A research agenda for information architecture
Information architecture is a fascinating field, but for some reason, it lacks research, and now it seems to lack innovation.
Now I'm not sure why. IA's are pretty open to input from other fields, at least, they used to be. But it seems that lately IA's are starting to miss out on interesting experiments around categorization. At the same time, IA's aren't coming up with any new ideas. Name one. Faceted classification, sure. A useful idea, that has been out for a few years now, was presented to the IA world by a semi-outsider (Peter Merholz), and is pretty old. And IA's were catching up with that one, not innovating. Experiments with faceted classification didn't come from IA's, and faceted classifications were already commercially implemented when IA's started catching up. Was that the best we could do?
The most recent "new" IA idea was folk classification, or folksonomy. But IA's are just observing what's going on, what sites like Delicious are experimenting with, they are not innovating. Not adding much value, imho. Sorry. We're not even doing the basic research: a lot has been written around folk classification in the social sciences, but has any IA taken the time to look at that experience and write about it, try to incorporate it into IA practice and knowledge. Don't think so.
It get worse. IA lives in the professional realm, it is practiced by in-house IA's and consultants, but is almost non-existant in the academic realm. Books are being written, and some new ideas are being formed, but there is a lack of deep IA research.
So in short: we, as IA's, are not coming up with the new ideas. We seem to have stopped taking ideas from other fields and adapting them. We're just observing experimentation with classification on the web, not participating. Are we becoming irrelevant?
In the past, IA has been clever to build on knowledge generated in other fields: information science, HCI, you name it we took it. In this post I am going to try to give some ideas for an IA specific research agenda.
Without research that lets us build our body of knowledge and answer the increasingly complex questions that are asked of us, IA as a practice might wither and fade away. I think that's conceivable.
First, Peter Morville wrote a good overview of research related or relevant to IA.
So here are some questions, to get us thinking. Please improve on this. This is not an article, it's a braindump blog entry.
Cognitive science.
Cognitive science looks at how we think. The field has build a strong body of work around how people classify or categorize the world.
For example, consider basic level categories. A fascinating concept. Surely, there is relevance to the IA world there. Who will investigate that?
"Compound categories": why are top level categories like "Home & Garden" so popular? And what exactly makes them work? What are the trade-offs? (I made that term up by the way.)
And there is much, much more. The structure of cognitive categories, for example. How does that relate to our categorization effort on the web?
Who will do a linguistic analysis of search terms? There is surprisingly little work being done on what people actually search for. (This BBC analysis is a good one, but we should see dozens like this, not just one, with source data so plentily available.)
Reading:
George Lakoff at Amazon
Business theory.
There are some fields of business theory that are directly relevant to an IA's work. Business processes are one. Tony (?) from CMSWatch asked last year at the IA summit when IA's would start working with business processes and develop a model that incorporates them into the IA body of knowledge (surely business processes should somehow help structure a website). When indeed? I know some work is being done in certain firms, but it's not being shared.
Recently IA discovered the KANO model.
I'm sure there's more I'm not aware of.
Social Science & Anthropology.
Anthropologists have long been interested in how we classify the world, and have developed a large body of work. IA's have mostly ignored this. So which parts of this body of knowledge can be useful for us? Anyone?
How is information sharing social, for example? Where does trust fit? Identity?
Reading:
"Sorting Things Out - classification and its consequences."
And more.
There is a sorry-for-shouting-HUGE amount of experimenting being done with classification, social classification and so on. Most outside of the IA disourse. Let's get with the program. Maybe we should let go of this idea that we know a lot, that we have a lot of value to bring, and start admitting we stand nowhere yet, and we need to experiment, be open to new ideas, and embrace the web. I'll shut up now.
Oh, wait. Finally, we should listen to outsiders. Don't dismiss them because we know stuff they don't.
Please add to the comments what you think someone should research. And if I have missed the point, let me know as well!