heyblog

« IDEA and Interpretation | Main | They should be thanking me »

November 27, 2006

Review: "Designing Interactions"

A few weeks ago Bill Moggridge’s new book Designing Interactions arrived at my house. It’s an enormous (more than 700 pages) and nicely designed book. It feels fresh and broad; there are interesting sections on mobile devices, patterns of technology adoption, play, service design, and critical design. Moggridge’s basic premise that good interaction design is the result of learning about people and prototyping solutions is very solid, and his examples illustrate it well. I think if you do this kind of work, you’ll especially enjoy the chapters on the mouse, the menu, or the Palm Pilot, and seeing lots of sketches and diagrams and screenshots. It’s fascinating to see things like Bill Atkinson’s sketches for the Apple Lisa’s menu system. Those glimpses into the field’s past make Designing Interactions an important book, the first attempt at a real cultural history of the field of interaction design, from its beginnings with Douglas Englebart and Xerox PARC, through current work designing for ubiquitous computing. Unfortunately, Designing Interactions suffers from some very serious flaws, and I hope that all readers will bring an especially critical eye to it.

In this review, I’ll focus on a few things in particular that bother me about the book and Moggridge’s approach to the material. First, he overuses (and misuses) the interview format, without providing authenticating evidence for the stories told by his subjects. Long stretches of the book feel like little more than mindless design-star fan journalism about the author’s pals and their companies. Finally, Moggridge never misses an opportunity to use his own company, IDEO, in a case study, or one of its employees in an interview. But by failing to make that vested interest clear, Moggridge turns the book into a marketing project. Bruce Sterling’s jacket blurb describes Designing Interactions as “a labor of love.” In this case, love is, if not blind, than pretty nearsighted. The book really should at the very least have the word “IDEO” in its title.

His story or history?

Designing Interactions reminds me of The Art of Unix Programming (a surprisingly fascinating and readable book, by the way). Like that book, Designing Interactions is a really both a cultural and technological history, made up of case studies, interviews, and technical material. There’s a variety of content, including some hand-drawn diagrams and a photo essay or two. But it’s those interviews that really make up a big part of the text. Moggridge excerpts his interviews at length, sometimes running more than a page of small text. (The complete video interviews come on a DVD with the book and are at the book’s website.)

It’s clear that Moggridge means this to be an approachable book, and the interviews have a good-natured, casual tone. But it’s a cumbersome way to structure a text. Throughout, the reader is constantly jumping between Moggridge’s voice and someone else’s. Several subjects have the annoying tendency to speak in informally dramatic ways, lapsing into quoted speech that’s hard to read. Inside a several-paragraph-long interview excerpt, there are passages like “If an expert comes along and says, ‘Yeah, yeah, I can do that. Where’s the super Orc? This is easy, perfect, perfect, perfect. This is so easy!’ Then we can say ‘Maybe you should be playing in expert mode.’ Then it’s ‘Ah, X, X, X…good? Hey where’d my perfect go? Good? Fair? Whoa, okay, I gotta pay attention.’” Catch all that? That’s an especially awful example (by Bing Gordan from the “Play” chapter), but Moggridge allows that kind of speech far too often.

Conversational speech isn’t an especially good format for deep personal reflection. So naturally, Moggridge’s interviewees tend to minimize their own failures, missteps, or unsuccessful designs, and gloss over difficulties. That makes the whole thing feel like a whitewash. I guess this kind of access is what the MIT Press means by an “industry insider’s viewpoint”, but it unfortunately makes the book something of a hagiography—complimentary biographies of designer-saints, whose every effort is nothing less than beautiful, innovative, useable and useful. Building a history out of fond memories means that Moggridge’s subjects are bathed in a sort of golden California glow of optimistic technophilia, bright ideas, and sudden success.

