Camera Absurda, The Case against Hockney


La Grande Odalisque, Ingres, 1814

La Grande Odalisque, Ingres, 1814

The painter David Hockney has garnered quite a bit of attention lately by saying that renowned artists from the past often used optical aids to help make their art. Camera obscura, cam-era lucida, and convex mirrors were all employed, says Hockney, to enhance the beauty and verisimilitude of many Old Master pictures. Hockney has written a big, lavishly illustrated book aimed at proving his case, and I have now read this book as well as another book with a similar point of view called Vermeer's Camera. I have diligently examined the graphs, weighed the arguments, studied the pictures, and after due consideration, I now feel absolutely confident in saying that this whole point of view from its initial conception to its elaborate proof is, in a word, kooky. And by that I don't mean just slightly kooky. I mean seriously, Bigfoot fanatic, UFO convention, world-class kooky.

Hockney's argument breaks down into seven major points:
(1) The line in a Warhol drawing resembles the line in an Ingres drawing. Warhol projected his images on to the paper, Ingres must have too.
(2) The detail work in Ingres's paintings is so complicated and so elaborate that it could only be produced by projecting and tracing.
(3) Art became dramatically more life-like during the same era that lenses started to be developed.
(4) The perspective in certain paintings is too difficult for someone to do freehand.
(5) Paintings of people drinking with their left hand show that a camera obscura (which reverses images) was used.
(6) The dramatic light and shadow of chiaroscuro paintings resembles sun-light more than studio light and sun-light is what's needed to get a clear image when using a camera obscura.
(7) Certain distortions in some paintings look like the kind of distortions that would be caused by optics.

That's it. Those are the arguments. To my mind what's not there is more striking than what is. I feel it's significant that what we don't find on this list is: "Such and such a great artist was often (or even once) seen using a special lens" or "Such and such a great artist was known to own a camera obscura." Or "When I posed for so and so he was often looking through a prism device." Hockney says that over the whole span of art history the reason we don't have any independent verification of such use is that optical aids were a closely guarded secret. I'll say. Four hundred years of world famous artists peering through things, climbing into dark boxes, and tracing images upside down and nobody mentions it? No model, no client, no apprentice utters a word? Clearly we are in grassy knoll territory.
Let's go through the list:

(1) Warhol's line looks like Ingres's line therefore...
This is an easy one. Warhol's line does not look like Ingres's line. At all. Warhol's line is stark and unvarying, Ingres's line is subtle, changing, shifting from sharp to soft, defined to vague and then back again. It, unlike Warhol's line, is expressive of the substance—skin, fabric, bone—that is being depicted. It's the kind of line that an empathetic draftsman (as opposed to an outline tracer) who is looking sensitively at his subject would use.

(2) The detail work in Ingres's patterned cloth is so complicated and so accurate that it could only be done with mechanical assistance, specifically, Hockney says, a camera obscura. To prove his point, Hockney compares Ingres's cloth with some patterned cloth painted by Cezanne. The Cezanne cloth was "certainly eyeballed," says Hockney, and "consequently" looks strikingly less convincing than the Ingres.2 This is one of those missing parts arguments: Detailed accuracy can only be achieved with a camera obscura; Ingres painted with detailed accuracy; therefore, he used a camera obscura. Hockney has skipped the part where he proves that detailed accuracy can only be done with a camera obscura. In fact, as it happens, if you need to paint a lot of detail, a camera obscura is the last thing you'd want to use. Here's why.
The camera obscura requires complete darkness. The viewer needs to be inside a very dark space and it's into this space that light from a pinhole projects itself. Ingres, therefore, would have to be working in, more or less, a pitch black chamber. He wouldn't be able to paint in there. He couldn't see. All he could do, and Hockney agrees with this, is draw the outline of the projected image onto his canvas, an outline that he would, presumably, fill in with paint after he'd gotten out. It's a lot of trouble just to get an outline, but even if we grant that Ingres might have been willing to go through this, there's still a big problem. Namely, the camera obscura reverses the image and flips it upside down. The upside down part is okay; Ingres could trace the outline on to his canvas and then turn the canvas right side up afterwards. But the reversal, as far as I can see, puts the kibosh on the whole project. How can it be helpful to have your subject outlined on your canvas in mirror image? Once the out-line is in place, you would then have to paint in the colors from life, looking directly at your subject, but everything on your canvas would be backwards. Hockney never really spells out this difficulty, but it seems to me to be an insurmountable one. Who would want to try and paint on a drawing that's the reverse of what they are looking at? Complicated detail is hard enough, but backward complicated detail?
Rather than let this reversal phenomenon undermine his theses, however, Hockney, in argument 5, uses it as evidence that artists worked with the camera obscura (a proof I'll get to later). The reversal is a difficulty though, and Hockney tries to counter it by demonstrating that a certain kind of convex mirror set up could re-project the image in a non-reversed way. But this solution is also problematic because the image you get is at most only twelve inches square. Here, a less dedicated theorist might conclude that this device would only be relevant for tiny paintings. The indefatigable Hockney, however, proposes that those using the convex mirror rig could piece large paintings together out of truncated little snippets. The artist, inside his dark box would, says Hockney, copy the projected image of a part of the subject on to the canvas, move the viewing opening to a new position, copy some more, move the opening, etc. until the whole image was hobbled together like a collage. Does Hockney seriously think that going to all this trouble would be faster or easier or more accurate than painting directly from life?
Getting back to that Cezanne comparison, I think the fact that Cezanne's cloth doesn't look as convincing as Ingres's has everything to do with Cezanne's cloth painting skills and nothing to do with mechanical assistance. Beautiful, fine detail work has been done from life for centuries without any assistance whatever, but per-haps anticipating this line of attack, Hockney quickly moves on to...

(3) Art got dramatically more lifelike at the same time that lenses were developed. Because, as I say, Ingres is not the only painter who did fine detail—art history is ruled with such skillful people—Hockney now must drag herds of previously innocent Old Masters into his mechanical aid plot. To prove that artists used lenses he points out that as the Renaissance flowers and painting geniuses start to emerge, lens technology is also developing. This is congruency being confused with causality. It's like, "There was a full moon when I got the flu; therefore, flu is caused by the full moon." Bad logic. Yes, lenses were being investigated during and after the Renaissance. Discoveries were being made. That's what the Renaissance was about. But Hockney can't show that artists used them. He shows us pictures from before the Renaissance which look flat and then pictures from after which have depth. This demonstrates, says Hockney, that lenses were used. How about this demonstrates that there was a Renaissance? That's what happened during that period; artists made discoveries about visual life, and paint, and beauty. They figured out how to get convincing depth onto their canvas, and as a matter of fact, depth is exactly what you don't get working from projected images. The pre-Renaissance artists look flat because they are copying surface effects. The great artists who follow are thinking more like sculptors, they are building depth into their canvas from the inside out. They are feeling the structure underneath and surface details are applied over their underlying structure. Using lenses to copy would undermine that sensibility. Tracing outlines, which is what lens use is all about, leads directly back to pre-Renaissance flatness.

The Lute Player, Caravaggio, 1596

The Lute Player, Caravaggio, 1596

(4) The complex foreshortening of certain subjects is too difficult for perspective alone. The artists must have used lenses. As an example, Hockney shows us a lute by Caravaggio and skeptically asks, "Could Caravaggio really have used Durer's method (perspective analysis) to paint that incredibly foreshortened lute?"3 Answer: yes. The more complicated the view of the subject, the more you need to utilize perspective. That's why the discovery of the laws of perspective was so significant. These laws allowed the artist to analyze complex objects and viewpoints and nail them. Hockney shows us various objects seen at difficult angles and says they would be too hard to depict with just perspective, but foreshortening and receding form are exactly the kind of things that "just" perspective would solve. Trying to get them to look right by, in effect, copying them, which is what Hockney's lens methods amounts to, would make the depiction less convincing because it wouldn't be derived from understanding.

