BLOG/WRITING
Visual Intelligence, By Donald D. Hoffman
I'm giving this thoroughly enjoyable and fascinating book by Donald D. Hoffman a second reading at the time of writing this. Hoffman has put together a compendium of the algorithmic rules that determine how we humans, and probably animals, extricate information and construct form from the chaotic 2D splatter of colour and value on the back of our retinas. Hoffman lays out how rules apply and how other rules might over-ride other rules under specific circumstances. Some rules seem well understood both conceptually and in terms of brain and eye physiology but for others current research can only speculate why they work. Any visualist who has taken the time to put in the thousands of hours required to draw well could find this book fascinating through the recognition that most of the brain's visual construction rules are the very same rules visual artists apply when constructing and arranging coherent form in the pictorial space of a two dimensional surface. Drawing might be a sort of reverse engineering of our innate read-only visual intelligence into a writable system, that through practice and skill, can be applied with stylus or brush onto a sheet of paper or a canvas. This is why the book will probably appeal to artists, or at least those artists who have taken the time to draw...to describe form...exceptionally well. It probably won't appeal to conceptual artists, or those artists who prefer words or ideas to images. It also probably won't, or shouldn't, appeal to those learning to draw, that is students, because, I suspect, being too self aware of processes while learning skill could undermine necessary unconscious 'flow' and hobble you. But, once you have a chunk of a lifetime spent drawing behind you this book becomes magic because it so comprehensively lays out much of what you already know.
This is a pipe! As a visual artist, on the one hand, i've always been annoyed at the somewhat 'anti-art' notion that this 2D representation is 'not a pipe', and also and also annoyed at Magritte's title for the painting 'The Treachery of Images'. It's a representation of a pipe, so it is a pipe we are perceiving. Surely Magritte might be being nuanced and even ironic in his title, because, if we see a pipe, then it is for us in an ever so astonishing, quite delusional way, a pipe? Just writing or saying 'This Is Not A Pipe', like some kind of hex, doesn't make it not a pipe. Surely it is at least both a pipe and not a pipe? The treachery could as easily be of words, and as a visual artist god only knows I find words as treacherous as images. Here is the 'other hand'; if this painting of a pipe were not a pipe, then looking at an actual pipe before our eyes is also not a pipe because it's actually a delusional perceptual construction from a two dimensional splatter of light in the back of our eyes. It's delusional pipes all the way down. There is magic and mystery (and creative perceptual intelligence Hoffman shows) in seeing, especially in the realm between two and three dimensionality. British artist Michael Craig-Martin recently said in a Guardian item “A painting of a shoe looks nothing like a shoe, being entirely flat and made of bits of colour put together. The miracle is that we look at it and can see a shoe as clearly as if there was a shoe in the room with us." This is a shoe, if it fits, wear it, although ultimately Hoffman's fascinating research is probably suggesting, somewhat alarmingly, that if it's not a pipe or a shoe you are seeing then nothing is actually what it appears to be, but rather an inferred fabrication created to make ends meet as best as possible negotiating in a complex world. Seeing isn't necessarily believing, and this is something that learning how to draw well makes eminently apparent.
Reading the Hoffman book, working artists who draw and paint will recognize the uncommon 'accidents of viewpoint' like co-linear and co-planar alignments that you try to avoid in composition, or the straightforward drawing of objects on a table. Should you happen to encounter accidents of viewpoint in your field of vision when drawing and put them in your image they would introduce confusion into the 'reading' of your image by a viewer. So you move things around, or move your point of view, left or right, even a little bit. These occasional accidents of viewpoint when encountered in the real world confuse the brain; one of my favorite examples of this, one that I often gave to drawing students, was a story printmaker David Blackwood once told me. He was from a small outport in Newfoundland where residents had been telling tales of seeing a headless woman in a pale diaphanous robe in the moonlight in the surrounding barrens. In Blackwood's tale the instinct was generally to run away as fast as possible. Blackwood himself claimed to have encountered the apparition walking home late at night and managed to suppress the instinct to run and instead risk a firm gaze. Within moments the apparition became aware of his presence and what was the arse end of a white horse turned sideways and gazed back at him. I always like to recount that story when making the point that vision is generally trustworthy but not always and that seeing should not, in fact, be believing. I think visualists who draw are uniquely placed to recognize this reality.
