Joe Clark, in a discussion of the iPad, states:

“While people will tolerate a lot of things, what we want are beautiful things that work well. There aren’t many nonexperts who can accomplish that. Expertise needs schooling, maturation, taste, and quite a lot of attitude.

The foregoing explains why open source has nothing to teach literature or indeed any artistic creation, since talent doesn’t scale as you give more and more developers check-in access to the version-control system set up for your novel.”

Open source has a surprisingly large amount to teach literature and all forms of artistic creation – at least, if you look beyond the simplistic view of open source presented here, and if you truly examine the myth of the romantic artist, the nearly Randian ideal many embed artists and artisans within.

Everyone’s familiar with (or should be, at any rate) the notion of artist as superman – someone with a talent simply inaccessible to those around them, an individualist creator who must be rewarded for the act of creating in order to utilize their talents and whose creation comes out of their very soul.  Some people even believe it.  But these notions are questionable, at best.

First, the incentive argument: in a survey of various creative media types (I don’t have the citation handy, but it’s somewhere in the pile of paper on my desk – email if you want it), roughly 2% cited a monetary motive as part of the reason they create.  The rest?  “I had to.” “I couldn’t help it.” “I wanted to make something that people would enjoy.” “I wanted to be remembered for something after I was gone.” “I needed to.”  You’ll note a distinct absence of the word ‘money’, and a large component of need – artists create because they feel they have to, whether or not they have any particular talent at it.  Open source principles argue that collaborative filtering will adequately serve (and this argument is probably *more* useful in the context of art than software, as the metric for “good” is more accessible to anyone in artistic media) to deal with the difference between good and bad taste, between ugly and beautiful – especially since people  have different conceptions of beautiful.

But the bigger problem with Clark’s argument is in the notion of the romantic artist as the creator whose creation is individual and internal.  Art doesn’t work that way.  All art is manipulation of common cultural symbols in order to achieve an artistic purpose – without these common symbols, art cannot truly reach the intended viewer.  The symbols can be anything from cultural tropes – the general, like heartbreak, or the more specific, like Dark Is Evil, the in-between like dog and shotgun references in mocking country music – to common structures – the GDC chord progression in pop music, traditional folk melodies – to in-jokes and references – sampling, or intentional reuse of musical chords to evoke another work.  Art is quite open source in the sense that artists expose their source code to others so that others can learn from that code and reuse it for their own artistic purposes.  No, you wouldn’t generally use version-control systems to write a novel (although don’t entirely underestimate the notion – the Sanctuary series, amongst many others, are shared universes – but if they’d been starting it today, perhaps it would have been styled an “open source” universe, where the basic modules are available for all to play with but are checked in and out of the central repository by the continuity editors) but that doesn’t mean the larger themes, lessons, and importance of the open source notion isn’t useful in arts.  In many ways, something even stronger than open source was the original model for art – copyright is a relatively recent innovation, after all, and in the medieval era most European musicians made their money from patronage and performance (you know, supporting code, whether they wrote it or borrowed it from someone else who wrote it).  Most artforms have always incorporated earlier works by reference, something that open source makes intentional and possible in the world of software.

Is open source the right model for everything?  I don’t think so, personally, but that’s not important – the flaws that Clark points out are very prevalent in open source software because the wrong types of people are exerting control over things like interface.  When you take someone with talent in UI design and give them control over the UI, open source produces excellent software – the problems arise when you give people without that skillset (programmers, alas, typically are not interface experts) the ability to control the UI, and don’t have adequate selection mechanisms (because interface expertise isn’t something that’s possessed commonly).  An “open source” field of art lacks these problems, as outlined above – selection expertise is inherent in the program.  Expertise is still desirable, but it isn’t incompatible with the open source ideology, despite Clark’s statement.