Preconditions

1945–1960. Cybernetics, oscilloscope art, mechanical analog computers. The groundwork before the mainframe. Before any artist had a mainframe, the apparatus had to exist. From the late 1940s through the early 1960s, an unusual coalition assembled it: Norbert Wiener and the Macy cybernetics conferences arguing that feedback was a general theory; Ben Laposky photographing oscilloscope traces in Cherokee, Iowa with a camera he had built himself; W. Grey Walter wiring up the cybernetic tortoises in Bristol; the Whirlwind team at MIT lighting the first real-time CRT display in 1951; John Whitney Sr. converting a war-surplus M-5 anti-aircraft gun director into the analog motion-graphics rig that would make Catalog (1961). None of this was art-world activity. It was engineering, neurology, military surplus, and amateur fascination — and yet by 1960 the technical, conceptual, and aesthetic vocabularies that the next sixty-five years of computational art would draw from were all in place.

Part of Digital Art

8 moments in this segment.

1. 1948 — Norbert Wiener — Cybernetics: Or Control and Communication in the Animal and the Machine. Artists: Norbert Wiener · Year: 1948 · Medium: Foundational theoretical book · Substrate: Printed book — mathematics, philosophy, biology · Place: Cambridge, Massachusetts, USA / Paris · Institution: MIT / MIT Press (originally co-published with Hermann & Cie, Paris) In 1948, the MIT mathematician Norbert Wiener published Cybernetics: Or Control and Communication in the Animal and the Machine. The book is dense, technical, and almost unreadably mathematical in places; it also invented a field. Wiener's argument — that feedback loops are the common structure of biological, mechanical, and social systems, and that the mathematical tools for studying one are the tools for studying all — gave later generations of artists and theorists a vocabulary for thinking about systems, control, responsiveness, and information. Every subsequent moment in this archive sits, in some way, within the conceptual space Cybernetics opened. The book sold unexpectedly well. Translations into a dozen languages followed within the decade. Artists read it — Gordon Pask cited it, Nicolas Schöffer cited it, the Gruppo T in Milan built their early work in explicit dialogue with it. By the 1960s, the word cybernetic had become a modifier for art practices that took system behavior as their primary subject — cybernetic sculpture, cybernetic poetry, cybernetic art — with Wiener's book as the theoretical anchor. The 1968 Cybernetic Serendipity exhibition at the ICA took his title verbatim. Wiener himself was uneasy about some of the consequences of the field he had founded. A 1960 essay, Some Moral and Technical Consequences of Automation, warned about the political risks of automated systems deployed faster than their implications could be understood. The warnings have aged well. He died in 1964. Lore Wiener trained at Harvard, Cambridge, and Göttingen, and had been a Harvard doctorate at eighteen and an MIT faculty member at twenty-four. Cybernetics was written in partnership with the Mexican physiologist Arturo Rosenblueth, whose influence on the book's treatment of biological feedback is extensive. Wiener's autobiography Ex-Prodigy (1953) remains a readable source for the period. He died in Stockholm in 1964 at sixty-nine, while on a lecture tour. Sources • Wiener, Norbert — Cybernetics (MIT Press, 1948; revised 1961) • Conway, Flo; Siegelman, Jim — Dark Hero of the Information Age (Basic Books, 2005) • MIT Institute Archives — Wiener papers
2. 1948 — W. Grey Walter — Elmer and Elsie (the Tortoises). Artists: W. Grey Walter · Year: 1948 · Medium: Autonomous cybernetic robots — two-neuron analog circuits · Substrate: Vacuum-tube amplifiers, photocells, motors, rechargeable batteries, protective shell · Place: Bristol, UK · Institution: Burden Neurological Institute In 1948–49, the British neurophysiologist W. Grey Walter built two small wheeled robots at the Burden Neurological Institute in Bristol. He called them Elmer and Elsie — jointly, the Tortoises — and demonstrated them publicly over the next several years. Each Tortoise was a simple analog device: a photocell eye, two vacuum-tube amplifiers, and motors, connected in a feedback loop that produced lifelike behavior. A Tortoise would move toward moderate light, turn away from strong light, and when its batteries ran low, seek a lighted charging