Tibor Kolley's photo, above, of John Turner for The Globe and Mail was published just before the 1984 election, in which Turner lost to Brian Mulroony. I remember seeing this photo and thinking how it would have been an editorial choice of the Globe, a conservative paper, to undermine Turner's election chances. The photo is, of course, an accident of viewpoint in which graphic culinary designs on a wall behind him appear to emerge like horns from his head. Kolley probably had dozens of photos of Turner at this event, but the accident of viewpoint was cunningly chosen to suggest Turner had horns and looked foolish.
When accidents of viewpoint are applied to carefully designed illusions they deliberately 'confuse' the visual mind, or, at the very least, lead it to construct alternate images or alternate readings of form. In the well known Necker Cube (above), for example, your brain flips the aspect of the cube back and forth repeatedly when you stare at it, showing how the mind can 'construct' what you see in two separate ways. The book is full of such illusions illustrating how the rules apply and how they might be broken. The overarching rule seems to be the 'rule of general views' which is a kind of wager that rules out the prospect of the uncommon accidental view through probability. In a way all the rules are wagers on probabilities based on accumulated information and experience that the brain makes to interpret the retina; it is far less likely that things appear one way than another so the brain goes with the much more likely scenario. It's when an unlikely 'accident of viewpoint' occurs the brain occasionally loses the wager and sees something quite different from actuality, like a headless female apparition in the moonlight or a cleverly designed illusion.
A flat graphic and low contrast animation background I painted that will allow for a character to be easily read upon it.
This book is doubly interesting for me because not only did I work long and hard to learn to try to draw well, but I also worked as a background painter for animation. Backgrounds, in animation (or live action) have to do several things; establish mood and location without distracting from the interacting characters. This means they almost always need to be a completely comprehensible pictorial space for the brain. If, for example a theatre-goer experienced some kind of confusion about what they were seeing in the background it would capture their attention away from the narrative and characters. Everything in the background should generally make complete visual sense. As well it should stay in the background and not come forward in the pictorial space (or come forward if there is an overlay element in front of characters) and allow characters to be easily 'read'. This is usually achieved by reducing contrast in the background and increasing contrast in characters. Furthermore, most scenes in a film, animated or live action, are around one or two seconds long. So you also have a severe time constraint on being able to make sense of what you are seeing behind the characters. Awareness of 'the rules' of visual intelligence are therefore constantly referenced by background artists, knowingly or not, as they make sure that the background painting is easy to 'read' on the retina by ensuring there is no ambiguity for a brain wagering on probabilities about what it is seeing. And the background must stay in the background. This lack of ambiguity in ideal background paintings is somewhat different than what might occur in an illustration (or a painting) where a viewer might have more time to lavish attention; in fact, some ambiguity and effort might actually make an image more interesting as in, 'hey wait a minute, what is going on here'.
Spot the predator. A complex, 'contrasty' background with a camouflaged character upon it. This image is the opposite of a general animation background/scene or live action film scene.
Visual intelligence, the book, helped me place another visual interest I've had and casually read about that at some point will require a post of it's own in future. I've always been a little perverse. As a visual artist I worked ever so hard to make form extremely comprehensible to the eye and the mind, accept on occasions in my own work where I might have wanted to use ambiguity to create interest. So I always wondered about how artists might reverse the rules of visual intelligence to obscure or even annihilate perceived form. If you are fascinated by, and your work entails, constructing form it follows that you might be just as fascinated by processes that involve de-constructiong it. This is of course the realm of visual 'stealth'. Yehudi Lights on aircraft mimicking 'lights' on deep sea fish. Dazzle camouflage on ships at sea to confuse submariners. Camouflage on uniforms and aircraft...and animals. 'Counter shading' on military aircraft and animals. And of course it makes a lot of sense that much of the early work on obscuring form was done by artists, including the first person to describe and analyse counter-shading, American artist Abbot H. Thayer.
Counter shading in nature. Light striking the darker top of the shark, above, and shading the lighter bottom of the fish will tend to flatten it out and make it blend with the background better. This is of course taken with a flash which negates the effect. Counter shading was not dealt with by Hoffman, but it is the sort of related perceptual phenomenon that artists discovered earlier in the 20th century and independent of research psychologists like Hoffman.
A counter shaded WW2 Focke-Wulf 190. Lit from above the darker top is illuminated and the lighter bottom is shaded which has a flattening effect on it's form against a flat sky making the aircraft less visible and harder to spot laterally. Viewed from above the lit dark top would have a flattening effect against the dark landscape. And from below the shaded light bottom would have a flattening effect...in terms of approximate 'value', against the sky.