Jonathan Raban writes about memory in his book Passage to Juneau:

Memory always has its own dark purpose, often hidden from the rememberer; and it is a ruthless editor, with a facile knack for supplying corroborative detail. It’s impossible to draw hard-and-fast distinctions between deep-dredge memory, retrieving material directly from the silt in which it has lain for many years, and the shallow-dredge variety, in which one remembers only an earlier act of remembering.

I kept thinking of those lines as I read Moggridge’s interviews. First-person accounts can add great insight and depth to otherwise dry historical facts or documentary evidence. But that “facile knack” for editing is why historians don’t rely only on other peoples’ memories to write histories. Memory of “an earlier act of remembering” easily becomes nostalgia or drifts into sentimentality, where all the rough edges are smoothed entirely over. This is why Moggridge’s methodology would cause any historian concern: the interviews are quite recent (2002 through 2005). That means many of Moggridge’s subjects are discussing events of ten, fifteen, or even twenty years ago. What Bill Atkinson says in 2005 about something he remembers happening in 1985 might be interesting, but on its own it’s unreliable and unauthenticated. Yet without exception, Moggridge presents his subjects’ recollections uncritically as historical fact, without other material to support (or disprove) them. To be clear, I don’t disbelieve Moggridge’s subjects’ statements. But the lack of secondary authentication means I can’t easily tell their own interests (publicizing themselves or their companies, winning old arguments, claiming turf or expertise) from fact.

I’m guessing Moggridge wouldn’t claim to be a historian, though that certainly doesn’t get him off the hook. Moggridge stakes out sequences of events, trends and forces, and the actors involved in them. He attempts to situate his subjects in particular cultural and temporal contexts; that’s history, and it should be held to a historian’s standards. In his preface, Moggridge calls his approach “tell[ing] true stories about designing interactions, rather than structuring the content around a more polemic point of view.” The worrying term “true stories” aside, would a “polemic” stance have really hurt the overall accessibility of the book? Questioning assumptions and supporting assertions with other evidence wouldn’t ruffle his buddies’ feathers, and might have forced Moggridge to a more critical and insightful point of view.

And who doesn’t imagine going home one evening and inventing a whole paradigm like the menu bar, as Larry Tesler says Bill Atkinson did? It’s a designer’s lucid dream: “He’d thought the whole thing up in one night! I can’t imagine what happened that night.” In fact, other parts of Tesler’s account suggest they’d been working methodically through ideas for months before arriving at that one. Was there really a midnight inspiration, or could 25 years have colored the memory? (Raban again: “All first-person narratives are like this. I thought it was a body. You thought it was a body. We were both wrong.”)

So what, you say? Isn’t the point just to inspire the reader with great stories about clever solutions? Isn’t this just like reading someone’s blog, where everything is obviously subjective anyway? No. The danger of accepting peoples’ memories at face value isn’t just that they might be misremember a fact here and there. Nor is it that the design of buttons and menus isn’t a moral, cultural, and aesthetic imperative; it is. But in “true stories” like Tesler’s, Moggridge implies that the “people and prototypes” approach to design is somehow “natural,” and that designers and their clients who practice it are guaranteed a successful outcome. Another review of Designing Interactions comments that Moggridge’s “team[s] almost naturally followed an iterative prototyping approach to bringing their ideas into the world” (emphasis mine).

I do believe “people and prototypes” is both an effective and appropriate approach, and I recognize Moggridge’s and IDEO’s contributions to it. But what’s missing in Designing Interactions is the understanding that the very idea of “a design process based on people and prototypes” is not “natural”, it is itself a piece of technology, culturally situated and rooted in Moggridge’s Silicon Valley of the last few decades. There’s mythmaking going on here, reinforced by the affable tone of the design gurus on parade. Unable to be critical of his own convictions, Moggridge has written a narrative that can do nothing but reinforce them.

IDEOs Mio!

