wikis and information architecture

I participated today in an IRC chat on "wiki gardening" for the Atom/Pie/Echo syndication format project wiki. There are over 700 pages on the wiki, which makes it pretty darn big by anyone's standards. Of those, more than 80 are "orphaned" pages (pages that no other pages point to).

The topic of the chat when I joined was "organization of the wiki," and it became painfully clear that most of the people generating and organizing content on the site had little or no background in information architecture. (I was particularly amused when someone asked what the difference was between the FrontPage and SiteNavigation items on the main navigation bar of the wiki, and I realized that despite the link, the SiteNavigation page didn't even exist.)

It got me thinking about Christina Wodtke's book Information Architecture: Blueprints for the Web. In her introduction, Christina has a section entitled "Why Bother?"--and I'm going to quote directly from it, because the example she provides is so perfectly suited to this situation. (I've done a minor edit in the third paragraph, however.)

In 1844, Sara Winchester, convinced that the ghosts of all the people who had been killed by the Winchester rifle would come after her for revenge, asked a psychic for advice. The psychic told her the way to keep the ghosts at bay was to build a house--that the sound of hammers must never stop. For 38 years, Sarah kept a team of workers building on to her house. She never made blueprints, although occasionally she would sketch out what she wanted on a random piece of paper or a tablecloth. The resulting house is a rambling four acres of stairs that lead to the ceiling, doors that open to brick walls, and windows in the floor: The Winchester Mystery House.

Flash forward to 1998. A CEO has heard about this Internet thing. Convinced his company will fail if he doesn't deal with it, he asks a consultant for advice. The consultant tells him that to keep his company afloat, he must build a web site. For all these years, until now, teams of people have been working on this site without a blueprint. Occasionally they have plans for one area or another drawn on a cocktail napkin by the east coast sales team and faxed to the design team on the west coast.

Now the web site is huge, with thousands of pages with links that lead nowhere, marketing speak that says nothing, and outdated facts: The Winchester Mystery Site a giant wiki!

Okay, okay. She didn't say wiki. But what a great parallel. Teams of people working on the site without a blueprint. Occasional partial plans generated by a subset of people in chats or meetings. Hundreds of pages, links that lead nowhere, and outdated/orphaned content.

Wikis, as Clay Shirky has said, remove personality from the process. But by placing the emphasis entirely on "bottom-up," emergent content, they also remove organization and structure from the process. The resulting "mystery site" becomes frustrating for users who aren't already intimately familiar with every nook and cranny.

I'm not a web services developer (and I don't want to be one when I grow up, either). But I am an information architect, and an interested observer of this process. And in general, I don't like to complain publicly about something that I'm not willing to help fix. So I'm going to take a stab at creating some navigation pages for the project wiki, focused on the kinds of users that the wiki currently has, as well as those it's intended to serve. That means not just experienced developers, but also interested laypeople. And probably some categories in between.

Because it's a wiki, it's by nature a collaborative space, and you can play, too. Once I've got some first cuts at organizational structure done, I'll put them on the SiteNavigation page of the wiki. And by then there may well be a documentation-focused blog to support the effort (work is underway to make that happen), where comments and suggestions for revisions will be welcome.

I leave on a family vacation this Monday, with limited or no Internet connectivity while I'm there, so I don't know how quickly this will happen. But I will try to get something up before I leave.