Adaptive Path Blog

Earlier this week, we invited Bret Victor to come to Adaptive Path to talk about his essay Magic Ink. His essay posits a number of interesting (and controversial) things and it is definitely worth a read. It does take guts to go to a room filled with interaction designers and tell them that not only is interaction design “mostly chuff,” but is actually harmful. So kudos to Bret.

Bret’s argument goes something like this:

There are two types of software, used for two different things. One is for creating content, the other is for learning. (Bret is one of the Neo Cons of Interaction Design.) He mostly sets aside the type for creating content (which as we’ll see, considering his conclusions, is kind of a cheat. Interaction design is most useful for those sorts of tools and applications.) and instead focuses on the kind for learning, which he claims make up the bulk of all applications. These learning applications suffer from lousy information design, that is both poorly designed in the Tufte-sense and takes no advantage of context–where I am, what I have done before, etc. On this point, we are in total agreement. Our devices take too little advantage in the increases in speed, power, memory, and environment awareness that have happened in the last decade.

Where our paths diverge is the dim view Bret takes of interaction design. For starters, he says it doesn’t really exist. All interaction design, he claims, is graphic design (for the display of information and feedback) and industrial design (for the means of input). Interaction design is essentially reduced to mere navigation, and navigation is difficult, ambiguous, cognitively challenging, and even physically harmful, in that it can cause repetitive stress injuries. Thus, Bret advocates reducing the amount of interaction, so that most of the work is done either visually or in the users’ minds instead of having to be manipulated onscreen.

Aside from the fact I worry about moving too much of the cognitive load onto users, I certainly can’t argue that good information design is underutilized in the digital realm. But I can argue that interaction design isn’t only about navigation, even for “learning” applications. In between the push of the physical button and the display of information, there are many decisions (some of which should be based on context) that dictate the behavior of the system: what is shown, when, and how. How is the context known and data about the context collected? All of these things, I would argue, are in the realm of interaction design (and of course the programming and engineering that makes it all possible). Robert Reimann and Jodi Forlizzi’s definition of interaction design (ppt) is about how systems respond in reaction to human behavior. Stuff doesn’t magically appear on a screen for visual designers to display.

And let’s not knock navigation too hard. Yes, no one uses a product for the navigation, but it is navigation that makes some products possible. The scroll wheel and accompanying display make the iPod usable. Without it…well, I guess you would have the Shuffle, I suppose. But navigation can be a pleasurable experience and not necessarily a chore.

It’s interesting to think of interaction design as a tool to only be used as a last resort (”In case of interaction emergency, break glass.”), but there is some truth to it. But since Bret’s dissection mostly ignores applications to create content — which is where interaction design comes considerably more often into play — his essay is less an inconvenient truth than it is an incomplete one.

This entry was posted on Thursday, May 17th, 2007 at 11:19 am and is filed under Interaction Design. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

This post is licensed under a Attribution-Noncommercial-No Derivative Works 3.0 Unported License.