by Justin Manley
A large canvas stretches across a wall, empty except for a few rows of text: a title, and a person’s name. Also, a list of specifications: timestamps, hardware models, software versions. As I watch, strokes of grey and brown paint appear one by one, slowly burying the text under an incipient image.
The canvas is a digital projection, and the brushstrokes are animated by a computer program. Before the text disappears under a layer of paint, I can make out the signature of the artist, San Francisco-based Clive McCarthy. The software will continue painting tirelessly for weeks, a new image every few minutes, but in these first moments, painterly brushstrokes fused with a programmer’s colophon offer a perfect metaphor for the artist’s hybrid career.
McCarthy, born in the United Kingdom, worked for nearly two decades in the computer electronics industry and rose to the position of chief engineer at the semiconductor company Altera. In the late ‘90s, he quit to become an artist, and in 2004 earned an MFA from the San Francisco Art Institute. Since 2010, he has worked on a series of computer programs that endlessly make and remake paintings: some original, some based on iconic works like the Mona Lisa, others derived from the artist’s own photos or images downloaded from the internet. A solo exhibition at the San Jose Institute of Contemporary Art, Clive McCarthy: Electric Paintings, shows a rotating selection of these works, some projected onto walls, others displayed on monitors.
McCarthy is not the first artist to experiment with generative painting systems. In the 1960s, British painter Harold Cohen began tinkering with a robotic painter he called AARON. Cohen preprogrammed AARON with a vocabulary of symbolic forms and equipped the robot with sensors and algorithms to observe and react to each painting as it took shape. Cohen worked on AARON until his death in 2016. Since 2001, computer scientist Simon Colton has developed a system called The Painting Fool, which transforms photographs into paintings stroke-by-stroke. In 2011, artist Benjamin Grosser created a robotic painting machine that paints according to what it hears over an attached microphone. Colton uses his system to research artistic creativity, while Grosser is most interested in feedback. Cohen pursued feedback as the elusive key to creativity. McCarthy’s work, in contrast, is distinguished by his persistent attention to the materiality of painting and the processes of visual perception.
Perceptual skill
Seeing feels effortless and instantaneous: I open my eyes, and the world appears automatically and faithfully before me. The truth is more complicated. When I look around, what I see is informed by experience, modulated by mood, and filtered through shifts of attention. Painting captures all of this. In painting, the movement of the brush over the canvas records the way that the eye moves over a scene. At the same time, it leaves traces of the mood and sensibility and character of the painter. In portraying the process of painting, McCarthy’s work conveys the effort and time and idiosyncratic distortion involved in every act of perception.
Each work in the show challenges a different set of perceptual skills: distinguishing characters of text on a page, identifying objects, making sense of how the human body is positioned in space. These puzzles are laborious but irresistible: it feels as though each image is always on the verge of resolving into something recognizable. McCarthy’s work encourages a slow, insistent, tantalizing way of looking that awakens me to the skill and pleasure of everyday perception.
Optical character recognition
The first work I see when I enter the gallery is a painting McCarthy calls James Joyce’s Ulysses. The picture shows a grid of letters, ten squares across, six down. Each square is layered thickly with letters painted one over another in different colors. Looking at the painting, I remember a box of brightly colored wooden alphabet blocks I had as a child.
Searching for patterns, I notice that an invisible brush predictably moves across the painting: from left to right, top to bottom, though it sometimes skips a square. I identify a few words, but they don’t make a coherent sentence. It’s like a broken crossword puzzle or a bad game of Scrabble.
It takes me a surprisingly long time to realize that the painting is a text, that every brushstroke is part of a word, and that the words fit sensibly together. (I later learn from the artist that the painting cycles through the entire text of James Joyce’s Ulysses, 60 characters at a time). Yet this understanding does not in any way simplify the experience of the piece. The effort of making out each letter against the many-colored background is consuming: it fills my attention and leaves me just enough mental capacity to hold a few letters at a time in memory.
With my eyes overwhelmed, I find that enlisting my ears eases the task: I speak the names of the letters under my breath, listening for combinations that sound right, like a child learning to read. Decoding an entire panel is a monumental feat of concentration, one I’m not able to muster in the time I stand in front of Ulysses. It’s just as well: the letters are always changing, and in just a few minutes, they will be submerged under a new page of text, a new challenge.
Wooden alphabet blocks are puzzles, tools for attuning babies to the textual world. Ulysses rekindles some of the experience of the baby playing with alphabet blocks, where letters and words still appear as pictures before they turn into ideas. In this state of temporary preliteracy, I marvel at the skill it takes to identify each letter and to mark the boundaries of each word, for it makes the ease of ordinary reading seem magical.