station to refuel before continuing its wandering. What Walter demonstrated was that extraordinary complexity of apparent behavior could emerge from minimal circuitry. The Tortoises had two neurons. They were not programmed in any modern sense. Yet they appeared to explore, hesitate, decide, dance with each other when placed together. Walter described his experiments in scientific papers, but he also presented the Tortoises to general audiences — they appeared on British television in 1950, in magazine articles, at the 1951 Festival of Britain — where they were received as small mechanical animals rather than as physiology demonstrations. The Tortoises were the first robots explicitly built to be observed as living creatures. The reception framework they established runs through every subsequent responsive-sculpture practice — Ihnatowicz's Senster, Pask's Colloquy of Mobiles, the entire lineage of cybernetic and robotic art. Walter's two-neuron demonstration remains the cleanest example of how simple systems produce emergent complexity. Walter died in 1977. Elmer and Elsie are held by the Smithsonian and the Science Museum London. Lore Walter came from an unusual background — American-British, trained in both medicine and physics — and was known for his work on EEG (electroencephalography) before the Tortoises. He called his robots machines that think in popular press appearances. He later wrote a popular book, The Living Brain (1953), that became a minor bestseller. The original Elmer and Elsie were damaged in storage; museum-held units are in most cases restored or reconstructed. Sources • Walter, W. Grey — The Living Brain (W.W. Norton, 1953) • Holland, Owen — "The First Biologically Inspired Robots" (Robotica, 2003) • Science Museum London — Tortoise collection record
3. 1950 — Ben F. Laposky — Oscillons. Artists: Ben F. Laposky · Year: 1950 · Medium: Photographs of oscilloscope traces · Substrate: Cathode ray oscilloscope, analog waveform generators, long-exposure film · Place: Cherokee, Iowa, USA · Institution: Sanford Museum (first exhibition, 1953) Ben F. Laposky, a mathematician and amateur engineer from the small town of Cherokee, Iowa, began in 1950 to photograph the traces of a cathode-ray oscilloscope with a handheld camera set to long exposure. The oscilloscope was an ordinary laboratory instrument: fed a sinusoidal signal on its horizontal and vertical axes simultaneously, it drew the characteristic Lissajous figures — curls, rosettes, overlapping ellipses — that every electrical-engineering student had seen. Laposky's innovation was to compound multiple signals, adjust the phase and frequency relationships with a bank of audio oscillators he had built himself, and photograph the resulting composite trace onto color film using color-filter attachments over the camera lens. He called the photographs Oscillons. The Oscillons are, by reasonable consensus, the earliest photographs of electronic images deliberately composed as artworks. Laposky made roughly fifty distinct compositions between 1950 and 1953, mounted them in a traveling exhibition (titled Electronic Abstractions) that opened at the Sanford Museum in Cherokee in December 1952, and sent the show on a circuit through regional American museums and science-education venues over the subsequent decade. The photographs appeared in Scientific American in 1953. They appeared in the catalog of Cybernetic Serendipity at the ICA in London in 1968, sixteen years later, as the earliest works in the first major museum exhibition of computer-adjacent art. What matters in this archive is a specific claim embedded in Laposky's procedure. He did not invent the Lissajous figure. He did not invent the oscilloscope. He did not invent the photographic emulsion. What he invented was the practice of treating electronic signal-composition as a visual-art medium. The oscilloscope trace was not an illustration of a mathematical function; it was the picture itself, and the picture was the product of a set of signal-generation decisions that were indistinguishable from compositional choices. Everything downstream — the analog video-synth work of the Vasulkas, the signal-based pieces of Paik, the entire lineage of algorithmic visual work — rests on the Laposky premise. The Sanford Museum still holds the complete archive of Oscillon photographs. The Victoria and Albert Museum's Computer Art Collection, established in 1969, has a representative set. Laposky died in Cherokee in 2000 at age 86. Lore Laposky ran a small machine shop in Cherokee County and worked on the Oscillons nights and weekends in a backyard studio he had built himself. He had no art training and essentially no contact with the coastal art world; he had been reading Scientific American his whole life, and the Oscillon project began as a private experiment in visualizing the signals he was already generating for fun on his home-built oscilloscope equipment. The Sanford Museum — a small regional natural-history and art museum in his hometown — took his work seriously when no gallery or museum anywhere else would. The 1952 exhibition was funded by Laposky himself. He paid the gallery to hang his work. He continued to make Oscillons, on and off, for the rest of his life, and never left Cherokee County. The complete Sanford archive was digitized in the 2010s; the originals are 35mm color slides, some still in their original Kodak cardboard mounts. Sources • Laposky, Ben F. — Electronic Abstractions (Sanford Museum, Cherokee IA, 1953). • Reichardt, Jasia (ed.) — Cybernetic Serendipity: The Computer and the Arts (Studio International, 1968). • Sanford Museum, Cherokee IA — Ben F. Laposky collection.
4. 1950 — Theseus — Claude Shannon's maze-learning mouse. Artists: Claude Shannon · Year: 1950 · Medium: Electromechanical demonstration — magnetic mouse navigating a maze, learning routes · Substrate: Electromagnetic mouse (Theseus) moved by relays beneath a 25-cell maze; telephone-relay logic encoded learning behavior · Place: Murray Hill, New Jersey, USA · Institution: Bell Telephone Laboratories In 1950 Claude Shannon, already the author of the paper that founded information theory, built a mouse. He called it Theseus. It was the size of a pack of cigarettes, cut from a block of wood and a set of relays, and it could solve a 5×5 maze. More important, it remembered the solution. Run it a second time against the same maze and it walked straight through without touching a wall. Theseus is not, strictly, a computer; it is a piece of electromechanical control built out of telephone-exchange relays Bell Labs had lying around. A grid of magnets under the maze floor pulled the mouse; a bank of about ninety relays beneath the maze — the real computer — stored the decisions. When the mouse bumped a wall the relay bank recorded which junction and which direction, and on the next trial it took a different branch. After one complete run, the stored state was a lookup table for every cell in the maze. What makes Theseus historic is not the mouse-hardware but the claim it made out loud, years before machines could play chess or translate a sentence: that a constructed object could learn from experience. Shannon filmed a demonstration for television — the mouse runs randomly, hits walls, backtracks, and then, on the second pass, glides through — and the film circulated widely through the nascent cybernetics community. It gave the word "learning" a concrete mechanism to point at. Theseus sits in this archive as a precondition, not a first work. Shannon was not making art. But the idea that a machine's behavior could carry memory, that memory could be edited by contact with the world, and that the editing could be shown to an audience — those are the premises on which half of the first-generation artists in this archive built their practices. When Harold Cohen later asked what it would mean for a machine to draw, he was inside a conceptual space Theseus had already opened. Lore Shannon kept Theseus on a shelf in his MIT office for the rest of his life. The maze was reconfigurable — the movable walls slotted in, and visitors were invited to redesign it and watch the mouse solve the new one. He used to demo it at cocktail parties. Shannon's own favorite demo was a cruel one: he would let the mouse run a maze once to learn it, then secretly rearrange the walls, and watch the mouse barrel confidently into the first new wall — his joke about the brittleness of learned behavior. The original Theseus is now in the MIT Museum. The surviving demonstration reel, black-and-white and shot on a tabletop at Bell Labs, is one of the earliest films of a machine being watched learn. Sources • Shannon, Claude E. — "Presentation of a maze-solving machine" (Trans. 8th Conf. on Cybernetics, Josiah Macy Jr. Foundation, 1951). • MIT Museum — Theseus maze-solving mouse (object file). • Nahin, Paul J. — The Logician and the Engineer (Princeton, 2013).
