DMI vs. Agents
I just finally read the Shneiderman vs. Maes debate in the Nov/Dec'97 "Interactions" on Direct Manipulation vs. Interface Agents. What's striking is that the issues are almost EXACTLY the UCD vs. LCD debate, but at a different level. Some points:
- Pattie Maes claims that we need agents because the common user is not a professional, as it has been in the past, but a non-expert who needs to know what to do when.
- Ben Shneiderman counters that direct manipulation interfaces have been shown to be easiest to learn – though I then wrote in the margin "learn about what? the interface? the underlying model? the domain? the process?"
- Pattie Maes responds the accusation of Jaron Lanier that agents "make us dumb" by saying "To some extent it's true. If I don't fix my car, then I'm not going to learn about fixing cars. However this does not constitute a problem. As long as there's always an agent available or I can call one by a motor association like AAA, then that's fine. It's too bad that I will never learn about cars, but I want to learn about other things instead." This sounds just like the issue of "glass box/black box" that Cindy Hmelo and I have raised, and the issue of scaffolding that never fades.
So, how do we deal with these two levels? How are they similar and different?