For your hand, it’s harder to draw a straight line than a crooked one. For your mind (and your computer), however, it’s the opposite — harder to draw crookedly than straight.

Such were the notes I jotted down a few weeks ago after participating in a desultory panel discussion convened for the Immaterial Materials show at Pratt. A curator from the Whitney, Chrissie Isles, made the claim that art has now expanded the notion of drawing to include all kinds of other things (”practices”). Some instances of performance art, she said, should be considered drawing — the notion being that the movement of bodies can be said to inscribe invisible lines across the time and space of the art event.

I found myself demurring, even though this inclusiveness was generously intended — among other things, it brought several of our digital artworks into the fold. Since these works take our old notion of hand-drawn space as a point of departure and actively engage not only the idea but also the responsive line of drawing, they can indeed seem to be a new form of drawing. In fact, nothing could please at least part of me more.

But to expand a definition is often to blur it, smudging crucial distinctions and differences. At such a point, it’s usually a good move to return to particulars to see what is really going on behind the generalization. Within our body of work, Loops is the purest (and most radical) example of a work that verges on drawing, so we can begin by comparing it to a canonical work of traditional drawing — say (to choose the very best) a Leonardo notebook page.

So what are the attributes of a Leonardo page? First, it’s comes about through the solitary and private act of the artist. Second, it’s an active attempt to arrive at a better understanding of its subject — with the effort of that attempt not hidden away, but rather with the mind’s rapid movements and decisions left bare (for example, an initial sketch is frequently redrawn from different perspectives or simply abandoned mid-picture once its conclusion has been glimpsed). Third, the notebook page is unique, with every mark bearing the unrepeatable character of Leonardo’s hand on that exact occasion. And, finally, while the original act of drawing unfolded in time, the resulting page is now fixed and unchanging. Time does hover over it, in a fashion, for the page gives us a sense of the original motion of Leonardo’s hand and mind, but this is at best a vague and tenuous awareness (it is certainly the case that the motions of our eyes in their saccades over the drawing don’t track those of his hand in sketching it).

Now to consider Loops in the same order of attributes:

• solitary and private act? — It’s neither of these, for it’s a layered series of divergent acts by a number of different actors: the underlying movement is that of Merce Cunningham performing a dance he has never set on any other dancer; the parameters for the visual patterns and decisions are made in collaboration by all three artists in our group; and the ultimate imagery is generated by a virtual intelligence operating autonomously within the given parameters.
• mind’s active attempt at understanding? — Yes, but again not one mind, or even one kind of mind, but several, in the order listed above. The most unusual aspect of this is the last “mind” in the sequence, if it can be considered such; in any case, the effort of the computer in its work of interpretation is made fully visible.
• unique? — No and yes: no, in that the underlying code may be duplicated exactly at any time (and not just in theory but in practice, since we’ve released it as open source); yes, in that no playback of the code is ever the same, so that what you glimpse on the screen now you will never see again.
• fixed and unchanging? — No, as indicated in the point above. Though one further thing to note is that while the present code for the work is fixed in its current version, it is open to revision and complete reinvention (a liberty we ourselves have taken three times and which anyone else is allowed to do freely).

So is Loops a form of drawing? A complicated rather than a simple question, which you may decide on your own.

(I notice now that I never expanded upon the thought noted down at the outset of this post, but if you follow its direction it will lead to the reason why one would even bother simulating a hand-drawn line in code).

A few days ago we announced the release of Field and Loops under open source licenses (both the choreography and performance by Merce Cunningham, and our new digital artwork of the same name). Was this altruism on our part? Not completely, there’s a plan: we’re trying to change the eco-system of digital art and performance in our favor.

For a long time now we’ve been frustrated with how digital art is taught and thought about, and I’ve come to believe that part of what’s gone astray is the very thing that we are trying to address in this “preservation project”: that is preservation itself.

Scholarship and preservation

Three things are almost constant in our field. Firstly, if you read any criticism about a digital artwork, it’s been written by someone who has been unable to study the object they are writing about in depth, hold it in their hands, take it apart, see how it works. Often they are reduced to guessing, to making things up (while defending making things up as a valid critical approach). Secondly, if you read anything interesting about an artwork, it’s almost impossible to track it down and see it for yourself. Photographs with captions in books are our canon, in a way that’s almost unbearably ironic for a field that often touts its ground-breaking multimedia nature on the very same page. These books are vital, but they are the botany, not the physics of our understanding of this art. And finally, should you locate the artists responsible, there’s a good chance that the artwork will have stopped working or be prohibitively hard to put back together again for installation. At best, all they will have is a wider range of photographs and captions for you to look at.