McCarthy reawakens the visual puzzle of reading by slowing and intensifying its perceptual challenge. Each block in the painting is a multicolored palimpsest with portions of many different letters visible. This pile-up increases the difficulty of recognizing individual letters. Other patterns make it difficult to assemble words. McCarthy’s painter — the computer program that wields the brush — changes colors regularly, so that a single word might span multiple colors. When there is a space between words in the text, the painter simply skips over a block, allowing the previous letter to remain at the top. This makes it nearly impossible to determine the boundaries between words.
McCarthy makes painters, not paintings. He does not place individual marks on the canvas, but in writing the computer programs that execute each piece, he determines the rhythm and logic of the brushstrokes, the pace and shifting attention of the creator. Each piece is a complete oeuvre that conjures a ghostly human presence.
McCarthy slows the experience of reading by rationing the information of the text. Even a single letter is made of many brushstrokes, laid down one by one, and each word takes several seconds to accumulate on the canvas. It is easiest to follow the brushstrokes, matching their slow pace, for the movement of the brush aids my eye in separating each letter from its background. Typically, it takes only a fraction of a second to read a single word. In Ulysses, reading happens a hundred or a thousand times more slowly. This enforced slowness makes it possible to reflect on the visual experience of reading itself.
In amplifying the challenge of reading, McCarthy’s Ulysses also revives some of its forgotten pleasures. Letters on children’s wooden alphabet blocks have a tactile presence and sensuous materiality that they seldom reacquire in the adult world of the printed and digital word. The recollection of these childhood pleasures is Ulysses’ reward for the labor of reading.
Object Recognition
One busy evening at the ICA, I sit in front of McCarthy’s Painting the Internet, watching a picture come together. Two yellow and tan masses occupy the center of the picture, set off against a backdrop of rich blues. A tangle of brushstrokes in shades of brown sits along the bottom of the frame. Beside me, spectators begin to guess aloud the subject of the painting: a waterfall pouring over boulders, a birds-eye view of sunbathers lounging on inflatable pool floats. Someone ventures another guess: an art auction. Suddenly, I see the picture differently: two large, pale paintings stationed against blue velvet curtains and attended by a crowd of brown-haired bidders.
The images depicted in Painting the Internet come from a collection of 10,000 images downloaded from the internet. The piece presents these images as a series of distinct paintings, rather than a single work. At the beginning of each picture, the brush moves with inhuman speed, laying down tight squiggles of color along a roving path. Most images are completed or abandoned within three or four minutes; a few take as long as ten minutes. At the end of each episode, the simulated canvas slides to the right, taking the painting out of sight and exposing a fresh surface for the next image. Successive images are entirely unrelated: a dark crowd of people gathered around a campfire, a row of mannequins in a department store, a person standing with a bicycle. Each painting begins as an abstract jumble of brushstrokes, recognizable as nothing but computer-generated paint. The most exciting part of each image is the middle phase, when the primary forms have stabilized to suggest many plausible interpretations, none of them certain. If the painting is not abandoned — if it endures for four or eight or ten minutes and acquires enough detail — it may resolve into a recognizable scene. This shift, if it happens, is sharp and irreversible. One moment I am looking at blue and yellow and brown paint; the next, I see an art auction.
Every glance at a painting is an exercise in object recognition. Usually, however, the process of identification happens below the threshold of consciousness. Painting the Internet makes me aware of this process by lengthening it and repeating it endlessly. By filling in the scene gradually, McCarthy’s computational painter withholds the moment of recognition, keeping me suspended for minutes in a process that usually happens in the blink of an eye. Yet, even when sustained over several minutes, the effort of identifying a scene in this painting makes it easy to forget in the complacent satisfaction of recognition. Each iteration is entirely fresh, unrelated to those preceding it. The perpetual novelty of successive scenes is what allows the protracted experience of recognition to be repeated endlessly, each perceptual struggle just as laborious as the one before. This relentless repetition rouses me from complacency and fixes my attention on the experience of object recognition.
Each iteration is entirely fresh, unrelated to those preceding it. The perpetual novelty of successive scenes is what allows the protracted experience of recognition to be repeated endlessly, each perceptual struggle just as laborious as the one before.
In doing so, Painting the Internet allows me to observe the way my awareness changes as I recognize a scene. In the moment of recognition, I feel a thrill of confirmation, the satisfaction of solving a puzzle correctly. But something is lost as well as gained in solving the riddle of the painting. As long as the painting remains a formless mass of color that I struggle to describe, my mind is active. I imagine new ways of interpreting the brushstrokes, and my eyes flick over the canvas, searching for details that might confirm or repudiate a hypothesis.
