Heading into Thanksgiving week, with my term at the School for Poetic Computation nearly done — just a final review and wrap-up week to follow — it seemed like a good time to reconsider what exactly Poetic Computation means, both to others and to me.
Like many rambly posts, this started off short and simple — where did text fit in with art? what exactly was the goal of this? — and then as I faced down my very real state of not-knowing, well, it all spiraled out of control and into discovery.
And so with all caveats made, where to start? With history, what brought me here? With classification? With semiotics, with the title itself, as our first week began? With distinction, to ask how is it different from entertainment, from advertising, from poetry-poetry? With a description of the maker, those attributes required by the code poet?
Oh, okay, who am I kidding. Let’s do them all.
Deciding to attend SFPC was both very slow and very quick. I’d been aware of the school from Zack Lieberman’s talk at the Eyeo Festival and when my initial plan to attend Hacker School fell though, it seemed like an interesting alternative. A place professing that writing code was creative writing was just what I needed.
Since I’d moved to San Francisco four years ago, heeding the seductive song of startup design roles, I’d been shocked by the absence of art and humanities thinking, of what I’d now call poetry, in the tech community, even as I fell in love with writing code itself and San Francisco’s steadfast proclamation that anyone can make anything (no matter how stupid).
So stubbornly, I looked at how to bring them together (I could make anything). Hectoring dudes I dated proved quickly to be a dead end. I would be no Joanie Applebook, spreading flexible thinking with flexible limbs, despite how well that aligns with other San Francisco myths.
Movements in data visualization were more promising. Unifying advances in web technology and interesting questions, data designers were telling compelling human stories, inventing new forms of narrative. That this appealed to someone who once took classes in things like Textuality and Narratology is perhaps unsurprising. Data vis took me to blogs like Infosthetics and the Creative Application Network, and eventually conferences like Eyeo. In each place I encountered work that worked as literature works — telling incomplete and elliptical “stories” in which we are able to find the most human parts of ourselves — but with the newest texture of our daily lives, with data. I was in love.
And as with such crushes, I stalked this work. Followed it, investigated its history, came up with strange theories about its meaning; I stayed up late imagining how I might offer it my own creations, so alike and so heartfelt, in hopes that we might, someday, be together.
In fact, I spent this summer giving talks about this kind of data art (including an Eyeo Ignite) and creating some of my own. In this way, winding up at SFPC was the culmination of this whole phase, a step into a community I had slowly discovered, hovered about, and finally intended to join.
Today when I pick up Holo, or otherwise encounter works data art, be they on the strict side of data or the generative end of the spectrum, rather than seeming the creations of unknowable beings with skills I can barely imagine, they become part of a conversation I understand and timidly enter into.
The historical narrative is the romantic narrative, in which code poetry both is the seductive language of a community and an artistic direction.
Standing away from the prosody of love, we have a more forthright question: if Poetic Computation is a style of work, what kind is it?
This slide, from Esolang’s Daniel Temkin is as good a start as any. (Though I should note, that since I have not seen the talk it was in and am using it merely as an artifact, I may be misrepresenting his meanings entirely.)
Code poetry is literature that intermixes notions of classical poetry and computer code. Unlike digital poetry, which prominently uses physical computers, code poems may or may not run through executable binaries.
But I would argue this differentiation is overdetermined and poetic computation encompasses all four levels of code-related art. These are then more logically distinguished by the position code holds in each: as material, tool, or object of investigation. I suspect this is what Temkin is after in his classification, where, to begin, generative art is related to output. In this case, code is merely the tool used to create works that are otherwise indistinguishable from traditional fine arts. That is, generative art is only as different from a painting as a sculpture is. Its questions are not about code.
At the next level, we have software art, which I find the hardest genre to corral, as it is essentially “art of and about software.” But drawing from the former level I might say this is where we place work whose questions are about software and computing culturally. This is the realm of artists like Aram Bartoll, Paolo Cirio, Jason Salavon, and nearly anyone who went through Eyebeam; it features works that straightforwardly or elliptically — nearly always conceptually — approach the membrane between tech and us.