This isn’t the situation with painting or music; but it is a condition that digital art shares with dance. And these three things severely limit the quality of teaching and criticism in general and the size and complexity of the objects that students and critics can study in particular. Ultimately they damage the art itself. Our large-scale, complex, and fragile artworks don’t fit well within this environment. Something has to change.

We’re not the first to care about the preservation of contemporary art (or dance). On the contrary, there has been considerable discussion, and funding for discussion, over the last decade. It’s amounted to, as far as we can see, very little that’s concrete and even less that’s interesting. It seems driven by either an academic desire to talk about a fascinating and fashionably interdisciplinary subject or an economic desire to transform digital artworks into something that can be bought and sold like a painting. Frustrated by this, we approached the Mellon Foundation, for funding for Loops, with a strong desire to indicate how we thought it should be done by actually doing it.

Digital artworks cannot be bought and sold like a painting — and all the talk and free conference food in the world will not change this. Our new strategy of open-sourcing everything we can find an excuse and the funding to open source, leads then to a different place, frankly, a more moral place. One where the theory (scholarship) and practice (pedagogy) of our field stands to be transformed by the possibility of “close reading” and deep understanding of the artworks themselves; where digital art, and dance scholarship, are no longer starved of actual, useful, stable, share-able examples.

Open-source software and open-source Art

That was the practical argument for “giving away” an artwork as open source, but are there deeper connections between the morals of software-making and art-making?

To readers arriving from the software-, rather than the art-, world, the previous section will sound very familiar. This is in fact nothing more than the philosophical goals, and the economic realities, of the open-source movement. The philosophical goals are often obscured today (given the success of the model, some have come to suggest that open source strategies are justifiable more simply, and less politically, in terms of the quality and cost of the software it produces). But, to my eye, the original point of this was both clear and extremely radical: that the attendant, moral price of giving somebody a piece of your code was that you also had to give them the ability to understand it, change it, and share it.

This is a different reading of the philosophical tenants of free software than most, but let’s rewrite this argument for the art-, rather than the software-, world: the price of having strangers come and look at your art in a gallery (or in a theatre, or especially on a street, in public), the price of putting your art in their head, is that you have to give them also the means of understanding, transforming and sharing it as well. Oliver Sacks’ most recent book, Musicophilia, hints at the psychological stakes of this exchange with a stark image, calling music that sticks in your head to the point of irritation an Earworm. Perhaps the cost of being in the Earworm distribution business should be that you have to allow your worms to be dissected, understood and shared. Perhaps, the self-confident Earworm maker might hope to get back more from a now flourishing community of Earwormologists than they lose from their more limited control over their ear-market.

My overview of “free software” here is backwards from the usual presentation of the open software’s core ideas: it’s typically given in terms of preserving the freedom of the recipient (of the software) to “tinker” with what they receive. I believe this (perhaps deliberately) downplays the heavy moral dimension of the life of the software creator in favor of underlining the wonderful freedoms to-be-enjoyed by the software consumer. Fun as “tinkering” is, it just doesn’t capture what makes software special. Software is simultaneously just like a screwdriver (in that if you sell me a screwdriver, you shouldn’t try to tell me what I can and can’t screw with it) but it’s also nothing like a screwdriver (in that your screwdriver doesn’t become a transformative extension of my brain with the moral responsibilities that entails). Software is in a special class of things, a class shared only by a few other things, including, perhaps, art.

What is “distribution” ?

But both perspectives can exploit the same strategy, and the third revision of the open source movement’s main tactic, its “core text”, the GNU Public License, was formulated quite recently. This is the very legalistically magic document that aims to maintain these freedoms (and codifies these responsibilities). Simply put: if you distribute software under this license, you must distribute the source code to it, in a way that allows understanding, modification and subsequent redistribution of those modifications under this very license in turn. Version 3 of this document updates the highly successful version 2 to defend this vision against new attacks, imagined and real, conducted by forces in favor of closed-software. (For example, some consumer product manufacturers, most notably Tivo, are dedicated to making hardware that, while having freely modifiable software, from other people, on it, won’t actually run any software that’s been modified, even by those people).

One attack (and defense) that was presumably deemed to be too radical, even for GNU, is being posed by web-based software. Should, say, Google turn out to be based on modifications of open software, should they be bound to release their code to me, an avid google-r, or is there something special about using a web browser that prevents this interaction counting as “software distribution?” If it works at all, my perspective, and my moral argument, still work despite the prophylactic Firefox and cable modem: Google is in the process of transforming not just society and scholarship, but individuals, and yet we are unable to study how it works, talk about it in any deep way, or change our own versions of it. Regardless of its intentions, can this ever really be morally acceptable?

