Steve sent around an interesting link the other day:

http://www.37signals.com/papers/introtopatterns/index.html

It’s an article that seeks to apply to web design the notion of A Pattern Language that was famously and wondrously developed by Christopher Alexander several decades ago in a book of the same name. Alexander’s book (the second of a trilogy actually) was about architecture, but his unique approach to thinking about how we create and use space has spawned a kind of minor cult, of which I’d say I’m a fringe member. Which is only to say that when I first came across Alexander’s book many years ago I was deeply affected by his approach to structuring knowledge via patterns, and it has remained an inspiration ever since, though somehow I keep giving away the copies I obtain and never have one of my own (sigh).

Alexander’s ‘pattern languages’ are sort of worlds within worlds, a way of expressing networked relations between objects so that when you think one you necessarily engage the others. In many ways his book entirely accidentally foresaw the basic structure of the hypertextual web, and has proven useful in devising and managing database taxonomies too (just learned this through a quick google search…google is all about pattern languages too I suppose). When I look at this very blog with its categories and trackbacks and various hierarchies, it too is a kind of pattern language.

Here is a link to a site that offers a surf-able (tho stripped down) version of his book:

http://downlode.org/etext/patterns/

One of the things that the 37signals essay picks up on is what I’d call the ’scribble’ effect that Alexander used, or what maybe is sometimes called back-of-a-napkin thinking. It’s a quick and dirty but amazingly effective way of conveying information and organizing thoughts. Of course it has become a bit trendy too, not always in a good way. I remember seeing the napkin on which Daniel Liebeskind submitted his $170-million design for the Royal Ontario Museum redesign, which he won. but it seems that in Liebeskind’s case his quick-and-dirty approach was really something of a masquerade for a lack of insight and research because the project’s way over budget and has gone back to the drawing board numerous times.

Anyway, I think Alexander’s pattern languages are as useful now as they were when he created them, if not more so. The book is an extraordinary resource for thinking about networks and knowledge and design. Not to mention building a house!