Secret Knowledge: Rediscovering the Lost Techniques of the Old Masters 
by David Hockney.
Thames and Hudson, 296 pp., £35, October 2001, 0 500 23785 9
Show More
Show More

David Hockney’s new study, Secret Knowledge, sets out a thesis with vast implications, both for the way we look at Old Master paintings and the way we think about painting’s relation to photography. The more attention you give the thesis, however, the more Hockney’s presentation starts to frustrate you. What you get is, first, a brisk illustrated lecture explaining how he hit on his ideas, a lecture that involves rushing every which way round the National Gallery, pointing out telling visual evidence and adding speculative asides. Next comes a gathering of supportive excerpts from art history; then, forming the bulk of the text, Hockney’s correspondence on the subject over the last two years, chiefly with the art historian Martin Kemp, author of The Science of Art (1990), a magisterial study of painting and optics. David and Martin, seemingly unedited, exchange chitchat, aperçus and mutual encouragement between Los Angeles and Oxford, artlessly and highly repetitiously: you really have to dig for the nuggets. An unexpurgated documentation of his thought processes is a less generous offer than Hockney seems to imagine.

The thesis is primarily historical: he wants to modify certain assumptions about the course pursued by Western art over the last six centuries. It’s hard to get the measure of what he’s saying without first recapitulating these assumptions. They are to do with the way in which Western painting has developed its claim to ‘imitate nature’ and I take them to run something like this.

Before the late 13th century, the imitation of nature, or the simulation of outward appearances, term it how you will, seems a relatively minor aspect of picture-making. What comes after, however, is a gradual accumulation of methods to capture ‘the look of things’, some more compatible with their forerunners than others, though none has been wholly supplanted to this day. To squeeze in six centuries under nine headings, these ways are: 1) from about 1300, illusionism – with Giotto, 3D qualities such as volume and recession start to loom large in painting. 2) Over the following few generations, observation: painters increasingly work with drawings taken from objects set directly before them (what Hockney calls ‘eyeballing’). 3) In 1420s Florence, perspective: Brunelleschi shows artists how to draw collections of objects as seen from a single static viewpoint, so that the picture becomes a ‘window’ set between them and the viewer. 4) In 1430s Flanders, oils: Campin and van Eyck take up the medium to expand the tonal range and textural diversity of painted objects. These ‘Northerners’, however, show each object coming forward separately to meet the roving eye of a mobile viewer. About eighty years later, we get 5) ‘painterliness’: Giorgione, emulated by Titian, leaves off applying colour to already drawn objects to rely on the open-ended, fluid brushstroke, as a way of prompting the viewer to share the painter’s pleasure in the process of conjuring up the object. Then in 1600 there is the phenomenon known in its own time as 6) ‘naturalism’, when Caravaggio starts to treat painting as the reception, rather than the construction of appearances; that’s to say, as a way of capturing the light that falls on the bodies in his studio, rather than of using light to enhance the volume of previously defined objects.

During the next two centuries: 7) this principle of light-catching naturalism is extended through its pairing with the ‘painterly’ manner (Velázquez, Hals, Chardin); through the development of plein-air practice (Velázquez, again, through to Constable); and through the assistance of optical devices that project images through a single aperture onto a plane where they can be transcribed (Vermeer, Canaletto). These devices climax, you might say, in 1839, with 8) the advent of photography. Now that the recording of appearances can be a mechanical process, it will increasingly be regarded as best done mechanically. Finally, during the later 19th century, we arrive at 9) Impressionism and its offshoots: Monet and then Cézanne bring to a head a century-old tendency for painters to treat outward appearances as the inward experiences of an embodied, mobile and motivated viewer. Painters working in Cézanne’s wake move decisively away from the notion of copying ‘what’s out there’.

This potted history concentrates, of course, on a single strand in Western art, disregarding the factors that stop painting ever merely ‘imitating nature’: the demands to emblematise, to idealise and to deliver an autonomous aesthetic experience. You might broadly argue that these were in the ascendant during the 20th century and, with very different results, during much of the 16th and 18th. But these concerns are peripheral to Hockney’s main theme.

This said, here is how Hockney, alternately spurred on and reined back by Kemp, seems to want to change the history of the imitation of nature. Stages 1) and 2) can stay; he’d like to play about with stage 3), but he’s not quite sure how. What he really wants, however, is to insert 7c) into the description of stage 4). That is, he asserts that the adoption of oils in 1430s Flanders went hand in hand with the use of image-projecting devices. The remarkable new look of van Eyck’s and Campin’s object depictions, he contends, is the result of their transcription of the inverted images that can be projected by concave mirrors. He ties this up with the way their pictures are organised. The projections you get using a concave mirror show quite small excerpts from a reflected scene; you need to bring together a multiplicity of them to create a whole composition. Many sharply focused small areas of transcription are then linked up in a composite perfectly adapted to a viewer’s roving attention.

From this point on, Hockney’s history of the imitation of nature wanders off along a rather different pathway before reaching the same finishing line of Cézanne: 5) The startling verisimilitude of the Flemish masters’ projection-based oil paintings sets new standards for other painters in the Northern tradition such as Holbein, whether these artists actually work with a concave mirror or by relying on the ‘eyeballing’ draughtsmanship of the Florentine tradition. Brought to Italy by Antonello da Messina, the secret technique may influence Leonardo and Giorgione. 6) Circa 1600, Caravaggio, in mid-career, gets hold of a new large lens, which permits transmission of a much broader image. His subsequent ‘cinematic’ approach to figure composition is based on the projections so cast. After which, concave mirror projections become a forgotten technique. 7) Velázquez, Hals and Chardin are all impelled towards their painterly naturalism by the visual qualities of soft-focus lens projections. Alongside the continuing eyeballing tradition (Rubens, Carracci, Rembrandt), lens-based devices (camera obscuras etc) play an important part in early modern portrait practice, from the time of van Dyck to that of Ingres (whose unerringly precise, tiny pencilled vignettes of sitters in 1820s Rome first led Hockney to explore the possibility of lost technical resources). 8) Come 1839 and photography, hand-and-lens techniques are quite soon forgotten. But 9), reacting to the advent of mechanical picturing, Manet, the Impressionists and Cézanne reassert the distinctiveness of painting by emphasising the ‘awkward’, hand-made mark and the bodily presence of the painter as active observer.

As to what happened to picture-making in the century of his birth, Hockney is filled with gloom. But his thoughts about the present predicament can wait: what is his evidence for this bold rewriting of the past?

There isn’t a written word to prove it, any more than textual evidence that refutes it. The projection techniques were known about at the times Hockney needs them to be but there’s no decisive documentary record of any named painter using them earlier than the 18th century. But this is what puts the swagger into his proposition: don’t just read, he tells us, look. Art historians have chased up every last document to embroider the cherished recital of artistic achievement but, he claims, they have forgotten to search the works themselves for clues as to how they were actually made. Looking on the paintings as a painter, he believes he has stumbled on the methods the Old Masters unspokenly employed from the 15th century onwards. Why unspokenly? Principally, because artists, as he can attest, are cagey: they like to keep their tricks to themselves.

At the outset, Hockney bought a camera lucida and with practice found a way to achieve results in drawing comparable to Ingres’s. From the conclusion that Ingres owned such a device – a fairly uncontroversial hypothesis – he leaped back in time to see who else in the canon might have made use of such assistance. As he explored the question, a conversation with the optical scientist Charles Falco opened his eyes to the existence, hitherto unknown to him, of concave mirror projections. With Falco, he started to analyse how paintings could be assembled from them, and then to find ways of distinguishing different eras in projection technique.

Repeatedly, Hockney goes out of his way to stress that this technique relies on both an optical device and a skilled hand, at least until the fateful stage 8). With pre-photographic technologies, the projected light makes information from the 3D world enticingly available on a flat surface, but the capacity to make good use of this information depends on a prior mastery of eyeballing, or direct drawing from the object. A powerful eyeballer may be able to produce an effect very similar to that of an image transcribed from projected light, although, so the argument runs, he would be impelled to do so only by the startling authority of such images. But this leads to the presiding problem: how, looking at the Old Masters, can we definitely isolate the traces of optical procedures?

