It has been suggested more than once that the renascence in the reputation of Vermeer (1632-75) in the second half of the 19th century may not have been unconnected with the invention and spread of photography. The description of Vermeer by the Goncourt brothers in 1861 as "the only master who has made a daguerrotype of the red-brick houses [of Holland]" is perhaps just one early symptom of this new way of appreciating the painter. It was the American engraver Joseph Pennell, writing in the British Journal of Photography, who first hinted that such comparisons might be more than metaphors: that Vermeer might indeed have relied on some kind of optical apparatus.
This possibility has been explored over many years by Arthur Wheelock, curator of Northern Baroque painting at the National Gallery of Art in Washington, and author of Vermeer and the Art of Painting. Wheelock is one of the organisers of the current exhibition of the painter's works - the largest ever held - which has just moved from Washington to The Hague. His doctoral thesis was about connections between optics and painting in Delft in the 17th century. His book, as well his contributions to the exhibition catalogue - here reviewed together - pursue this theme of the technical basis of Vermeer's art. He takes the results of examination using X-rays and infrared reflectography, analysis of pigments, paint layers and brushstrokes, studies of perspective geometry, and asks what these can teach us about Vermeer's style, about how he achieved his miraculous effects of light, space and psychological depth.
A major part of this analysis of technique is devoted to the possibility, much discussed by other critics, that Vermeer used a camera obscura, and an instrument is on display in the exhibition. The camera obscura took different forms, but comprised in every case a lens by which an image could be projected on to a screen. This image could then be studied and traced. Some were made as closed boxes with ground-glass screens, just like the 19th-century photographic cameras of Daguerre and Fox Talbot (but without light-sensitive plates). Others were tents or cubicles, inside which the artist could work in near-darkness. In these the images could be projected on to paper or canvas.
There have, until recently, been two main reasons for imagining Vermeer to have used the camera. The first has to do with certain features of Vermeer's perspective, which Pennell originally diagnosed as "photographic". In the Officer and Laughing Girl the head of the soldier is half as large again as that of the girl, despite the fact that they sit together across the corner of a table. This is a simple geometrical consequence of the close-up viewpoint of the picture. As Wheelock says, the sense of space is "intensified" by Vermeer seeming to exaggerate the difference in scale between the figures.
Then there is the fact that some passages, especially in the foregrounds of pictures, appear to be out of focus. The suggestion is that Vermeer is reproducing not what he sees directly, but what he sees in slightly deficient optical images. Some of the most striking examples are in the tiny portrait panel Girl with a Red Hat. Here the sculpted brass lions' heads on the chair back, on which the girl rests her arm, dissolve into soft, loose, fluidly treated patches of light and shade. Similar effects are to be seen in the skeins of thread that spill like liquid from the sewing cushion of The Lacemaker.
In a special case of this loss of focus, Vermeer treats the highlights on shiny surfaces such as polished furniture or pottery as white or yellow circles. These again are artefacts of optical images: the bright spots spread out on the viewing screen into "circles of confusion". In Wheelock's words, the artist's "conscious attempt to exploit optical phenomena visible in a camera obscura - intense colors, accentuated contrasts of light and dark, and halation of highlights - suggest that Vermeer sought to recreate the impression of such an image."
My own research has provided new evidence about this question, from a different direction: from detailed study of the geometry of the rooms shown in Vermeer's interiors. It is possible to work out, with great precision, the three-dimensional arrangement of the spaces Vermeer depicts, by carrying out the conventional process of perspective construction in reverse. There are many objects in these paintings whose exact dimensions are known. Assuming Vermeer to have painted them at their true sizes, they provide a means for inferring the dimensions of the rooms as a whole. They include the maps Vermeer shows hanging on the walls, copies of all of which survive; some real paintings belonging to Vermeer's family; the mass-produced Delftware tiles used for skirtings; and various musical instruments including virginals recognisable as the work of the famous firm of Ruckers. It proves possible to demonstrate that the architecture of the rooms in some dozen pictures is the same (with minor variations in the patterns of floor tiles): that these all show one and the same room.
