From "The Atlantic", August 12, 1999

Exquisite source

Heads turned in June when Linus Torvalds's Linux operating system was awarded first prize by the judges of an international art festival. How far, one wonders, can the open source model go?

by Harvey Blume

Where open source is concerned, no hyperbole seems too hyper. The distribution of Linux source code by Linus Torvalds, in 1991, has been compared to Martin Luther's translating the Bible into the vernacular. Larry Wall, the inventor of Perl, declares that open source programming is the expression in software of a fundamental Christian message: creation is not fixed in advance; free will is included and collaboration is encouraged. Even the Chinese Communist Party smiles on open source. China Youth Daily reports that open source programming has met with resistance from "software companies" trying to impose the norms of a "traditional market-economy age upon the new 'age of the information economy.'" Closer to home, but in no less political a vein, FEED's Steven Johnson compares Eric Raymond's open source manifesto, "The Cathedral and the Bazaar," to the Port Huron Statement, Tom Hayden's white paper for 1960s student radicalism.

The Kwakiutl Indians of the Pacific Northwest once flourished with their potlatch (or gift) economy, but other Americans have had little experience with the idea of prospering by giving wealth away instead of hoarding it. No wonder, then, that we're uncertain about how to label open source: whether as animal or vegetable, politics or theology. Open source pits the virtues of collaboration and participation against the habits of consolidation and control -- and, as far as the development of software infrastructure goes, it works. Still, how much cultural significance can be bundled into a software package?

This past June the jury of the Prix Ars Electronica added yet another dimension to open source by awarding Linux a Golden Nica for first prize in the ".net" category. (The 1999 awards will be presented on September 6 as part of the Ars Electronica Festival in Linz, Austria.) For twenty years, Ars Electronica has held festivals on the theme of "cultural transformation from the analog to the digital era." Its jurors are unimpressed by "recycling conventional art forms on the Net (e.g. Web galleries)" and unmoved by brilliant home pages. In Linux they found an alternative form, one that contributes to global networking even as it foments discussion about whether "code itself can be an artwork." At first blush this discussion doesn't seem all that promising. Why shouldn't code be art? From Chartres to the Brooklyn Bridge, feats of engineering have been appreciated for their aesthetic properties. Why should software engineering be any different?

But the Prix Ars Electronica went not only to the content of Linux (those efficient, bug-free lines of C) but also to the process of producing them -- in other words, to open source itself. As Eric Raymond observes in "The Cathedral and the Bazaar," "Linus's cleverest and most consequential hack was not the construction of the Linux kernel itself, but rather his invention of the Linux development model." The Golden Nica invites us to detach the code from the process for the moment and to ask, If open source can lead to computer code worthy of being called art, can it serve as a foundation for other kinds of art as well?

Of course, art has been made along open source lines in the past. In the 1920s, for example, the Surrealists explored a form they called the "exquisite corpse," in which a drawing or poem in progress was circulated to a number of artists for elaboration. (Examples are on display in the Surrealist exhibition at the Guggenheim Museum.) Participants in the project did not get to see the entire work (unlike in open source software development), only the section or line to which they would add -- but that shouldn't prevent us from viewing the exquisite corpse as a way of applying open source to poetry. It's our notion of open source that should be opened up and made flexible. The very market success of open source demands this; as it continues to prove its viability, open source enters into any number of partnerships with traditional forms, without necessarily abandoning its original inspiration. More generally, the open source phenomenon ought to be understood as an electronically networked and rapidly evolving expression of a long-standing collaborationist dream. As the Surrealists put it: "Poetry must be made by all and not by one."

An expanded view of open source sheds new light on one of twentieth-century art's signature techniques: quotation, or, in the digital context, sampling. Quotation is a kind of "open-sourcing" of artistic material. Picasso started by quoting freely from earlier styles (and wound up quoting freely from himself). Copyright was never at issue in Picasso's case -- only originality, its unregulated counterpart. Today, of course, the purview of quotation has expanded enormously. Sampling almost anyone and anything is just a simple cut-and-paste operation. From one point of view this is liberating. From another it is theft. And neatly separating these two aspects of the same process may be well beyond our powers while we are in the midst of it.

Consider the case of John Myatt, described last month in The New York Times Magazine as being part of an art-forgery scam so profound that it was said to have "altered art history." Myatt didn't grow ever more expert in copying one painter, as have forgers before him; he copied many twentieth-century artists, and (in his own opinion) not very adeptly: "There was a negligence to everything I did," he confessed. Your favorite Giacometti could easily be yet another Myatt, and certainly would be if it had any K-Y Jelly slopped on the canvas (the use of fast-drying K-Y being one of Myatt's techniques for increasing his output). Astonishingly, the K-Y almost never tipped off the experts, many of whom now despair of getting the canon right again, separating the real Giacomettis, Braques, and Chagalls from the jellied frauds. But other experts admit that Myatt was doing just what a modern artist should: quoting like crazy and making us think about what artistic originality amounts to in the first place. This is exactly the kind of thing the Ars Electronica jurors like to ponder. Perhaps next year, if he can get online, Myatt will be rewarded with a Golden Nica, making him a hero of the open source underground: the Linus Torvalds of the dark side.

