Forward to next post: Robert Legault
Somewhere in the war between enthusiasm and cynicism, the content of Patrick’s notes on the O’Reilly Tools for Change for Publishing conference went undiscussed. And I’m sorry about that, because some of them really got my attention. They looked neat. I wanted to hear more.
Then, less than a week later, I was sitting in a restaurant in the Alps, nursing a bruised tendon in one knee while one of the smartest people I know dozed beside me. And suddenly two of those one-line comments tied together a mass of notions I’d collected from a dozen disparate sources. It was pure high country cascade, and I’d like to pick apart some of the ideas from it here on Making Light.
‘Content is king! Content is king!’ No. Nor is ‘context.’ Contact is king.
Since when is the Internet a monarchy?
Content is important, since the internet is basically an information delivery system. But we’re social animals, so we bring community into everything we do. Look at the stories of the desert fathers; these men were hermits, and still their lessons are about community. We’re the same today, even when we’re alone at our keyboards.
Really good communities tend to grow around valuable or interesting content. Google and StumbleUpon pull new readers in. If the content keeps moving and stays good, a certain proportion stay around and chat, get to know each other, hang out. The growing community at BoingBoing is a recent example.
Contrariwise, really good content is about community. Writers are motivated by readers; creation implies audience. And a good community can generate even better content than an individual, if conversation and dissent are used to refine the ideas under discussion.
The two are one, like hands joined together, like the end and the way.
Curating the conversation will be a whole new kind of editorial job.
Communities that generate content generally do it in conversation (not always; my old community didn’t). That conversation needs to be gardened and nurtured while it’s in progress — and that’s a whole ‘nother ball of fishy kettles — but it also needs care afterwards.
Online conversations can strain even Sturgeon’s Law. Some of them are astonishingly crap. And even — or especially — the good ones wander all over the place, cross-fertilize, merge and separate. Finding anything again is a matter of luck and memory. Finding it the first time is more about miracles and Google-fu.
That’s not going to be good enough in the future. More and more new ideas are first proposed online. Without a record of how they were developed, without the initial arguments against and reactions to them, they lose a significant part of their meaning. (For instance, only without context can Wikipedia seriously describe disemvoweling as “a technique by forum moderators to censor unwanted posting”.) Valuable historical information is being lost.
One could manually index conversations, of course, but that’s enormously time-consuming. Kelly McCullough’s excellent work indexing Making Light and Electrolite is the product of great effort as well as love. To index every conversation like that is way over the event horizon of practicality.
So what should be done?
If I may go all Charlie Stross for a moment (by which I mean buzzword-rich and science fictional), we need some way of automatically tagging and flagging conversations.
Tagging can indicate the content of a discussion. Word-frequency analysis might be useful here, perhaps in a moving average over multiple comments to pick up threaded discussions. There is a tension between using a controlled vocabulary and a free one; one is more searchable, the other more adaptable. Perhaps some form of evolving or semi-controlled vocabulary? Bayesian algorithms already figure out which emails are spam and which are ham; could they figure out which tags are applicable and which have been superseded?
But conversations should also be flagged for quality. A good conversation is idea-rich; a poor one reiterates old notions and beats them to death. Text mining and search systems should prioritise the former over the latter. And that’s yet another question: what are the markers of a good online conversation? I bet that’s harder to detect automatically than trolling, even, but what a treasure for searchability if we could!
At this point, my companion woke from his sleep and listened to my excited handwaving exposition. He was unconvinced.
What say you?
Nota bene: constructive criticism of goals, means and methods is keenly sought. Rants might be better left at the door, unless they are funny or in verse.