Sunday, March 07, 2004

Tried to comment on Boxes and Arrows: We Are All Connected: The Path from Architecture to Information Architecture but the comment server seems down. So here's the rant.





The biggest problem with the article is that *nothing* in his discussion of building web-sites actually follows from his initial assertion that Information Architecture is like Architecture.

I'm increasingly attracted to the analogies between architecture and information system design (both at the programming level and at the user interaction level) so I was really hoping for something interesting here.

But as far as I can see, the only message we're meant to take from the analogy is that we need a separation between "grand designers" and the builders and engineers who do the work. And that we need a lot of contracts and specifications.

Is this the lesson from architecture? One of my favourite books is "How Buildings Learn" which is partly an extended rant against this kind of thinking, and a celebration of bottom-up design that emerges from users and builders making continuous small modifications. Although the author of this essay namechecks the book and uses the "e" word, it doesn't seem that he allows it to disturb his off-the-shelf preconceptions about web-development in the slightest.

Another theme of HBL is the danger of "Magzine Architecture" and a complaint that architects don't think through or evaluate long-term usability of buildings. Yet this author puts "test before putting it together for real" in his check-list of things IA can learn from A.

So, what testing is it that architects do? (Note : I'm sure the *engineers* do a lot of testing.) And in what sense is this something which IA - which got it's testing obsession via UI and human factors people (like Nielsen), and maybe ultimately from *industrial* design - can learn from architecture? Possibly IA is ahead of A here.

Once we get to the real life example, this is all standard web-design stuff. The architecture comparison isn't even mentioned. There don't seem to be any innovations inspired by architecture. Or even attempts to show parallels.

Seems like the author knows a lot of interesting ideas : Alexander, Brand, Nielsen, emergence etc; believes in connectedness; but doesn't seem to understand *how* these things are connected : the logical implications between them, and the structural constraints that they place on each other.

No comments: