Blake Stacey, who is a good guy to have by your side in a firefight, has a wonderfully complicated post on this thing called science blogging. He’s mostly stating the obvious: it’s anarchic, it’s very hard to pull out, say, introductory material on a specific topic in science, there are problems of accountability, we don’t produce anything as coherently useful as a basic textbook, etc. Well, yeah. This is a general problem with solutions that bubble up from the ground rather than being defined from above — they do something very, very well, but it usually isn’t the something that a planner would design, and they often won’t easily do something else that you think they ought to do. Blake is entirely aware of this, obviously.

Nobody is acting as the central regulator of online science writing, though some would like to try. The interactions and evolutions we see are the result of the incentives at work, playing themselves out. If we want to change the way science blogging happens, or if we want our loose community to start generating something new, central decrees are no good: we have to make our desiderata the natural products of volunteer enthusiasm. Furthermore, science blogs are not a central authority for anybody else, so if we want to change their behavior, we have to find ways to put new “motivator units” in place.

He’s pointing to a real problem, and we’re aware that establishing a central authority is not going to solve anything, and I’d say it would even ruin everything. Incentives have to come from somewhere. My answer is…don’t worry, be happy, the solution will come from somewhere where you least expect it.

As an example, blogs themselves. I kept a web page for a long time, since the early 1990s, when all we did is write static html. When blogs started to emerge, I didn’t quite see the point. I could see exactly what they were: they were nothing but web-based front ends for personal databases. That’s all they are still. I couldn’t quite see the point — the data being stored was rather idiosyncratic, and personally, I couldn’t imagine myself writing enough stuff that it would warrant database tools to manage it. But then something odd happened: it turned out to be very useful to be able to compose something, and have it stored away in a manner that made it easy to access again. And then populating the database with useful stuff started to become an end in itself, the new motivator unit, and as the database grew, it became more useful, and that in turn made it more compelling to put more stuff in it, and so on.

Feed forward loops are powerful forces, people.

Then the other big force was Google. Google is the one significant tool we have for poking around in other people’s personal databases. That’s another powerful motivator, that we find ourselves able to plumb other people’s words and experiences fairly easily, and what do you know, other people are interesting, so we want to look more, and we want other people to find us interesting, so we stuff more and more goodies into our own databases. Feed forward, feed forward, feed forward.

However, and this is the limitation that Blake bemoans, the motivator is too general: all we’ve got is a criterion based on how many people find a particular database entry interesting, which is terribly vague. We want to know how many people find an entry informative, and even more, what kind of people (novices, experts, whatever) find the entry useful, and in what way. Google is terrible at this. I’ll be the first to admit that Pharyngula entries rank high in the Google indexes not because they are necessarily the best at explaining overall, but because they tap into subjects and attitudes that are popular. If popularity were synonymous with accurate or useful or expert, then America would have been created 6,000 years ago by a magical giant with a long white beard, and everyone in the world would be a porn star.

So right now we are waiting for the next piece of the toolbox to fall into place, adding new metrics that will feed forward into new capabilities. We can’t design them — design is a terrible paradigm for adding unexpected newness and potential (which any evolutionary biologist would tell you). What will happen is a surprising and unpredictable side-effect of something else on the web. It could be something like social networking software adding a new criterion, like whuffie, that focuses people’s energies on productive contributions — only it won’t be the social networking software as it is now, and it won’t be whuffie, and it will have multiple effects, some of which may not be desirable. We will be surprised. It will take time for it to take over and become useful. In the early stages, almost all of us will be scratching our heads and wondering why anyone would find that interesting (cf. Twitter), and many of the solutions that are promising early on will fail. We are waiting for something new to evolve, and the best way to promote that is to encourage diversity and look at everything sideways, not by pushing for a specific solution. You don’t get emergent properties by forcing a result, or they wouldn’t be emergent properties.

The only answer is to keep playing. Don’t worry about it. Expect and embrace serendipity.