Beneath this membrane, into tech, we have code art, where code transforms from fuel (practical or cultural) into an object in itself. This happens is when code is regarded as a system of signs no different from any other type of language. It is as expressive as generative art but the rules of talking to machines are foregrounded; the question is humanity and our new pets.
And then finally, what could be said to be the purpose of the esolang, where the focus is the construction of programming languages themselves? Some work as Fluxus projects, where the description is more important than the implementation but many, including the canonical Brainfuck are intended to be experienced. This of course limits the works’ audiences to those familiar with software development. While it would be easy to take this limitation to dismiss esolangs as toys for devs, I like to imagine them as performing one of the most important functions of poetry: examining the underpinnings of languages we take for granted by taking them to extremes. In this case, computation has become the universe in which the works take place, its principles questionable exactly as the principles of physics are in this universe: more to be understood than to be obliterated.
And so we move from computer as medium to cultural commodity to interface to playground, from incidental to required. Given the breadth of these ways and means, does it make sense to group such disparate types of work together as one thing, as poetic computation?
In search of answers, I return to SFPC itself. Since the school is what brought the term into my life, it only makes sense to revisit the Eyeo talk in which Zach Lieberman first introduced the school, and which, conveniently for this exploration, approaches things semiotically.
I am going to skip the definition of school, since it is not what I am most interested in, and skip straight to poetry, which Zach identifies with smallness, with hiddenness: poetry is the weird section at the back of the bookstore, filled with self-published chapbooks; it is the piece where two small boats are better able to charm and reflect us than a million-dollar blimp powered by brainwaves.
In contrast computation is Big Tech. It is the monster driven to be “newer, bigger, faster” and can only be restrained by the injection of poetry, which here becomes the inoculation for capitalism, computation’s twin face. And poetry is the vaccine because it is impractical and because it is human. The school for poetic computation, which advocates in favor of said work, is an un-vocational school, in Zach’s description: anti-job, anti-corporate, anti-money.
This formulation worries me a bit, not only because I enjoy eating, but because it seems to give short shrift to computation (which is not reflective of any actual discussion I’ve ever had with Zach). Nevertheless, the idea that computation is the footman to our drive to a technocratic dystopia does a great disservice to the work of non-poetic computators whose nearest bridge to us is a love of creation.
For to me, then, poetic computation is the work that complicates our relationships with code and tech by demanding they consume and reflect the human. It is at the level of code that we can search for computation with a soul, computation that works for us, an alternative view to the technocratic assumption that we are all finally little more than defective meat robots who must adopt logic to be worthy of our machines.
In this, computation itself is value neutral, just the way we have right now of talking to machines. And the poetic is the romantically human. Poetry uses aesthetics to express what straightforward declaration cannot; it is the opposite of what machines seem to demand: the explanation of everything. Poetic computation stands as a bulwark against a-poetic computation.
And a-poetic computation is the technocratic narrative. It is not inherent to the work of machines but to those who would seek power and for whom machines are a convenient means. The most poetic work then is not a certain type and does not demand a particular manner of interaction with computation (though I love the language focussed ones myself) — rather, it is the work that privileges the dignity of the human and the examination of itself.
When we look at poetic computation this way — as work with computers organized around a particular goal — it is clear how it is distinguished from being just a genre of new media art, just a method of literature, just a clever game. It is distinct to the extent that it work against technocracy and for people: as simple and as hard as that.
It is possible for the output to be co-opted and integrated into advertising. It is possible for work to sprout anywhere anyone is working with aesthetics.
Throughout these past few weeks, I keep coming back, as a touchstone, to two different ideas. The first is the core conceit of The Ship Who Sang series: in the future deformed humans are sealed up into ships, giving what might be objects life and personality. The best AI is augmented human.
The second is the talk given by Sara Hendren, wherein she discusses the notion that all tech is adaptive tech. If we begin at this point it is so easy to question what we do, what we adapt for, what we want to note, augment, expand.
Works that keep these same spirits, kindred, are works of poetic computation.
Willing to keep looking. Willing to fail. A little bit in love with machines and the spirits that live inside them.