5. 1951 — Jay Forrester (project director) — MIT Whirlwind real-time display. Artists: Jay Forrester (project director), Robert Everett (associate director) · Year: 1951 · Medium: Real-time graphical computer display · Substrate: Whirlwind I digital computer with vector CRT display; magnetic-core memory · Place: Cambridge, Massachusetts, USA · Institution: MIT Lincoln Laboratory / MIT Servomechanisms Laboratory Whirlwind I, the real-time digital computer developed at MIT from 1945 and operational by 1951, was not designed to make art. It was designed to solve the problem of real-time flight simulation for Navy pilots, and later to track Soviet bombers through the SAGE air-defense network. But it produced, as a byproduct of its architecture, the first real-time graphical display in computing history: a cathode-ray tube on which the computer could draw in real time, at human-comprehensible rates, under program control. The Whirlwind display is, technically, the direct ancestor of every computer-graphics display since. The Sketchpad demonstration that Ivan Sutherland would mount on the TX-2 in 1963 ran on hardware whose architecture descended directly from Whirlwind. The Stromberg-Carlson microfilm plotters that Bell Labs used for their 1960s computer art ran on CRT technology that MIT had proven. The contemporary frame-buffer graphics pipeline, through the Evans & Sutherland era and beyond, is continuous with the Whirlwind architecture. On April 20, 1951, Whirlwind was featured on Edward R. Murrow's See It Now television program. Project director Jay Forrester demonstrated the computer drawing a rocket's trajectory on its display, in real time, as Murrow spoke. This is the first time a general American audience saw a computer display real-time graphics. The moment is not widely remembered in art history, but every subsequent computer-art development depends on the hardware Murrow's cameras documented. Lore Whirlwind was a massive project — tens of millions of 1950s dollars, a full building at MIT, a team of dozens of engineers — and it was close to being cancelled multiple times before SAGE justified its continuing development. Whirlwind's magnetic-core memory, developed by Jay Forrester's team, was later licensed to every computer manufacturer and generated royalty income sufficient to underwrite MIT research for decades. The See It Now broadcast is preserved in the CBS archive. Whirlwind itself was retired in 1959; fragments are held at MIT Museum and Computer History Museum. Sources • MIT Lincoln Laboratory — Whirlwind technical reports • Redmond, Kent C.; Smith, Thomas M. — Project Whirlwind (Digital Press, 1980) • Computer History Museum — Whirlwind collection
6. 1957 — Lejaren Hiller — Illiac Suite for String Quartet. Artists: Lejaren Hiller, Leonard Isaacson · Year: 1957 · Medium: Musical composition for string quartet — score generated by computer · Substrate: ILLIAC I mainframe; Markov-chain and rule-based filtering algorithms; score printed for human performers · Place: Urbana, Illinois, USA · Institution: University of Illinois at Urbana-Champaign Lejaren Hiller and Leonard Isaacson composed the Illiac Suite for String Quartet in 1957 on the University of Illinois's ILLIAC I computer, and it is the earliest completed composition in Western music whose pitches, rhythms, and internal structure were determined by a computer program executing composition rules. The suite has four movements, each written to demonstrate a different procedure: first-species counterpoint, chromatic modification of triadic harmony, rhythmic and metrical transformation, and Markov-chain-based generation using probabilities derived from traditional harmonic progressions. The output — pitches on punched cards — was transcribed by Hiller into standard musical notation and premiered by a string quartet at the University of Illinois in August 1956, with the full four-movement suite completed by early 1957. Hiller was a professional chemist before he was a composer. He had earned a PhD in chemistry from Princeton in 1947, worked at DuPont, and taken a position in the University of Illinois chemistry department in 1952 before shifting, at age 34, toward composition, eventually founding the university's Experimental Music Studio in 1958. Isaacson, his collaborator, was a mathematics graduate student. The ILLIAC I — one of the first-generation American academic supercomputers, 2,800 vacuum tubes, roughly 1,024 words of memory — was shared by the entire university research community; Hiller and Isaacson had to work at night to get enough machine time to run the compositional experiments. What makes the Illiac Suite matter in this archive is that it is the first publicly-realized case of what would become, a decade later, the European Stuttgart-school argument in computer graphics: that a machine can be programmed to execute a set of constraints, produce an output that satisfies those constraints, and that the output can be assessed aesthetically as a work of art. The suite's musical language is not radical — it is largely tonal, recognizably in a post-Bartók idiom — but its procedural logic is. Hiller and Isaacson published their methods in Experimental Music: Composition with an Electronic Computer (1959), the first methodological treatise on algorithmic composition. The suite has been performed continuously since 1957 and is in the repertoire of serious contemporary-music ensembles. It is also a staple reference for any serious genealogy of computer music. Lore Hiller and Isaacson worked on the ILLIAC at night because the machine was oversubscribed by physics and engineering users during the day. They would book three-hour blocks starting at 2 AM, carry their punched-card decks across campus in paper boxes, and spend the first hour finding and correcting the single-character bugs that had caused a previous run to crash. Hiller later said the suite's first movement was more-or-less generated in a single night of successful runs; the fourth movement, which used Markov-chain probabilities, took nearly six months of nightly sessions as Hiller and Isaacson iterated on the probability tables. The ILLIAC I itself was decommissioned in 1962 and scrapped; a single rack of its tubes survives in the University of Illinois history collection. Hiller died in 1994; Isaacson remained at UIUC. Sources • Hiller, Lejaren; Isaacson, Leonard — Experimental Music: Composition with an Electronic Computer (McGraw-Hill, 1959). • University of Illinois Archives — Lejaren Hiller Papers. • Ames, Charles — "Automated Composition in Retrospect: 1956–1986" (Leonardo, 1987).
7. 1958 — Haroldo de Campos — Noigandres — Brazilian concrete poetry founding. Artists: Haroldo de Campos, Augusto de Campos, Décio Pignatari · Year: 1958 · Medium: Concrete poetry — text as visual and phonetic material · Substrate: Printed magazines Noigandres 1-4 (1952-1962); typographic experimentation treating text as visual material · Place: São Paulo, Brazil · Institution: Noigandres group (self-organized) The Noigandres group formed in São Paulo in 1952 around Haroldo de Campos, his brother Augusto de Campos, and Décio Pignatari. Its name was lifted from a line in Ezra Pound's Canto XX — a word no scholar had been able to gloss, and which Pound had adopted as a private sign for the untranslatable. The group's 1958 "Plano-piloto para poesia concreta" was a manifesto but also, more usefully, a specification. It laid down the operational rules of what they called concrete poetry: the poem as an object, not a line; typography as syntax; the visual arrangement of words on a page treated as part of the signal, not decoration of it. This matters to a computational archive because the Plano-piloto described, years before any of its authors saw a computer, the same decomposition of language that would be required to make a machine read or write poetry. A word was no longer a unit of meaning alone; it was a unit of meaning, shape, sound, and position. Each of those could be varied independently. Each could be operated on. "Poetic function" became, in their formulation, the layered result of a set of transformations applied to a substrate — a description indistinguishable from what a program does. When Augusto de Campos began, years later, to work with computer-generated layouts, animated typography, and holography, he was not adopting a new tool. He was finding machines that could finally execute the specification Noigandres had written out by hand. The group anticipated, in the specific sense that engineers mean the word, the formal preconditions of algorithmic writing. The 1958 manifesto is in this archive because concrete poetry is the most consequential pre-computational instance of work that treats language as a computational substrate. Its lineage runs straight into the sound poetry experiments of the 1960s, the text-machines of Alison Knowles, and every subsequent generation of programmers who have tried to make a computer write. Lore Haroldo de Campos translated the Iliad, Pound's Cantos, Mallarmé, Joyce, classical Chinese, and Hebrew scripture into Portuguese, and treated each translation as a formal experiment — a way of exposing the machinery of the source. He and Augusto were as much engineers of reading as they were poets of writing. The brothers collaborated for decades without ever writing a joint poem. The Plano-piloto was laid out in the magazine Noigandres 4 (1958) in grids of enumerated paragraphs, like a technical datasheet rather than a manifesto. The São Paulo art world called it a provocation. The engineers — and, eventually, the programmers — read it as a specification. Sources • de Campos, Haroldo; de Campos, Augusto; Pignatari, Décio — "Plano-piloto para poesia concreta" (Noigandres 4, 1958). • Bessa, Antonio Sergio; Cisneros, Odile (eds.) — Novas: Selected Writings of Haroldo de Campos (Northwestern, 2007). • Solt, Mary Ellen — Concrete Poetry: A World View (Indiana, 1968).
8. 1961 — John Whitney Sr. — Catalog. Artists: John Whitney Sr. · Year: 1961 · Medium: 16mm film · Substrate: Repurposed M-5 / M-7 anti-aircraft gun director (WWII surplus mechanical analog computer) · Place: Los Angeles, USA · Institution: Motion Graphics Inc. (his own studio) John Whitney Sr. bought a surplus World War II anti-aircraft gun director — an M-5 or M-7, accounts differ — some time in the late 1950s. The machine was designed to solve a hard real-time ballistics problem: given the speed, heading, and altitude of an approaching bomber, calculate where its flight path would intersect the muzzle of an anti-aircraft gun in three to five seconds' time, and lay the gun accordingly. It was a mechanical analog computer built for killing pilots. Whitney rewired it to drive cameras and artwork stages, and used it instead to make films. Catalog, released in 1961, is a demonstration reel — a short film that showed in sequence what his converted machine could do. Rotating patterns, concentric shapes expanding and collapsing in harmonic intervals, compositions that broke into subcompositions in real time. It is not a narrative, not a documentary, not even an experimental art film in the Brakhage or Deren tradition. It is a technical demonstration of a new kind of image-making, and as a technical demonstration it announced something that had not existed before: motion graphics as a designable medium. The audience mattered. Whitney showed Catalog to the commercial film industry, and the film got him title-design work and technical consultations. His son, John Whitney Jr., took what he learned from his father directly into the slit-scan sequences of 2001: A Space Odyssey in 1968. The entire vocabulary of Star Wars title crawls, television station idents, and network broadcast graphics of the 1970s and 1980s — every rotating logo, every expanding ring — descends through the Whitneys from a weapon of war that was supposed to shoot down bombers. That the first serious production system for computational motion graphics was built out of military surplus is not a curiosity of provenance. It is a structural feature of the whole field. The computer, at its origins, was a war machine. Everything computational art has done since has been done on tools whose funding histories run through the Pentagon or its counterparts. Whitney's gun director is the condition of the medium made visible. Lore Whitney bought a surplus WWII anti-aircraft gun director — a multi-axis mechanical analog computer originally designed to calculate the lead for shooting down bombers — and rewired it to move a camera and artwork stages in synchronized motion. 'Catalog' is literally what it sounds like: a demo reel of effects he had coded into the machine, offered as a showcase for film-title work. His son, John Whitney Jr., would later apply this lineage directly to the slit-scan 'Stargate' sequence in Kubrick's 2001: A Space Odyssey (1968). A weapon of war became, in the most literal sense, the seed machine of motion graphics. Sources • Youngblood, Gene — Expanded Cinema (Dutton, 1970) • Whitney, John — Digital Harmony (Byte Books, 1980) • Academy Film Archive