The second half of the third chapter left me feeling like I was swimming in pretty deep water. This isn’t a bad feeling. In fact, reading material that I’m not immediately understanding makes me think sometimes. However, I feel like I’m approaching the heart of this book where Dourish is explaining his own research program, and I’m feeling a bit at sea, like I’ll have to read this section again.

Here are a few notions that floated my way in this section of the book. One important concept that I want to keep thinking about has to do with how designing software requires one to use abstractions and metaphors. And this relates to the comments that Joe wrote in response to one of me earlier postings about Dourish’s book. Part of an answer to Joe—and others who might wonder whether graphic interfaces and metaphoric designs artificially cover complexity—is Dourish’s comments that “abstraction makes it possible for us to treat a complex set of computational behaviors as a simple, higher-level object out of which we can build something new and even more useful” (p. 82).

It’s interesting to follow Dourish’s complication of the notion of abstractions of interfaces hiding features of interaction. His example on pages 83 and 84 of how folders on a desktop might represent files that act differently depending whether or not they are on the desktop of a computer helps explain the problem of abstract representations that the user can’t control.

So if I’m getting this right—and I think I have more questions than answers—Dourish is asking for computer software and systems that provide the user with specific, detailed, ethnographic information. Is changing from WYSIWYG to HTML an example of “computational reflection"? (p. 85) Not really, right? Because you can’t reconfigure the program. Ah, but with Manilla you can, if you have the patience.

Dourish makes a distinction that I like between “space” and “place” (p. 89 & 90). This is something that I could relate to, thinking about how I’ve designed chat rooms (spaces or places?) in my office in the TappedIn MOO. And he also mentions what has been my theme of late when it comes to computers: language variation. “Structure of the environment is often a key issue in controlling how interactions develop” (p. 90). Yes, that makes this seem worth slugging through if it helps me to understand how computers structure environments that control linguistic (and other) interactions.

In general, I do get this notion of providing “frameworks that help system designers invest interactive systems with sociological understandings” (p. 96). At least I think I do, but I’m spinning a bit from all of this theory.