I was listening today, as I am wont to do to an old addition of the Boing Boing Boing, a podcast from the editors of Boing Boing. It's good stuff, and even though it's not relevant to the news of the day, I usually don't care. So the guy they were talking to, wrote a book on the 1854 Cholera Epidemic, which is interesting, but his previous work is on dynamic systems theory--more or less.

So here's the interesting thing. In systems work on cities, the sort of "cutting edge" as it were is analyzing the bottom up stuff--neighborhoods, informal communities that build around places like playgrounds, schools, and market places, and other contact [1] interactions--rather than top down stuff like governments and cultural factors. (Sorry about that sentence folks!)

In cognitive psychology, the more cutting edge models, the ones that probably do the best to explain psychological reality are the top-down ones. They're epistemologically difficult because it's hard to isolate variables in top-down systems, but the bottom up systems don't tend to scale well [2], and don't mirror some key experiences. Clearly both have to work at the same time (and most cognitive systems include a black box, to be fair) so there's a lot left to be discovered.

I guess what this post is all about is the fact that I never really thought to take cognitive-style systems theory out of the mind, nor did I think that when I did, the "radical" positions would be reversed. Part of the issue is that psychology has always been a very bottom-up field, Wundt was a bottom uper and Skinner was the very picture of a bottom-upper. [3]

The key part of the distinction between top-down and bottom-up is what you take for the unit in the system, of course. Chunks of information in your mind, is way different that a person in a city, or a city block on the city, or an IRC channel online.

Anyway. That's what I've been crunching through. Hope you're interested.

[1]To borrow a model from Samuel Delany's Times Square Red: Times Square Blue. Contact is the random, unplanned interactions that happen as a result of urban living, and is in contrast to "networking," which is a goal-based social activity designed to further specific goals in specific situations.
[2]Bottom up explanations tend to work best when you have cognitive systems that work on very simple kinds of problems, and very simple inputs. When you get into "real" life situations its hard to imagine that even the brain in all it's glory, can parse everything that it needs to.
[3]...And if you leave out Totem and Taboo and Civilization and it's Discontents, you could probably make the argument that Freud's work was bottom-up for the most part. But I don't know if that's even worthwhile.