The best view of this room is given by The Music Lesson in the Queen's collection. A mirror hangs above the virginals, which reflects the girl's head turned slightly towards her visitor, the carpet-covered table, Vermeer's easel and stool and, at the extreme top left, a tiny fragment of the back wall of the room. The painting's theoretical viewpoint - the position in principle of Vermeer's eye - can be located in this space. It is indeed at the level of the head of someone sitting on the stool. The viewpoints for other pictures can be found in a similar way. They are all in the same general area of the room, although not exactly at the same spot.
Each painting takes in a slightly different angle of view. This angle can be carried back in each case, through the viewpoint, to define a rectangle on the back wall. I have discovered that, in at least six instances, this rectangle is the precise size of the canvas in question.
It is extremely difficult to explain this strange geometrical phenomenon, except in one way. Vermeer had a cubicle-type camera, whose lens was positioned for each painting at its theoretical viewpoint. With this arrangement he cast a full-size image on to the back wall of his room. I tested this hypothesis in two experiments. The first was to build a scale-model room, with furniture and figurines, and with a photographic camera - its plate in the plane of the back wall - playing the role of Vermeer's camera obscura. By this means my colleagues and I recreated several pictures photographically, including The Music Lesson.
The second was to rebuild the room at full size for a BBC television programme, again furnished as for The Music Lesson. This showed how it was perfectly feasible, using a simple lens readily available in 17th-century Holland, to create a sharp, bright and detailed image at the size of the painting itself.
And yet - despite his embracing the hypothesis of Vermeer's use of the camera - Wheelock would strongly resist the idea this discovery implies, that the artist might have traced large parts of his compositions wholesale from optical images. Throughout his book and the exhibition catalogue, Wheelock is at pains to insist repeatedly that Vermeer used the instrument as an incidental aid only. He uses the findings of X-ray analysis to show how Vermeer continually adjusted his compositions, overpainted features, moved the boundaries of objects; and argues from this that the artist was therefore not copying literal appearances. The artist selected, altered, manipulated, in the service of compositional objectives and "realism" of a higher kind.
One example among many is provided by Wheelock's discussion of The Music Lesson. He argues that Vermeer has artificially contrived the pattern of illumination in the picture. "That direct sunlight enters the room is clear from the bright rear wall and pronounced diagonal shadow falling from the window sill. Logically, however, sunlight that would create such a shadow would form similar ones at the juncture of the windows and the ceiling and behind the horizontal frame at the top of the lower window." My photographic reconstruction shows nevertheless that Wheelock's logic is wrong. The shadows in question fall in the model precisely as they do in Vermeer's painting. (The model was lit by large diffusers outside the windows to simulate north light, not by direct sunlight. We did not have to try hard to reproduce Vermeer's shadows; they fell out immediately.) The precise positions of a few shadows are not perhaps matters of enormous import. Nor would I want to argue that Vermeer simply copied everything he saw. The larger and more interesting question is why Wheelock, and other critics, should be so very anxious to save Vermeer from this charge. The explanation has to do, of course, with the old - and one would have thought long-resolved - debate as to whether photography can be an art. A false analogy is being drawn with a certain kind of modern photographic practice. We are used to thinking of journalistic photographs as fixing the fleeting moment, of catching accidental appearances. This overlooks the obvious potential of both camera obscura and modern camera as tools for meticulous and carefully considered composition.
Both instruments are means for projecting the three dimensions of space on to two-dimensional surfaces, for turning scenes instantaneously into pictures. One can imagine Vermeer going back and forth between his studio and his darkened cubicle, endlessly moving his furniture, tapestries, wine jugs and wall maps, adjusting the postures of his sitters, studying the fall of light and shadow in his camera image, the resulting perspective alignments, the "negative spaces" between objects - all of which, as Wheelock rightly emphasises, are so precisely judged in the final results. Wheelock argues that the "refinement" and "purification" of Vermeer's compositions are reasons for believing he cannot possibly have relied on the camera. To my mind they are just the opposite.
Such technical controversies have little relevance perhaps to an appreciation of the human and psychological subtleties of Vermeer's works. Insofar as the subjects of his painting are geometry, space and light however, they go to the very centre of his art.
Philip Steadman is professor of architectural and urban morphology, the Open University. His essay, "In the studio of Vermeer", was published in The Artful Eye, edited by R. Gregory et al (1995).
Editor - Arthur K. Wheelock Jr
ISBN - 0 300 06558 2
Publisher - Yale University Press
Price - £25.00
Pages - 229