Reflecting on artful tricks of the trade

A joy to behold, David Hockney's Secret Knowledge is his private adventure now shared with his readers and lookers. With his claim that paintings as early as van Eyck's Marriage of the Arnolfini (1434) were created by tracing optical images, there is much to look at and read in new ways. With this challenge, Hockney has thrown down the gauntlet to show his hand grasping a dangerous nettle. Dangerous, for it not only diminishes the skill of great painters, but implies they deliberately kept secret their technical aids.

But how could the most famous painters keep the use of lenses and concave mirrors for projecting images secret? Surely the sitters would spill the beans, talking of the dark enclosure with the instant magic pictures? Slide projectors were not called magic lanterns without reason. They brought magic to village halls and enhanced mundane lectures. They were certainly talked about - so how were artists' earlier, and more surprising, projected images kept secret? This is even more puzzling when we remember that aids for perspective drawing were fully described and freely used in the Renaissance.

This initial doubt, with its hint of a shameful secret, makes the idea even more intriguing and worth pursuing. If Hockney has opened a can of worms, it is remarkably interesting. Appealing, too, is the way Hockney gathers the evidence and assesses it using his remarkable skills of picture-making. He knows what is hard to do and spots clues from his special knowledge.

Hockney says of van Eyck's elaborate, perfectly formed chandelier: "The chandelier has always fascinated me. It was done without any detailed underdrawing or corrections, amazing for such a complicated foreshortened form. Van Eyck could have hung the panel upside down next to the viewing hole and painted it directly, following the forms he could see on the surface... It would have been a superb subject to show off his skills. Artists think of these things." Hockney is surely right to say that artists use, and have every right to use, any technique for picture-making that comes to hand and pleases the eye. Leonardo recommended checking a picture by looking at it reversed in a looking-glass.

Here it is large concave mirrors, for throwing images on a screen, that are being considered. Were they available in the early 1400s? Yes, just what is needed is there in van Eyck's picture - for in the centre is a large convex glass mirror. As Hockney says, this could very easily be turned around for its concave back to form images on a screen. Possibly this very mirror was used for painting the chandelier, though the mirror itself appears in "photographic" detail, and includes the chandelier.

It is well established that the camera obscura was used by Canaletto in the first half of the 18th century for his architectural studies of Venice and London; evidently he added the figures later, as they are twice too large for the buildings. It has been shown convincingly by Philip Steadman, using experiments, that in the mid-17th century Vermeer converted his studio into a large camera obscura. What is startling about Hockney's claim is the early date and the secrecy. But it could have been done. Large concave mirrors were used by alchemists for melting metals. Convex lenses for spectacles and magnifying glasses became available somewhat later.

Hockney makes amusing and instructive play with the reversals of mirrors and lenses - which give us a strong clue as to which would have been used for a particular picture. An image can be cast from either, and it is upside down for both; but the mirror's projected image is not right-left reversed, though the image from a lens is horizontally reversed on the screen. The reason is not altogether obvious. Both are optically symmetrical, up-down and right-left. So why are their images not symmetrical? Why are they reversed? The same question applies to objects reflected in a looking-glass. A book looks right-left reversed, but not upside down; though the glass is optically symmetrical. The reversal comes from rotating the book to face the mirror. For Hockney's concave mirror, the subject will be turned away from the screen, facing the mirror behind him. But with a lens he will be facing the screen. It is this reversal of the subject, to face the mirror behind him or the lens in front, that gives the game away. Hockney notes that there are far too many left-handed people in some of the later pictures. This is evidence that the use of concave mirrors gave way to lenses; which is likely, as large lenses became available later.

For large scenes, or for several objects of interest, perspective viewpoints are usually inconsistent for various regions. This is impossible for a camera, unless the picture is a montage from several positions. Hockney suggests just this for Holbein's The Ambassadors (1533), where there are several perspectives for the various wonderfully painted musical and scientific instruments lying on the table. As the optical images would have been much smaller than the whole picture, it must have been sensible to deal with each separately, from effective viewpoints, and combine them into a single composition. Our eyes are tolerant of such inconsistencies, which must help Hockney, as he produces wonderful montages using cameras.

One point is hard to follow. He says that optical projections help to capture fleeting facial expressions, and notes that faces are painted more individually when these aids became available. But surely a fleeting smile is no more captured in an image than by looking directly. It is the film that freezes motion; but this came much later.

One might think that to get a single perspective for a whole scene, tracing on a sheet of glass, placed between the painter and the scene, would be effective and relatively easy. However all this may be, my admiration for the skill of the old masters, and of Hockney, is undiminished. If they had a bit of help from optics, they were making intelligent use of science. Where's the harm in that?

Richard Gregory is emeritus professor of neuropsychology, University of Bristol.