Seeing this issue arising, but perhaps unwilling to demand that everybody play hardball with Google, the Free Software Foundation (the force behind the GNU Public License) worked on an optional addition to its license that would cover this: the confusingly entitled Affero GPL. Works distributed under this license explicitly count web-services as distribution and providing such services automatically triggers the legal responsibility to distribute the underlying code. If Google incorporate this code into a web-service, the whole service becomes “tainted” and requires the distribution of its source-code.

This will, of course, never happen, but the existence of this license shows that the very nature of what “distribution” is is transforming, and up for debate.

Distributing Loops?

Which brings me to an utterly hypothetical, but intriguing point back in the (digital) art world: what if performance (or installation) of an artwork counted as distribution? What if works made with tools under GPL-like licenses triggered the code distribution clauses of the GPL as soon as the doors to the gallery opened? What if the audience had to be able demand a CD of the source code for the piece on the way out of the auditorium? Critical thought about digital art would be transformed, new forms of scholarship would appear, the techniques of digital art, poorly taught right now, would practically teach themselves in self-assembling online forums.

Fantasy, perhaps. One thing is for sure, we are sufficiently tickled by this idea that Loops, when it is installed “for real” in a gallery (rather than a press conference), will be accompanied by a stack of DVD-Rs. Like the Loops preservation project itself: a small gesture, but one that we hope becomes exemplary.

And with that, back to cleaning up Loop’s source code, so that I can post it all online.

In Loops there are 25 points for each of Merce Cunningham’s hands. Each point is a kind of tiny creature, highly susceptible to three kinds of impulses:

1. peer pressure — the desire to act just like neighboring points;
2. excitation — the sensitivity to exciting news from neighbors near and far; and
3. boredom — the mounting impulse not to be stuck in the same pattern forever.

Boredom, for example, is calculated in a 2d behavioral space, as pictured in the sample diagrams above. As a cluster of points persists with the same behavior, a circle of boredom starts expanding around them, which increases their sensitivity to excitation from without.

As we remake Loops, we’re explicitly annotating some of these previously invisible forces within the artwork. Thus, the Circles of Boredom can appear explicitly from time to time, looking like this:

As an annotation title, Circles of Boredom sounds like a translation from an obscure Jules Laforgue poem. The other titles are more straightforward: Corners Found; Extrema; Fastest; and Center of Speed –> Proximal Points.

Revisiting Loops reminds me of the sea-change that Shelley and I experienced when we started working with Marc in 2001. What had once taken the two of us hours, days and weeks to render, Marc’s code could now blast to the screen in realtime. Any decisions we made were reflected almost immediately on the screen we were all three gazing into.

An immediate consequence of this is that we could try any number of variations and outright changes, pursuing many more possibilities than we could ever have even begun to contemplate before. And since the probabilistic strategies of Marc’s software gave back imagery that we could never quite predict (or, for that matter, quite repeat), the process of creation became a matter more of searching and finding than of designing or constructing.

We would glimpse oddly compelling configurations that we’d try to refine by encircling parametrically. What was especially strange was recognizing in the flux of pixels weirdly familiar styles and forms that grew and mutated one into another — an abstract skeleton, for example, with lines like a Giacometti drawing once became a fleeting series of forms that we named in honor of Duchamp’s Large Glass but were never quite able to corral.

The Blind Watchmaker (1996)

Richard Dawkins experienced something similar in the simple Biomorph program he once created to generate artificial 2d creatures. He put his finger on the phenomenon:

Technically, all we are doing, when we play the computer biomorph game, is finding animals that are waiting to be found.
What it feels like is a process of artistic creation. … Effective searching procedures become, when the search-space is sufficiently large, indistinguishable from true creativity.

As we remake Loops from scratch (see the Loops open source project), we aim among other things to document the process completely.

Thus, for instance, automatic stills are generated every twenty seconds and constitute a visual archive of the piece’s development (see Loops stills for a sampling of these).

Similarly, our authoring system, Field, time-stamps all changes to the scripting code and maintains a versioning archive that preserves each and every build of the piece as we put it together.

We’re equally intent on turning all the key diagrams we scribble on folio paper into readable pdfs so that anyone interested can examine the conceptual process we’ve pursued.

One such diagram shows all the key elements of Loops in its reconstructed form. A detail of that diagram is reproduced above; the complete version is available as this pdf.