Make no mistake -- there is such a thing as an open source underground, where the distinctions between literature and software, not to mention sharing and stealing, get really and truly fouled. This underground is personified by the legendary figure of Luther Blissett -- "Luther Blissett" being (according to a site called the "Luther Blissett Project: a mythopoetic on-line guide") "a multi-use name that can be adopted by anyone and is used every day and every night in the rest of Europe and the world." When Blissett hijacked a bus in Rome -- "with drums, confetti, drinks and ghetto blasters tuned in to Radio Blissett" -- the hijackers "bought only one ticket, because they all shared the same open identity, that of Luther Blissett." Blissett would just as soon crack Web sites as hijack buses. This past spring he paid a surprise visit to (an invitation-only portal to select art sites), downloaded all its files, and copied them to a free site. As he put it: "What is a computer if not something that benefits by the free flow of information?"

Blissett is also an open source novelist. Inspired by the papal encyclical Ut Unum Sint ("That all may be one"), he wrote A Survivor From the XVIth Century: Q (not yet in English), announcing that the process of composition requires "no boss, no mysterious scholar," only the contribution of many Blissetts, all of whom would be credited separately if their names were not identical. "Creative writing," Blissett declared, "is an utterly collective operation: concepts can't be anyone's property, the genius doesn't exist, there's just a Great Recombination." Clearly Blissett's idea of recombination owes more than a little to a talent for real and virtual safe cracking. His slogans might be: "Release early, release often -- or we'll do it for you!" and "Open source: by any means necessary!"

But we don't need Blissett to point out that open source practices of one sort or another are more common than might be supposed. There are signs, for example, that computer-game characters are beginning to offer themselves up to the great recombination. Phil Hood, an analyst at the Alliance for Converging Technologies, writes that computer games are "slowly moving to a model in which users and programmers can add characters and actions to existing proprietary games." In his view, game-makers will find it makes good business sense to collaborate with users "to extend, enhance and enlarge a brand." Hood explains that "By acknowledging fans and customers as 'co-owners,' companies open up new opportunities to tap the creative powers of their fan base." Those powers are, of course, the engine of open source. The New York Times recently reported on a game in which "the real action happens in edit mode, where you can customize and choreograph every facet of a fighting character's movement." And there have been reports of players fashioning high-powered characters for existing games, then auctioning them off on eBay.

Perhaps some comparable process will be coming to literature. Fictional characters have long been reused by writers -- think of how many authors have helped themselves to Sherlock Holmes, or how Shakespeare drew on a potpourri of characters as it suited him -- but characters may soon be snatched from author's galleys. Not plots, not whole books, but simply characters: literary action figures drafted into interaction with other such figures. It's relevant that the Nabokov estate has recently settled with Pia Pera, an Italian writer who has rewritten Lolita from Lolita's point of view, while the copyright to the original novel remains in effect. If this deal had followed open source protocol, the character of Lolita would be henceforth available for rewriting by any willing writer. One day, she might encounter Hamlet.

We think of open source as arising on the cutting edge of digital technology -- certainly Linux and, say, Apache, are inconceivable without an Internet. And yet the dream of a vast collaborative and communal enterprise is primal, whether expressed in the dictum that "Poetry must be made by all and not by one," or in recent allusions to an electronic noosphere, a region of ideas that encircles and engages us. Versions of the dream are as likely to turn up in the creation of comics as in the dissemination of scientific information.

Art Spiegelman, for example, based his last book -- The Narrative Corpse: A Chain-Story by 69 Artists! (1995) -- on the idea of the exquisite corpse, with each artist forwarding his contribution to the story line to the next cartoonist in the chain. And, this past June, NIH director Dr. Harold E. Varmus proposed to put all scientific research on what is effectively an open source footing. Under Varmus's plan, which is still being debated, researchers would skip mediation by scientific journals and upload their results directly to the Internet, to be freely examined by anyone. This system would strongly promote the collaborative side of science, upsetting researchers wed to the proprietary approach. According to The New York Times, however, many concede an open source revolution in research is "just a question of when and how."

The conjunction of old dream and new media gives open source redoubled force. Its popularity, and its presence in many fields of intellectual endeavor, make it tempting to look back on the production of cultural goods in the twentieth century as a halting march toward some grand collaborative consummation. Of course, the attractions of this sort of open source utopianism haven't been countered, as yet, with anything like open source realism. Hence the theorists of open source theology, open source politics, and open source business plans. The truth is, we don't know how far the open source model can go. The Golden Nica awarded to Linux reminds us it can go as far as art.