Hockney and Falco think up some quite bad ways to do so. Their analysis of various paintings as composites leads them to probe at the joins between projections; when van Dyck’s or Chardin’s figures seem out of proportion, they suggest, they register the strain of patching together different focal areas. But, as a painter, I know I can eyeball a figure drawing as disproportionate as any they cite if, as I often do, I concentrate on the most complex passages, such as the head and hands, to the neglect of what connects them. Falco calculates the way that the perspective co-ordinates shift across paintings, contending that this is caused by the painter shifting his mirror or lens. He goes to town on one of the earliest still lifes, painted by Hans Memling c.1485. A vase of irises and columbines sits on a tabletop covered by a patterned rug, whose repeating units are seen to recede in two horizontal rows from the table’s forward edge. The vanishing point of the front row is slightly different from that of the back. For Falco, only an obscurantist would claim ‘that Memling, rather than using a camera obscura, inexplicably decided to change the perspective part way through’, so that ‘anyone arguing at that point that Memling didn’t use a lens would be a person too thick to worry about.’ Falco will be unruffled, therefore, when I point out that if I check the co-ordinates of my own freehand drawings from life, they frequently shift from level to level (Florentine perspective, in any case, was probably not a major issue for a Fleming in 1485); moreover, Memling’s flower-jug, with a vaguely conceived highlight and a painted decoration that doesn’t quite sit on its body, cannot possibly have been transcribed from projection; the flowers, finally, are evidently, if very beautifully, adapted from a botanical pattern-book. Don’t just measure, look.

Hockney is very keen on the problem of the way patterns sit on objects. How – it’s an old question – did the likes of Bronzino get all those fancy hearts and curlicues to flow so precisely and persuasively over the ruckled satin worn by Eleonora of Toledo? How long did she have to sit there? How else could he have done it, Hockney asks, if not with optics? But optics don’t really bypass the task at hand, which is one of visual analysis. First, you need to model the fall of light over the whole garment, then, maybe on an overlaid sheet, make a linear note of the way the bands of colour cross its folds, pointing out where the highlights are and where the colours glow strongest; it’s a two-stage process that will take at least two sessions of painting. Since the princess can’t wait, take a note of her pose, then borrow the dress and drape it on a dummy or paid assistant.

All of this applies whether you are eyeballing or working from a projection, but it hasn’t much to do with the precepts of 20th-century British art-school training. Post-Cézanne, post-Coldstream drawing procedures of the sort Hockney probably grew up under ask for a systematic, rationalised response to all visual stimuli; they don’t initiate you into a warren of compartmentalised tricks of the trade, which is what pre-modern art training seems to have been largely about. Though his fascination has turned from the moderns to the ancients, Hockney retains the modern hankering for a single master-key to artistic procedure, turning to optics as if they would grant artists uniquely privileged access to verisimilitude.

This blinds him, despite his experience as a painter, to some of the practical problems of oils. He writes as if the chandelier that van Eyck painted in the Arnolfini Marriage was brushed on directly from a projection. If projected, the ornate brass would have appeared on the ground of van Eyck’s panel chiefly as a set of highlights, but highlights would be quite literally the last information he would require for the rendering he’s giving it, using a method which proceeds upwards and downwards from the mid-tones; it would be like baking a cake by starting with the icing. Likewise, Hockney’s zeal leads him to be very selective in his choice of evidence. Next to the van Eyck in the National Gallery sit a nameless burgher couple of the 1430s, painted by Campin on matching panels. Hockney focuses on the ‘modern’-looking husband, but not on the wife; she, though equally closely scrutinised, has clearly had her features finely but firmly rectified to align them with current standards of beauty. Couldn’t the hand that idealised her have also generated the complex morosity of her Belgian business partner? It’s good to demystify Old Master practice as far as we can, but with Campin and van Eyck, there remains, however you care to look at it, a stupefying mystery, or mastery of linear control. If they used projections, they didn’t need to.

Not trusting to migrants from the ‘primitive’ Western Pacific to have dreamed up the megalithic statuary of Easter Island, Thor Heyerdahl created a balsa raft of the sort he thought might have been sailed across the ocean by ancient Peruvians, bearing civilisation: he made a bestseller out of the expedition, but his diffusionist hypothesis of 1950 now seems completely unwarranted. Is Secret Knowledge a Kon-Tiki of art studies, an intriguing but superfluous recourse to Factor X? It isn’t. It will – or at least it should – change our views of Western painting because its central expert evidence clinches the case for any with eyes to see. Speaking as a draughtsman, Hockney looks at the quality of marking in the Ingres drawings that first interested him; refers it forward to the 1960s line sketches of Andy Warhol; refers it back to Holbein’s English Court drawings of the 1520s and 1530s; refers yet further back to the one preparatory silverpoint known to come from van Eyck’s hand; and forces you to see that the same thing is happening across five centuries. In each case, the border between light and dark has been lying visible to the artist on the paper before him, and rather than will it into being, his pencil has merely followed it, unpressured and unfaltering. In the nakedness of the drawing, the relaxed muscular record of the tracing hand is unmistakably different from that of the forward-groping, carving-and-correcting eyeballer, struggling with his preconceptions and expressive or idealistic yearnings. On the strength of this demonstration, the use of projections seems quite certain in the 1530s; most likely a century earlier.

