datacloud: After Usability

October 23, 2003

After Usability

Even as the computer industry--and society--have increased attention paid to "< href="http://usability.gov/">usability in software and hardware design (and product design, communication, and other circles), important work has begun to coalesce around the notion of usefulness. As an article in the Guardian reports,

As Pentagram's Robert Brunner argued last week at the HITS (Humans/Interaction/Technology/Strategy conference in Chicago: "It doesn't matter if something is usable. What matters is that it is useful. Even better if it is desirable." To break the deadlock in user-interface and product design, we need radical innovations. We could start by going beyond the text- and list-based interface of Google, and should be debating what could be learned from Grokker's information visualisation-driven product.

I've long thought that the most interesting and useful computer experiences emerged from breakdowns. This is true of education in general: one doesn't learn important things if they're transparent. One learns when theories and concepts collide, and when realities smack them upside the head. The most unusable technologies are frequently the most useful. Usability models, though, tend to conceive of breakdown--even visibility--as a bad thing, to be avoided. Sure, we can't survive if everything breaks down completely. But if things aren't stress facturing, if the system doesn't crash occasionally, you're not pushing hard enough. Posted by johndan at October 23, 2003 04:01 PM | TrackBack

Comments

This is a great point�as technical communication embraces usability more closely, and it influences teaching more directly, the focus on error could cause some of the same problems which occurred in other forms of composition.

I wonder if we need a Mina Shaughnessy-style theory for the �errors� which occur during usability tests. Or a better way to account for the successes which occur during testing. Right now, success is transparent�not exactly the best situation for instruction.

cbd.

Posted by: cbd at November 2, 2003 07:04 PM

[deleted some info here about an internal server error Brad reported via private email, but I'll reprint the rest of the email I sent him back here.]

I like your idea of applying Shaughnessy to this whole usability/usefulness issue. There doesn't seem to be a good taxonomy or other framework for understanding error (or "successes" that don't really help). TC has long suffered from the "transparency" issue--good communication is transparent--and, unfortunately, we seem to be moving away from that. (Even Norman's new work on "emotion" and technology, which gets beyond "affordances" and transparency [finally] still ends up mystifying the issue.)

- Johndan

Posted by: Johndan at November 2, 2003 07:50 PM