows). This may have taken him a few minutes to do. Then, he finished drawing her face from observation, eyeballing — the delicacy of the portrait suggests he spent an hour or two doing this. Later, perhaps after lunch, he did the clothes, but for this he would probably have had to move his camera lucida slightly.* It is now we notice that the lady’s head seems very large in relation to her body, especially compared with her shoulders and hands. If Ingres had moved his camera lucida to get in the clothes, a slight change in the magnification would have occurred, explaining the difference in scale we have noticed. I think this is what happened, because when we reduced the head by 8% (it seemed to ‘fit’ her body much better.
… The three small drawings by Ingres on this page, on the other hand, are all obviously eyeballed. The lines are groped for, there are signs of hesitation. my word ‘groping’ suggests uncertainty — ‘Exactly where is the correct position?’ he seems to ask.
Notice the difference between the sleeve to the right (done in a conventional eyeballed way) with the one on the opposite page, which is drawn with confidence and continuous lines.
There is no groping here … It seems to have been done with a certain speed. All drawn lines have a speed that can usually be deduced: they have a beginning and an end, and therefore represent time, as well as space. Even a tracing of a photograph contains more ‘time’ than the original photograph …
[* With a mirror-lens projection, the usable image is never much more than a foot (thirty centimetres) across — this is an optical characteristic of all concave mirrors, no matter how big they are. Outside this ‘sweet spot’ it is impossible to get the image into sharp focus. Paintings made with the help of a mirror-lens must therefore be very small, or must be a collage of small glimpses: portraits; details of hands, clothes, feet; fragments of landscape — and still lifes.]
I find it interesting that he seems to assume the perfect line is just … there … to be traced using the camera lucida. As any photographer knows, that’s not true. The struggle (the ‘groping’) to find it has to happen before you can begin. Granted there is further ‘groping’ for the painter or sketch artist to make the transition from eye to hand, but that can’t happen until the first and to my mind, far harder exercise of finding the line “out there.”
If you are interested in learning more about Hockney’s theories but don’t want to buy the book, there’s a pretty good description of the arguments for and against in the Wikipedia entry for the Hockney–Falco thesis (as it is called). [ link ]