Friday, January 14, 2005

Centers and the web



I just had an interesting dream, about an extraordinary community in the woods I just visited, known as Breitenbush. There was a meeting space, with quite a lot of steam, very atmospheric in the way Breitenbush is with its countless hot springs. Somehow, everything we said in this space was being recorded, as a community memory. Over time, slowly, community administrators would go through and transcribe the meetings. Also, over time, everyone would create links between the transcriptions. The result was searchable by watching the recordings, and stopping at any point to fast forward or backwards, and the recordings would get louder & warmer when the topic was related, so you could stop and watch the unfolding of the topic over the years. It was a dream, so actually when you saw these recordings, you relived the meetings from that time. Somebody (I think it was the voice of my conciousness) said "this is terribly disorganized! We need to create a hierarchy of topics, and put all these memories into specific labeled boxes." And the community didn't allow this, but said anyone could create a category, an abstraction, and link it to as many points in the network of transcripts that he liked.

A very Gestalt dream. But it makes sense that people building an egalitarian sustainable community in a forest of hot springs, near a dormant volcano, would see such a picture naturally.

So we come to the web itself, and the fact that the Open Directory and Yahoo Directory, a.k.a. 'ontologies', are kind of like the office-librarian-voice, and searches weighted by connections are a lot closer to the community-voice. Google's success is based on the weight of clusters of connections. These clusters are kind of like fields, but specifically like centers.

In The Nature of Order, Christopher Alexander advocates looking at the world as intertwined, overlapping centers. I can describe centers best by example: I'm typing on a keyboard. That is a center. The keys are centers. The keyboard is a center. The spaces between the keys are centers. These spaces plus the surrounding keys are centers. The clusters of keys are centers. The keys in various states are yet different centers. My finger hitting a key is a center. My hands hovering over the keyboard, the space between keyboard & hands, along with the impact of a specific key in time, also, together, makes a center. The idea I have, and the impact it has on the keyboard and ultimately on the web, is a center.

Centers are a kind of boundary-less field in space, overlapping each other, so that these structures all exist at once, with some changing faster than others.

But they also have weight. A very coherent field of centers has more life, or weight, depending on the strength of the connections among all of these intersections, and how the whole of them creates something far more than any of them would in isolation.

Those of us who've built computer networks for years (I first used email in 1974), knew that weighted connections were important, and when the web came along, with its potent version of the hypertext vision, it was quite natural to think in terms of weighing clusters of connections. In 1998, a friend and I sat at a computer and starting building some software to search based on numbers & cycles of links ... we were 150 feet from Larry Page's apartment, and we didn't know that he'd started to do this a year earlier. But that just goes to show that searching for connections, and the strength and importance of clusters of connections, is a perfectly natural thing.

The force of wholeness is one of these things you find in the strangest places. Right about the same time, 1999, I was in Haight Ashbury with some friends, at a pub called "Hobsons' Victorian Punchbowl House". Everyone was ordering a fashionable drink, but, as far as I could tell, no one in this crowded place had ordered the victorian punchbowl. We did, and we put it in the middle of the table, but there were six cups around the edge of this bowl, and only three of us, all sitting on one side of a table with five chairs. It was unbalanced. It wasn't whole. That is the only possible explanation I have for why two girls, from another table, suddenly left their dates and sat down on the other side of the punch bowl, and without asking, picked up a cup and took a drink and started chatting with us. They were actually kind of surprised that they'd done it, and we talked about how they'd done it completely unconciously. Despite all the funny possible explanations, I believe, in truth, they had been drawn to complete the whole.

Looking at my Breitenbush dream, it seems like there are a few things missing on the web today. One is communiy memory, which of course was the name of the first community computer network project, started by Efrem Lipkin and others in Berkeley in the 1970's. The idea was empowerment -- it was very easy to manipulate people if they couldn't follow the sequence of events. I'd also add that watching the creation of a project over time, something Christopher Alexander would call its unfolding, is the fastest way to learn why the project is the way it is. Adele Goldberg is working on applying this fact to educational software. A quality of no history on the World Wide Web was apparent very early on, and archive.org was built to help correct this situation. Still, we have a long way to go.

But another thing that's missing is making more connections in the production of content for the web. I keep a lot of blogs, but very often a blog entry (which is a center) should be in more than one blog. Other times, it should only be in one. I really have no way of managing this properly. When I can call anything a center, and consolidate, and cross-post, and give my overall output more coherence, which I think is inevitable, then we'll be closer to working with patterns that connect.

0 Comments:

Post a Comment

<< Home