(5) Paintings with left-handed drinkers in them illustrate the use of camera obscura because people don't typically drink from their left hand. As I said earlier, the reversal effect of the camera obscura should, realistically, completely rule it out as a painting aid, but here Hockney turns a lemon into lemonade. He uses the reversal to buttress his case. This Caravaggio youth is holding a glass in his left hand, and Hockney flips the image to show us that it looks more natural in the right. Putting aside the slander factor against those-who-are-left-hand-ed, the picture, as Caravaggio painted it, requires that the glass remain just where it is. As great painters were once aware, paintings generally read from left to right. The eye moves across the canvas in a right-ward direction and it's essential to keep the strongest visual elements on the right-hand side. The glass functions as a coda for the painting—a kind of climax—and, as you see, if it is reversed the picture, in a sense, is over before it's begun.

(6) The dramatic light and shadow of chiaroscuro paintings resembles sunlight more than studio light and sunlight is what's needed to get a clear image when using a camera obscura and mirrors. "With the exception of direct sunlight, you don't get such a strong light naturally, certainly not in scenes that are inside apparently darkened rooms like these; but you need such strong light to use optics."4 This is another lemon/lemonade situation. Hockney should be throwing in the towel at this point having confessed that sunlight is necessary for the equipment to work, but he surges forward undaunted saying that this explains why we have so many high contrast pictures. In fact, optics needing sunlight essentially proves that optics weren't used by the Old Masters and it proves it for the following reasons: (a) The depicted light in paintings by Caravaggio, Vermeer, Ingres, etc. is indisputably cool light, not sunlight; (b) Painters tend to avoid indoor sunlight because it's too transitory. When it comes in from one window it can only last for, at most, half a day, and during that time it moves around the room; (c) Artists for centuries have prized north light studios specifically because they don't get sunlight; and (d) If artists from northern Europe depended on sunlight to paint their paintings, museums would be empty.

(7) Certain paintings' distortions and drawing errors are the kind of thing that would occur with lens use. Hockney here is disregarding all that earlier business about how the only explanation for the lack of error and distortions in Old Master paintings is lenses, but I really don't think he can have it both ways. Either lenses help or they don't. And as if this contradictory argument weren't misleading enough, Hockney also, I'm sorry to say, cheats on the accompanying illustration. That's right. After a careful study of one of the book's pictures, I discovered a clear case of evidence fudging. To illustrate an example of a lens-caused distortion, Hockney offers us a picture which he claims has two different vanishing points where there should be just one. He says that the artist "probably paint-ed the front first, and then, because of depth of field problems, refocused his lens before painting the back, hence a second vanishing point." Aside from the fact that, again, this whole line of argument contradicts his earlier position, what he says is occurring in the painting isn't. I got out my straight edge and double checked the rug's receding pattern edges and found that there is only one vanishing point. Not two. Hockney's overlaid lines are a distortion. A distortion, I must repeat, that is in the service of a completely contradictory and meaningless line of argument.

This isn't the only instance of a picture not representing what Hockney claims it does. On page 81 and 82 there's a two-page spread of a carefully painted chandelier from van Eyck's Arnolfini Wedding that Hockney says indicates the use of optics. "Notice how the chandelier is seen head on (not from below, as you would expect)," says Hockney."This is the effect you would expect with a mirror-lens, which must be level with the objects you want to draw or paint." What? Now, aside from all the other complications with lenses, the objects have to be at eye-level for the lenses to work? Hockney keeps sneaking in new restrictive information about these "visual aids" that make them seem less and less useful. But that's not the biggest issue related to this illustration. There's another problem here which suggests that Hockney is assuming no one would ever actually look at the evidence he's presenting. Namely, the chandelier isn't at eye level. He presents the chandelier to us as clear evidence that a mirror lens was used because it is "seen head on (not from below as you would expect)." Only it isn't. The upward curve of the center piece and the higher near candles show us that this chandelier is most explicitly seen from below! Which means, and here I'm only passing on what I have just learned from Hockney, that it could not have been painted with a lens.