Throughout Designing Interactions, the IDEO design firm is never far from the reader’s awareness. Few sections fail to include an IDEO product or IDEO contribution. IDEO employees make up many of Moggridge’s interviews. The book’s final section on process and prototyping includes several pretty photos of stylish IDEOers in at work in action-designer poses, includes excerpts from IDEO publications, and promotes the company’s Method Cards. Yet for all of this, Moggridge’s is oddly vague about his own role in founding the company. The fact is stated in the one-paragraph author biography in the book’s flyleaf, then never appears again, though it’s alluded to through mentions of his previous company IDTwo. Moggridge should have made his role in IDEO and his relationships with his interview subjects abundantly clear. Most historians would have difficulty including that sort of biased material in their work, and would at the very least try to broaden it with other sources.

Don Norman’s jacket blurb says: “This will be the book—the book that summarizes how the technology of interaction came into being and prescribes how it will advance in the future.” But the IDEO bias in Designing Interactions should bother all readers. Yes, IDEO played a significant role in shaping some of the major products in interaction design history. But at times it’s as if Moggridge has included a section or interview specifically to give more ink for the company, which hardly needs help marketing itself. Moggridge should have been far more transparent about his own role at IDEO and about his relationship with the people, products, and clients in the book. (I wonder, were IDEO employees paid for their interview time?) After all, he stands to gain personally and professionally from promoting IDEO as the origin of (and exclusive experts on) interaction design. Again, from a historian’s point of view, the conflict of interest here is serious. I hope that the first part of Norman’s blurb won’t turn out to be true.

I found myself asking again and again, who is Designing Interactions really meant for? Its biases would make it a poor student textbook. Is it for other designers? I suppose so, though how useful is a long series of a-ha! moments and reminders to “learn about people” and “prototype often” as a design text? I’m worried that Moggridge is writing mostly for his peers—the ones who appear in the book—as a way of “telling their story”, an approach that’s better suited for magazine articles (maybe) than for a significant work of research. That might explain why there’s so much emphasis on company histories and intrigues: who worked for whom when, who managed what group at what company, and so on. For instance, the section on Will Wright’s company Maxis deals mostly with the company’s financial history, and how its acquisition by Electronic Arts affected internal management. Why include this? Steven Johnson’s Emergence has a far better discussion of the interaction design of Wright’s games. The chapter on the Internet is all but useless, as we get more about Google’s corporate history than on its product development process. That’s a shame, considering Google’s emphasis on rapid iteration on ideas is such a great example of Moggridge’s methods.

Another view: Dan Saffer

It’s interesting to compare Designing Interactions with Dan Saffer’s recent book with a slightly different title: Designing for Interaction. Both authors use interviews, even with some of the same people (Larry Tesler and Brenda Laurel), but are limited to three or four questions and tightly-edited answers. (I’ll also admit that a couple of Dan’s subjects are friends of mine.) Designing for Interaction’s interviews are short sidebars that expand on each chapters’ main idea, without feeling distracting. That’s an appropriate and useful use of that kind of research.

Moggridge’s last two chapters make up a practical section, about the length of Saffer’s whole book. It feels a little rushed. The first 500 pages left me really wanting some solid instructional content that’s not here: task diagrams, wireframes, radio buttons, that sort of thing. Yet Saffer still manages to cover a lot more of the nuts and bolts and day to day work of interaction design. And even though the two books’ titles are close, for me the difference is pretty telling. “Designing for Interaction” follows from Saffer’s definition of the field as “the art of facilitating or instigating interactions between humans (or their agents), mediated by products.” Isn’t there something vital and slightly humble in Dan’s inclusion of “for” and the singular form of “interaction”? Designing for interaction means getting out of the way, designing interactions means deciding what “the way” itself will be.

Posted by Andrew at November 27, 2006 09:55 PM

Trackback Pings

TrackBack URL for this entry:
http://www.heyotwell.com/cgi-bin/mt/mt-tb.cgi/149

Comments

Posted by: molly at November 29, 2006 03:30 PM

Posted by: heyotwell at November 29, 2006 04:10 PM

Posted by: Steve Portigal at November 30, 2006 03:16 PM