Definition Not Found

The last refuge from #content might just be asemic writing

There’s a catharsis in writing without something to say. Your pen becomes a needle or a mosquito’s proboscis sucking from a well of lactic acid, the kind that settles just under your skin like cellulite. You do it when absently doodling in class or on the phone, making crosshatches, the coiled spirals of a rotary telephone wire, the onion sections of topographic contour lines. You might do it just because, or to see if you can, as if testing out a pen you’re not going to buy. Containing neither content nor value, these marks might be scanned as data but can never be parsed. It feels singularly seductive at a time when everything is made surveillable and where you don’t need to speak to be heard or write to be read.

Whether or not you know it, and perhaps especially if you don’t, what you’re doing is a kind of asemic mark making, where meaning looks possible — are the crosshatches hiding something; is the doodle a code? — but easy interpretation is denied. Per its etymology, “asemic writing,” a mode coined as a term by the visual poets Jim Leftwich and Tim Gaze in 1997, is writing without any specific semantic content. Although they looked to describe their own textual experiments, the term inspired a new generation of artists and writers and, buoyed by its circulation on the blogs and listservs of the late ’90s, soon grew to become a global movement. There is no small irony in assigning a name to a form predicated upon resisting meaning, like penning a press release for a protest without demands. Yet it feels very right that asemic writing should emerge from that particular Y2K moment of hurtling globalization, techno-pessimistic paranoia and neon-lit fishtanks; a time of semiotic overstimulation where signs swarmed like white blood cells and where, in the immortal words of Horse_ebooks, everything happens so much.

Asemic writing turns on apophenia, or the tendency to identify a signal where there is only noise

What we mean by asemic writing, though, dates back to two Tang Dynasty calligraphers, “crazy” Zhang Xu and “drunk” Huai Su. Revered for their cursive styles, their scripts are at once tender and wildly explosive, with all the expressive aggression of a ribbon worm shooting out its proboscis. The 1200 years since have produced numerous other proto-asemic examples, from the “interior gestures” of Henri Michaux and Roland Barthes’ “contre-écritures,” to the illegible writing of artists like Mark Tobey, Rachid Koraïchi, and Cy Twombly (a former Army cryptographer). We might consider legibility, a successful end-to-end transfer of discrete information, as the liminal boundary here. Asemic writing is to ‘legible’ writing what abstract art is to its more representational analogues. And just like abstraction in art — consider the evolution from Cubism and Futurism to Suprematism, for example — legibility exists along a continuum.


Take the ribbon worm knot on the left. Perhaps you understand it as a doughy coil or some kind of felt alphabet toy for a non-Latinate writing system. Perhaps you don’t read it as writing at all. Someone used to Latinate or Cyrillic scripts, however, would likely see in the worm on the right a ‘6’ or ‘b’ or ‘б,“ while an Arabic, Farsi or Urdu speaker might see a ط or a strong ‘T.’ Speakers of other languages may just see a pair of scribbles; what is legible to some might be entirely asemic for others. This is key: asemic writing turns on apophenia, or the terribly human tendency to perceive meaningful patterns in random data and to identify a signal where there is only noise. The intent of the creator thus becomes vital. For a work to be truly asemic, it should be illegible not just to readers and viewers but to its maker too, lest it be something more akin to a cipher. Conversely, some of history’s most impenetrable ciphers have later been revealed to be asemic, as in the case of Luigi Serafini’s 1981 illustrated encyclopedia The Codex Seraphinianus.

The absence of a specific semantic content does not mean that asemic writing is not inescapably semantic in form. We recognize something as asemic precisely because it bears the hallmarks of what we understand to be script. On the page or canvas or screen, the marks deploy a variety of lines, along with curves, strokes, serifs, and other fontlike or ideographic accoutrements. There is little to no attempt to depict depth, dimension or color, but there might be a sense of movement along vertical or horizontal axes. Sometimes, as in the work of artist Mirtha Dermisache, there is a lightly skeuomorphic attention to the modular columns and paragraph markers of the printed page. In this way, asemic writing might be better understood not as illegible but as ‘post-literate,’ to use the phrase of one of the contemporary movement’s most important hubs, the New Post-Literate. To encounter a piece of asemic writing is to engage in a kind of pattern recognition, a database query that heavily relies on what we visually interpret as writing-or-not. As with the replacement Unicode characters �, □, or ▯ that you might see when there is an error in rendering text or displaying foreign character sets, you cannot read it but you agree to understand it as language.