A further sign of projection in this range of drawings, one which Hockney mentions only glancingly, is the presence of anomalous, undigested information. Transmitted light doesn’t convey an object precisely in the form you thought you understood it, as anyone tracing from a snapshot will soon find; it snags and spreads over the little bumps and dips of bodies in ways that you couldn’t have foreseen. The drawings of Holbein and Ingres are scattered with small records of these oddities, folds of clothing going nowhere, isolated facial features in a desert of over-exposure. You may even see one in the strange chink of light that cuts the bridge of the Doge’s nose in Bellini’s famous National Gallery portrait of c.1503 – though for practical reasons, and whatever Hockney says, I think that most paintwork before Caravaggio would have been done at at least one remove from optical tracing. When it comes to the history of early modern naturalism from 1600 onwards, however, his thoughts about lenses seem to converge quite closely both with the look of the pictures and with the drift of contemporary art studies.

Historically, then, this book says something important. It floats many claims that probably can’t be sustained; but in its wake, that list of assumptions about the course of Western painting will need to be revised. When it comes to the modern era proper, however, Hockney’s argument becomes one about values rather than visual evidence. If we buy his narrative, eyeballing and optical projection have coexisted ever since the Western idea of ‘art’ came into being. He loves their interplay and the display of manual mastery of which the imitation of nature once consisted. But that project started to go off the rails in 1839. Photography and its sequelae, film and TV, began to alienate humanity from its own picture-making; and to seek out the deepest roots of the problem, he looks back to eyeballing Florence, with its principle of the single specified viewpoint, rather than to projection-favouring Flanders, with its multiple mobile foci. The fixed observer was destined ultimately, he reckons, to issue into the present as the controlled consumer. The static viewpoint posited by Brunelleschi’s geometry has become that of the couch potato.

Our images have become at once too many and too narrow, constricted by the tunnel vision of the camera. They induce fatigue and demoralisation and that iconoclastic urge so common to the art of his century, which so often threatened to part company with imagery altogether. But riding over the horizon just in time to save Hockney from terminal cantankerousness, here comes the digitalised image. Hurrah for manipulation! ‘The hand is back inside the camera’: imagery can once again be reclaimed for human participation, for conscious, creative interaction. Maybe this new ‘post-photographic age’ will reunite our divided heritage of picture-making.

Or maybe not. What Hockney ignores is that we have a need, part of the time, to be alienated by images; and that this is the habitual function of the light-projected image whose early history he reconstructs. We have a nagging urge to reach out and touch, through pictures, something that has nothing to do with our own expectations or creative participation. What is strange, what is anomalous, what is startlingly modern-looking (to use one of Hockney’s criteria for reliance on optics), might just be factual and worthy of our trust. Through the history of images, going back from the CCTV and photojournalism to the naturalism of Vermeer and Caravaggio and ultimately to the Arnolfini Marriage with its self-proclaimed evidential status (Johannes de eyck fuit hic, 1434), the claim that we are witnessing a record of transmitted light has always catered to this impulse. Whether we actually are witnessing a direct transcription and what kind of editing has occurred may always be a matter for suspicion; but somehow this taste for truth will have to be served even when all is digitalised.

And yet this longing is at odds with the activity of making, which is where art has to start. People make pictures with their minds and hearts and muscles; and from pictures, people want to make themselves money and fame. Western painting has developed alongside commodity capitalism, as a specialism in luxury properties. In this light, the imitation of nature is a deception the painter practises, not so much on the viewer, as on his own motivating rubric. I tell myself that I want you to look on this painting as if it were an unwilled work of nature; but really, I want you to look on it as the work of a single, supremely wilful individual, myself. No wonder the old projectors were concerned to cover their tracks.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences