Versificators

There’s an almost heartwarming story at the end of an otherwise alarming Economist story about “music intelligence” — the use of computers to analyze music and determine what people will find appealing and program tunes accordingly (much like Orwell’s “versificator” in 1984 that spewed out songs to mollify the masses). Apparently, according to the article, Frederic Monneron, who lives in a small French village and publishes equestrian books, composed an album of ballads after “a setback in his love life” and submitted it to a song-analyzing company. With computer-driven analysis, the company determined that his songs would be big, and he went from an anoymous unknown to having 200,000 copies of his record pressed in no time, with the benefit of no agents, no network, no insiders to get him signed. It seems like a meritocratic triumph, bringing talent (if making music that has scientifically certified universal appeal counts as talent) to the fore while minimizing the slogging and socializing and “dues paying” that are usually required. It replaces the inefficient filter of one’s willingness to struggle that much, sacrifice enough to the process, with a filter that obviates all the human qualities of the musicians altogether. The Economist argues that Monerron proves that music intelligence won’t stifle human creativity but enhance its distribution. But it seems that Monneron may merely be a transitional artist, taking advantage of a new cultural-production system before it perfects itself and eliminates the need for human musicians altogether, just as Reason and such software has removed the need for instruments.

Labels are already using music intelligence to pick singles from established artists’ records, sequence an album’s songs into a “flow”, and generate tweaks that will enhance sales– the article names 50 Cent’s “Candy Shop” and James Blunt’s “You’re Beautiful” as two examples of computer-selected hits. And ringtones are being made with the help of machines. The computers apparently take sales data from past records and associate them with 30 different aspects of the sound itself (pitch, timbre, melody, etc.) and produce mathematical equations that can generate new songs with similar parameters that can be expected to have the same success. (This is probably similar to the technology that powers pandora.com, a website that tries to recommend music to you based on what bands you tell it that you already like. In a previous post I wondered if this destroys the serendipity that often makes music personally meaningful or if it just gives you more meaningful music. Meaning from culture product –especially pop songs –seems to depend so much on context: a song that played when she broke my heart, the song that played right before I started puking outside the bar, etc.) Essentially it codifies and perfects the instincts of bubblegum-music producers, whose transcendent faith in the formulas they’ve stumbled on have kept the charts populated for the past few decades, if not since the advent of recorded music.

Ultimately this mechanization of pop formula will widen the divide that already exists between artisanal music made for live performance (various forms of “roots” music) and for those nostalgic about human craftsmanship, and pop music made for mass consumption and quick disposability. The artisinal kind of music seems more and more like sepia-toned nostalgia (unless it’s me who is trying to play it in the privacy of a basement somewhere). Pop music seems much more pleasant to consume because it’s easy and immediate and requires no great understanding to appreciate. It’s quick to bring a sense that your culture understands you, no matter what emotional upset you may be feeling — in part because it allows you to shape such upset in a preconditioned package that comes equipped with soothing simple answers and platitudinous remedies. “True love will win in the end” “Time heals all” “Tonight is a great night to have fun.” “This time it will really be a good time.” Separating the creation of music from the celebrities recruited to sell it makes more gains in efficiency too, I’m sure. Obviously talent is no longer an alibi for celebrity. With better technology, no more mismatch between appearance and talent (the Jessica Simpson conundrum?)– the computer will supply the talent and the star just needs to live the preselected image (which perhaps computers will generate as well).

I guess this all appeals to me because it foregrounds the fact that pop is product and encourages us all to accept it as such instead of succumbing to marketing illusions that the songs have something unique and important to say, that they might help us discover something true about ourselves. When pop music is demystified, we can turn our self-producing energies elsewhere, perhaps use them to make things rather than consume them. That’s the theory anyway. Another way of saying it: the more ephemeral our pop pleasures are, the more energy we have left to bring to other things. But what are those other things?