It’s worth emphasizing that asemic writing can extend beyond representations of typography, and graphic notation presents a particularly beautiful example. Here, you might see some familiar markers of musical scores: staves, notes, dynamics markings, sharps ♯, flats♭, naturals♮and other accidentals, key signatures, the slurs and accents of articulation, the angular slashes of ligatures, and so on. Sometimes there are directions, as in the score for Earle Brown’s seminal 1954 work, Four Systems, which instructs the performer that it “may be played in any sequence, either side up, at any tempo(i). The continuous lines from left to right define the outer limits of the keyboard. Thickness may indicate dynamics or clusters.” To consider the score’s geometric rectangles today is to feel a sense of frisson at its prefigurative qualities, at the way they resemble the horizontal bars of a midi editing software.

Asemic writing might be better understood not as illegible but as ‘post-literate’

Musicologist and historian Richard Taruskin, in his weighty title The Oxford History of Western Music, says that electronic technologies have resulted in us entering a post-literate sonic era in which standard notation and conventional musical literacy have lost their primacy as a means of musical preservation. It’s a phenomenon that graphic scores would seem to have presaged. Yet what graphic scores best illustrate are qualities beyond musical notation’s communicative potential, as with intensifications of emotion in Marco Fusinato’s 2007–13 series Mass Black Implosion. On archival facsimiles of avant-garde graphic scores, the artist rules a line from each note to a central point “as a proposition for a new composition, in which every note is played at once, as a moment of consolidation and singular impact.” The result is a series of arresting sinkholes that suggest obsessive SEO link building for the end of the world.

How best might we represent sound? The delightful Wikipedia page Cross-linguistic onomatopoeias points to the failure of textual language to do so. I am particularly fond of its lists of common animal sounds in different languages: A horse trotting is “clip clop” in English, “deregin-deregin” in Arabic, “pocotó pocotó” in Portuguese, “tsok-tsok” in Russian, and “gadagung gadagung” in Danish. A pig grunting, meanwhile, is “ghnot ghnot” in Bengali, “kkul kkul” in Korean, “röh röh röh” in Estonian, and “oinc oinc” in Catalan. In an interview at Scriptjr, Finnish artist Satu Kaikkonen gestures to asemics as a possible democratizing, universalizing solution, saying “asemic art can serve as a sort of common language — albeit an abstract, post-literate one — that we can use to understand one another regardless of background or nationality. For all its limping-functionality, semantic language all too often divides and asymmetrically empowers while asemic texts can’t help but put people of all literacy-levels and identities on equal footing.” There is, on the other hand, a pleasing harmony between terms for TV static, which in Swedish, Danish, and Indonesian all translate to “war of the ants,” while Hungarian uses “ant soccer” and Romanian “fleas,” even if it portends a flattening of meanings engendered by technology gone global.

To the analogous question of how we might represent silence, John Cage’s 1952 opus rasa 4’33 provides the obvious answer, yet far more exciting are the works of sound artist Christine Sun Kim, who splices American Sign Language and musical notation in visual scores that deftly reconfigure concepts of both duration and futurity. In her drawings, time becomes spatialized, with waterlogged lines that loop and multiply like a series of dance notations rent asunder, all post-bombing filamentous rebars and Russian smileys. Sun Kim is herself deaf, and prefers to instead consider silence in terms of quietness, “because I still do not quite get what ‘silence’ means, especially since I grew up instilling your perception of it, not mine.” In particular, her work How to Measure Quietness (2014) suggests that we might consider quietness as degrees of interiority with a series of pianissimos that run the gamut from sleep (mp) and deaf breath (p) through to heartburn (pppppp), anxiety (ppppppp) and silent treatment (pppppppp). The fortissimos of How to Measure Pauses (2014), meanwhile, offer that silence can be very, very loud.

