Social Networks and Power

Here’s an interesting article about the social-network modeling of Duncan Watts, a researcher for Yahoo! who contests the popular marketing idea that influencing so-called Influentials is the key to making your idea or product go viral.

I first heard about him at a CSN conference last year; he had just published research suggesting that the success of any particular pop song or musician was essentially random, linked to early positive feedback from consumers.

The quality of a song accounted for something like half the variability in the success of a song in the artificial culture market he constructed. In short: if you suck, don’t expect to be successful, and if you’re amazing, you’ll do OK no matter what, but blessed are the mediocre, for they shall inherit the earth. But only if they’re lucky.

This sort of research matters for how we model literary history. What makes a book–let’s say, Invisible Man, which I’m writing about now–become a massive, runaway success. Was its success the result its outperforming the field of relevant competitors? Of Ellison’s being very well connected to elite literary and cultural circles? Of an essentially random processes of cultural selection, independent of its content (relative to the whole literary field)?

To answer these questions, we’d need a more precise definition of how networks of social power work, but history is unfortunately very messy, not anything like a controlled experiment. My gut tells me that we can’t abandon the idea of Influencers just yet, at least not in a slow-moving field like literary production (as opposed to, say, faster-paced music and clothing markets). If my research into the literary culture of the ’40s and ’50s has taught me anything, it’s that there are centers of cultural power, journals and magazines that don’t merely reflect the latest trends but rather influence the cultural conversation far more than you’d suspect by looking at data like relative circulation.