Allegory of Painting, Vermeer 1666

Allegory of Painting, Vermeer 1666

I think that when a theory can be completely discredited solely on the basis of its own evidence, what we can safely say about that theory is that it's not good. Hockney's big idea—that the Old Masters used optics to help them out—is propped up with skewed logic, faulty premises, unwarranted conclusions, and, as we've seen, downright dishonesty. If it had been put together and published by a scientist rather than an artist, we'd have a scandal on our hands. But Hockney is being given a free pass, and his whole cockamamie notion is getting swallowed whole by artists and art historians alike. Why? Why are so many people so willing to assume that the great artists of the past copied their pictures from projections? Some, obviously, may be just assuming that Hockney, being an artist himself, knows what he's talking about. Others may feel that even if he's wrong about occasional details, there must be something to it or there wouldn't be all this fuss. But my suspicion is that the main reason so many are being taken in by this theory is that it meets a need. The giants of art history painted pictures that raised the bar of what is possible and on some level maybe that poses a threat. Great art is inspiring but it also can be diminishing. Hockney's theory pulls everyone down to the same comfortable level.

As for Steadman's book Vermeer's Camera, which also attempts to prove the use of optic assistance, it is no less cockamamie than Hockney's, but the nuttiness is buried under an awe-inspiring amount of scholarly commotion. I don't think I've ever seen a book about art filled with so many graphs and models and grids. The author even assembles miniature three-dimensional reconstructions of Vermeer's paintings which he then reproduces and draws perspective lines all over to show us that.. .well, I couldn't tell exactly what it showed us except that the author is not afraid of hard work. Ultimately though, despite it all, I could find no convincing evidence anywhere in this book that Vermeer ever used the camera obscura. In fact, once again, like Hockney, we are help-fully provided with strong evidence that he didn't. We are told, for example, that in x-rays of the paintings there are no indications that Vermeer made preliminary out-lines. As we learned earlier, outlines are all you can do when you're using a camera obscura because it's too dark in there to paint. Thanks to this x-ray information, therefore, we now have a good indicator that Vermeer did not use the camera obscura. The author also generously points out another flaw in his theory. He tells us that conservators have discovered that six Vermeer paintings each have tiny little picture marks on their vanishing points. This indicates, we're told, that Vermeer used some kind of a pin to mark the spot where the painting's perspective lines converge. Why did Vermeer need to mark his vanishing points? If the author's theory is true that Vermeer used the camera obscura to outline all the subject matter, why would he then need a perspective aid like a marked vanishing point? The author speculates that Vermeer, after he had emerged from his camera obscura, might need to supplement some of his outlines with lines based on a vanishing point. The only reason for a vanishing point is to make sure that all the receding forms in the picture are on the correct diagonals. Does it really make sense for someone to go through all the trouble of working with the camera obscura—no light, the reversal effect—if later it all has to be redone—or "supplemented" — with a perspective system? Is this really a process any sensible artist would ever want to go through? Doesn't it seem much more reasonable that Vermeer painted the way Vermeer depicts an artist painting? In his Allegory of Painting we see an artist in his studio. This painter is not looking through lenses, or crouching in a dark box. He's simply painting from the model—not just using any out-lines—painting directly onto the canvas. Can't we just assume this is how Vermeer painted? Here with this picture, we are fortunate to have a great artist showing us, in effect, how he worked. Why doubt it? Why complicate it? Why assume Vermeer was being deceitful? Or put it another way, why be kooky?

1. Philip Steadman, Vermeer's Camera: Uncovering the Truth behind the Masterpieces
(New York: Oxford University Press, 2001).
2. Hockney, Secret Knowledge, p. 33., 3. Ibid., p. 54., 4. Ibid., p. [126]., 5. Ibid.,
p. [82]., 6. Ibid., p. [82].