An illustration (above) from American artist Abbot Handerson Thayer's 1909 book on camouflage. Thayer was first to discover/identify countershading. The caterpillar, if illuminated from the side, would be darker on top and lighter on bottom, like the shark and the FW-190. In the above picture, with light coming from above, the larvae's form is flattened out making it more difficult to see. If illuminate from below, as in the bottom picture, the opposite effect is incurred. The caterpillar becomes easier to see. However, in nature it is far more likely for the caterpillar to be viewed right side up with the sun, the source of illumination, high above it.
Hoffman claims that humans (and animals) have extraordinary intelligence to be able to take the jumble of information on the retina and be able to extract shapes and then forms from it. But significantly more visual intelligence is required, I think, to not just do what everyone seems normally quite capable of doing, but also to take that perception of form and then describe it, as seen and visualized, by dint of sculpture, drawing or painting. This is what artists spend their...minimum...ten thousand hours doing when they learn how to draw. I find Hoffman's exuberant celebration of the normal a bit like the now ubiquitous and ridiculous neo-Duchampian 'everyone is an artist' dogma. It takes incredible embodied intelligence to learn how to draw well, and superb describers of form demonstrate even greater intelligence and skill than someone who can manage it adequately. The reason that the existing subject of camouflage fascinates a modestly skilled (by comparison to the pioneers of camo) artist like me, is, I suspect the same as what captivated the very forward thinking artists who first figured it out. If you know how to describe, and accentuate, form, as a trained artist does, it is therefore possible to use those same skills to deconstruct form. The pioneers of camouflage recognized nature doing this and copied and experimented and elaborated on the products of evolution.
Although apparently not actually applied during WW2, Yahoodi lights were experimented with and it was shown that they would break up the silhouette of an approaching aircraft and therefore significantly delay the ability to recognize it's approach.
Yahoodi lighting was inspired by the counter intuitive use, in nature, of deep sea fish to illuminate their forms (above). Evolution had 'discovered' over millions of years that deconstructing the form of a fish made it harder to recognize by predators or prey.
Dazzle camouflage was used on shipping in WW2 because it broke up the form of the ship. This made it harder for U-boat skippers, observing the target on the horizon, to estimate a target's speed and size in order to fire a torpedo to intercept it. No camouflage works one hundred percent. But it reduces the chance of successful visual targeting and identification which in nature, through evolution, selects for distinct traits passed on and magnified over many generations. The skunk, below, arguably has it's form deconstructed in much the same way as the shipping above. Baby fauns have their forms broken up by the 'dappling' pattern on their backs which helps hide them when laying low in the dappled shade of shrubbery. Tigers have peculiar parallel markings on their faces which arguabley allow them to 'loom' more effectively. Looming is worth looking up; a predator will approach it's prey carefully on trajectories that reduce the eye's capacity to recognize movement. By approaching 'straight on' and gradually looming larger, in peripheral vision, they seem to be able to make it harder for the prey's eye cells attuned to movement to be triggered. In tall grasses the patterning on a tiger's looming face might make it even more difficult to recognize. Worth noting, if you look up looming, that attack strategies designed by fighter pilots in WW2 involved approaching the target on trajectories that utilized looming that would reduce apparent movement. These strategic approaches to targets mimic the evolved behavior of raptors attacking airborne prey. None of these camouflage examples are directly covered in visual intelligence, but many of the mechanisms by which they work are. I include them because they are generally things that were recognized first by artists, and artists who know how to draw well, or skilled film makers will also recognize them as tricks of their trade.
Stealth aircraft, as below, are really about deconstructing form in a similar way to dazzle camouflage, but with microwave radiation rather than at the wavelengths of light. As a result they might be quite visible to the eye when in close proximity, but much less visible on radar at a distance. Radar is a technology that has formerly given great advantages in early warning of aerial and nautical attack.
Because of or in spite of my fervent enthusiasm for this book, as a visual artist I was struck by a glaring and annoying omission. Hoffman lists the 'tourists' who will appreciate and enjoy the journey through his book on visual intelligence as; people in marketing, advertising, graphic design; readers of popular science; undergraduates with undeclared majors; developers of virtual worlds; lawyers concerned with eyewitness testimony; and philosophers. But no mention of artists. WTF Donald! Art and artists, often working hundreds of years ago, are referenced on a number of occasions in the book for showing an awareness and understanding of how to apply the rules of visual intelligence to describing form. What kind of accident of viewpoint, blind spot, or illusion prevented you from including us artists among the sorts of people who might appreciate your writing and research on vision?