Musical notation differs from other kinds of writing in that it is both a record and a set of reproducible instructions. Even as any standard score allows the performer a certain amount of liberty — a loosening of time in a rubato section, or a different color picked from a field of tone and timbre — it requires a fairly strict adherence to the piece as written. Another Sun Kim work, Eighth Note’s Worst Nightmare (2014), nods to graphic notation’s terrifying freedom with its jottings “no stem/no flag/no staff.” Failing to sound a whole note, or to sound a different one in its place, is as unthinkable as skipping or replacing the words in a recitation of a poem, the form of writing in which, at its best, the words and shapes are most fixed and the associations, even the meanings, most free. But if you don’t know and can’t hear the sounds that are signified as notes, a score on a page can become something close to asemic.

Perhaps those hi-def photographs of the ribbon worm aren’t the best depictions of asemic or post-literate writing, even as their proto-textual forms invoke the delicious possibility that insect life might one day evolve to camouflage itself against the regime of signs that now surrounds us — stick mantises that blend in with data center cables, maybe, or peppered moths with newsprint-patterned wings. Still, there’s something in their soft, corrosive brutality that speaks to the loss inherent in writing. There is always something we take away by standardization — that move from fluid line to letter or character; tag yourself I’m the one who lost my mother tongue — and refuse to give back. Think teaching children cursive handwriting, or the linguistic devastation wrought by Canada’s residential school system, whose program of forced assimilation resulted in the decimation of a number of First Nations languages. Or consider the plight of the nastaliq script of Urdu and Farsi, along with many other languages in the Central Asian stretch between Iran and China’s Xinjiang province. With a writing system that moves both diagonally and horizontally it is notoriously difficult to code and is increasingly replaced online with the Naskh script of Arabic, or forced to circulate as .png or as .jpeg. A viral project Tag Clouds by French artist Mathieu Tremblin illustrates this especially neatly. Drawing analogies between graffiti tags and online depictions of keyword metadata, he paints over existing street art with a machine-readable translation that “makes shit graffiti legible,” or more generously, privileges easily extractable semantic data over form and expression.

There’s something in @reverseocr’s yearning to be understood that makes me think its labor is a kind of unrequited love

Within the sphere of green anarchist thought there is a current that bills itself as primitivist, with all the condescending fetishism that “primitive” invokes. Avowedly anti-technology, the anti-civilizationist critique of capitalism extends beyond the environmental degradation and forms of domination of contemporary production to rail against the concept of civilization itself. The sphere of alienation is extended beyond labor; as theorist John Zerzan lays out in Running on Emptiness, it is the regime of symbolic thought that is believed to most deeply distance us from our authentic selves, which are arbitrarily defined as the way we once existed as hunter-gatherers. Art, music, mathematics, literature, speech: any mode of representation is highly suspect. It’s the paleo diet, but for culture. Zerzan’s vision for the “future primitive” would have us living in a silent, pre-pastoralist utopia where we exist wordlessly amongst the trees — beyond art and agriculture and beyond semiotics, or perhaps more aptly, before and unsullied by it. While Zerzan’s concepts seem attractive as a thought exercise, they are unconvincingly and rather petulantly argued. Who would want to do away with the back catalogue of some of the only good things to come out of the morass of humanity as we know it? Perversely, a reading of these texts makes me wonder about the possibility of an asemic writing made not by humans, but by bots and other algorithms.

In 2011, So Kanno and Takahiro Yamaguchi created the Senseless Drawing Bot, a kinetic drawing machine that is Jean Tinguely-meets-Mars rover. It pairs a motorized skateboard with an arduino, and a long-short double pendulum that induces an element of chaos, to spray graffiti on the wall. There’s a lot of empty swinging and swaggering, a louche calisthenics. It makes a mark only when its randomized wobbles pass a certain pre-coded threshold, when it’s sure all eyes are on it, and then its gestures are fast, flashy, and nonchalant, as if drawn from immense, tumescent muscle memory. It’s all big words and it’s trying hard to flex; if ever a bot has seemed like a blustery fuckboy, this is it. The outcome is surprisingly great, a dense accumulation of multicolored freneticism, neat on the bottom and looping wildly on top like an overgrown hedge. Unlike the aforementioned Tag Clouds, it points to a machinic tagging that does not mandoline work into strict taxonomies, is unreadable by human viewers, and does not — yet — appear to be machine readable, either, as well as the delightful paradox of generative bots which are programmed by people, yet also enjoy their own agency.

In the realm of graphic notation, Emma Winston’s @GraphicScoreBot tweets out an image resembling a graphic score every hour. Each tweet features an outlined white rectangle, usually with stave lines, and often with a bass or treble clef and dynamic markings, so it’s clear we are to read this as music. Except, instead of conventional note forms, its markup includes an array of colorful geometric shapes, squiggles, and dashes. Circles of varying sizes and transparencies especially make the images feel like musical infographics (to me, they seem to suggest duration; others might see in them chords or orchestra stabs). There are semantic ruptures: the bot will, at random, tweet out cards from Brian Eno and Peter Schmidt’s Oblique Strategies, entreaties like “Trust in the you of now,” “A very small object. Its center,” and “Slow preparation, fast execution.” Less bombastic are the double-spaced “B E G I N” and “E N D” that pepper the scores, which Winston suggests can be taken as start and end points or altogether ignored. Though the scores are generally sparse, occasional plaintive adverbs and phrases like “sadly,” “casually,” and “as if tired” make suggestions as to mood. Cameos by Italian terms like con moto (with movement), andante (at a walking pace), and quasi niente (fade away to nothing) make the scores feel somehow more official. If the “post-literate” leads us to interrogate what we consider to be writing, this bot’s relative adherence to notational convention, more Fauvism than De Stijl, does the same for the musical score.

Also on Twitter, Darius Kazemi’s @reverseocr tweets out asemicisms more akin to those absentminded doodles, each cryptic scrawl accompanied by a random word, like “subtlety,” four times a day. It’s a study in impenetrable handwriting, only here the writer is not a shrink with a prescription pad but a bot. Without that accompanying word, the marks, while elegantly spare, are unrecognizable as anything but marks. So far, so asemic. Yet the way the bot works is by selecting a word and then trying — badly, endearingly — to draw it out. It keeps drawing, and failing, until an OCR or Optical Character Recognition program (the question of literacy is transposed to the algorithm, here) identifies a character. If that character matches the first letter of the word, “s” in the case of “subtlety,” that character gets drawn and the bot turns its attentions to the second character, “u.” If not, it perseveres until it gets a match, and eventually it manages, through trial and a lot of error, to draw out the whole word; we only see these successes. Of course all of these computational processes happen at lightning speed, but in a 2014 adaptation of the work for a show at Boston’s now-shuttered Find and Form Space Kazemi slows the algorithm down to a human timescale and makes visible the otherwise hidden work performed by the bot. The word here is, appropriately, “labor.” Yet there’s something in @reverseocr’s yearning to be understood — to be read, to be recognized by another — that makes me think it’s a kind of unrequited love. There is a 1973 interview with James Baldwin in the Black Scholar in which he says, in response to a question about the role of political themes in his writing,

The people produce the artist, and it’s true. The artist also produces the people. And that’s a very violent and terrifying act of love. The role of the artist and the role of the lover. If I love you, I have to make you conscious of the things you don’t see. Insofar as that is true, in that effort, I became conscious of the things that I don’t see. And I will not see without you, and vice versa, you will not see without me. No one wants to see more than he sees. You have to be driven to see what you see. The only way you can get through it is to accept that two-way street which I call love. You can call it a poem, you can call it whatever you like. That’s how people grow up. An artist is here not to give you answers but to ask you questions.

Kazemi’s bot expands the field of how we might understand asemic writing. Illegible though its drawings may be to our eyes, it is without doubt trying very, very hard to communicate meaning. Humans are not its intended audience; rather, its visual language, like barcodes or the computer vision markup of Amazon warehouses, is entirely for bots, machines, scripts, and other denizens of the algorithmic world. It’s a robot laughing alone with salad, and its inner life, its own well of lactic acid that it draws from to express itself, is off-limits to us. We, however, are on view to them, from the moment we press our thumbprints into our iPhones in the morning to the moment we touch-type a 2 a.m. text message whose characters are so drunkenly scrambled as to form complete non-words, which an algorithm gently corrects to other words we did or did not mean, so long as they’re legible. Perhaps this is an imposition on our freedoms; perhaps this is that two-way street between us and the algorithms, learning from each other; perhaps this is love.

Rahel Aima is a writer based between Brooklyn and Dubai. She runs NIGHTLIFE, a newsletter about staying in, and is currently working on a book about oil, water, and digital culture in the Arabian Gulf.