After I recognize the subject of a painting, the quality of my gaze changes. Without realizing, I subside into complacent inattention. Although I still register the colors before my eyes, I no longer note the texture and pattern of the brushstrokes. Instead, I see a set of stock figures — auctioneer, bicyclist, mannequin.
Perhaps seeing the world as an artist simply means prolonging the period of inquisitive, active looking as long as possible — endlessly coming up with new questions and new puzzles to solve so that vision never lapses into inattention.
Pose estimation
At the back of the gallery, a piece called Gardening shows three views of a nude woman on a ladder pruning a tree. The subject is unchanging: always the same woman and the same ladder. As the painter works and reworks each panel, the woman’s pose, her place on the ladder, and our vantage continue to shift. Gardening challenges me to resolve the brushstrokes of pink and beige and grey into the proper configuration of woman and ladder — to identify arms, legs, breasts, thighs, and arrange them properly in space — before the painting is abandoned and “gessoed” over to make way for a new one.
At any moment, there is usually at least one panel in Gardening which is fully finished, which shows, blurred but unmistakable, a woman on a ladder, rendered in luminous layers of semitransparent glaze. Others appear in varying states of progress or abandonment, perhaps no more than a welter of brushstrokes against a backdrop of white gesso. Yet the finished panel contextualizes its companions and makes it possible to infer from even a small number of brushstrokes the outline of a human figure.
The rest is an old story by now: the slow pace of the brushstrokes allows me to savor the experience of pose estimation. But in this one, there’s more. The brushstrokes of Ulysses, Painting the Internet, and Gardening are all laden with feeling. But because of its singular, sensual subject, the changing sentiments of McCarthy’s painter come across most clearly in Gardening.
Coda: Sentiment Analysis
Where Ulysses feels patient and Painting from the Internet is restless, Gardening feels personal, even obsessive. In one scene, the gardener’s back is turned, and the painter articulates her bottom with a curving brushstroke that hugs the contours of her body. The stroke seems saturated with loving attention. As the figures of the woman and the ladder sharpen and the puzzle of pose estimation becomes only a memory, I begin to think about a different kind of pose: the attitude of the painter towards the subject.
In Gardening, each iteration of the painting begins in a frenzy of brushstrokes distributed evenly across the canvas. The brushstrokes go down impossibly fast, tracing the edges of figures in broken, impatient dashes of color. Quickly, though, the painter settles down to shape the figures in earnest, working at a more sedate pace. The brush might linger for minutes at a time on a leaf in the background while the gardeners are still only a cursory outline. Then, abruptly, thick strokes of gesso obscure the figures, and the painter begins anew on a fresh impression.
Although I know the painters in McCarthy’s works are always computer programs, I feel a human presence as I watch each panel in Gardening take shape. At the beginning, I envision a burst of inspiration, the brush attacking the canvas, placing marks everywhere. I imagine the painter’s eyes moving over the scene in the garden, just as my own eyes follow the brush. When for several minutes, the canvas seems entirely still, and I must lean close to see the brush worrying the same dark patch of leaves again and again, I imagine that the painter is frustrated. And when the scene is finally abandoned, I imagine the painter scowling, restless, and dissatisfied.
Imagining the body and mind of the painter feels natural. From our first finger paintings and experiments with watercolors in elementary school, we come to understand every picture as the product of human hands and feelings. Reading our own emotions is a skill, and in learning, we search everywhere for signs: in the way our bodies feel, in the things we say — and in the kinds of marks we make. Imagining the feelings of the painters behind McCarthy’s paintings is a natural exercise of this emotional perception, more subtle and ethereal than the guesses about the subject of a painting, but no less real.
Of all the perceptual skills exercised in the exhibition, this empathetic way of looking at paintings is the slipperiest, but it may be the most important for understanding McCarthy’s work. He makes painters, not paintings. He does not place individual marks on the canvas, but in writing the computer programs that execute each piece, he determines the rhythm and logic of the brushstrokes, the pace and shifting attention of the creator. Each piece is a complete oeuvre that conjures a ghostly human presence. This is the great challenge and pleasure of McCarthy’s work: the tantalizing feeling of proximity to the creative hand, the sensitive eye, and the changing moods of the painter.
# # #
Clive McCarthy: “Electric Paintings” @ San Jose Institute of Contemporary Art through March 22, 2020.
Justin Manley is a writer and programmer.