heyblog

Digital Ground is essentially a response to what John Thackara called "the design challenge of pervasive computing" at the Doors of Perception 7 conference in 2002. McCullough builds on the content of Doors 6 ("Lightness") and 7 ("Flow"), and on Thackara's statements in particular. Digital Ground's contribution is an exploration of "embodiment" and of the practice of design for situated technologies.

McCullough's own talks at both Doors 6 ("No Place Like Anywhere: Environmental Knowing and Design") and Doors 7 (also titled Digital Ground) are good introductions to his ideas. A version of the Doors 7 talk he wrote for Archis is reprinted at Doors' site.

McCullough was, as I recall, the only one at Doors 7 to say what seemed to me obvious, that anything that flows needs containers in which to flow:

Anyway, this is the heart of the matter: Flow needs fixity, like a river needs riverbanks. Intentionally fixed settings, also known as architecture, provide a necessary context for Flow.

Digital Ground attempts to define what those fixed settings are, and how they can be designed so that "being someplace digital" has meaning. McCullough's term "digital ground" is a nice one: as digital systems become embedded into the contexts of our lives, they form a background, present but unobtrusive. They function, like the physical built environment, as a context in which other behaviors and relationships exist. "The usability of well-made traditional places (i.e. workplaces, home, a corner of the library) now appears as a rich basis for design of context-aware technology." In other words, McCullough says, "life takes place." Providing a sense of embodiment—fixity and a sense of "place"—is the architectural challenge facing interaction design. The first third of "Digital Ground" explores the ideas of "embodiment," "place", and "context" in detail, and connects them to more concepts familiar to interaction designers (such as mental models or affordances).

McCullough introduces a set of typologies for pervasive computing products. Types are "generative design abstractions", which "unite periphery, passivity, phenomenology, adaptability, affordance, facility, appropriateness, and scale." Although that sounds overwhelming, consider a simple urban architectural type: the sidewalk cafe, which probably suggests to you not just some formal requirements, but some social patterns, behavioral etiquette, and may trigger some memories of experiences. This is embodiment at work. In the middle section of the book, McCullough explores a set of "situated types" for pervasive computing at length, along with ten technological "functions" that form pervasive computing systems (such as "sensors detect action"). The types read something like a cross between Alexander's architectural design patterns and interaction design scenarios, and are the most exciting section of the book. It would be a productive exercise to use the ten functions to brainstorm ideas for pervasive computing within the types.

McCullough isn't alone in exploring questions of embodiment in interaction design (although as far as I know, his use of architecture, and of types in particular, is unique). Inevitably, situated software has a blog (though a google search will find you some rather more human voices). Paul Dourish wrote that "embodiment is an emergent property of interactions" in "Where The Action Is" in 2001. "Embodiment" has concerned several other design writers as well, such as Mark Hansen, who seems to be considering the larger context of "new media art." Many of the Interaction Design Institute Ivrea's 2004 thesis projects deal with situated interaction in some way. (Other Ivrea student projects provide some of Digital Ground's design examples.) Clay Shirky's students' best projects all turned out to revolve around projects "form-fit" to few users, often in a specific physical location, for one-time situations like collecting money from fellow students for some shared expenses. Johan Redström's "Designing Everyday Computational Things" PhD dissertation (2001, but not cited in McCullough) also describes some fascinating and creative situated design projects. Redström's realized projects should speak to working designers and technologists; his dissertation may be more the more instructive text for readers interested in applications instead of theory:

Using experimental design, I have investigated how to design computational things for meaningful presence in everyday life.... The specific results from this investigation ... illustrate a set of parameters, and design opportunities, of this design space.... The more general results of this investigation are formulated as a design philosophy for everyday computational things.

All of this should be somewhat reassuring to designers. "Anytime, anywhere, infinitely scalable" design is a hard problem for designers (not to mention for engineers). And always-on, anytime, anywhere turns out to be a huge waste of resources, if nothing else. It's far smarter and more useful to know enough about local contexts of use to provide those resources at the right time, to the right person. In contrast, true "human-centered design" is primarily about delivering products that fit within human-scaled structures and situations. (I'll point out again that McCullough's typologies, Shirky's form-fit software, and Redström's projects aren't built on hierarchies, the default structural form of most software intended to scale.)

Digital Ground has its faults, though. McCullough brings up, but doesn't deeply engage, ethical and social consequences. Potential military or surviellance applications of pervasive computing are mentioned in passing, but there's no call for designers to take strong critical positions against these. It's easy to imagine how someone coming at this topic from a genuinely critical, even subversive approach would see "the design challenge of pervasive computing" wholly differently. For example, see Adam Greenfield's response to Ubicomp, or much of Anne Galloway's work. And although McCullough cites their work, his outlook is less critical than Anthony Dunne and Fiona Raby.

It's also worth mentioning that the use of the word "architecture" in interaction design, or in software in general, is far from uncontroversial (Peter Lindberg has written several great posts at Tesugen.com unpacking the concept of "software architecture".)

Throughout, I often wondered how McCullough might treat the business side of design. A long discursive section ("Accumulating Value") at the end of the book proposes that traditional economic measures of value that center on utilitarian measures' failure to take into account sustainability as a goal, and that "places remain great accumulators of value." Essentialy: trust me, pervasive computing will find its own valuation criteria, just as all other places do. It's important to differentiate valuation from evaluation, a topic that gets no attention in Digital Ground, yet which is also of market importance.

That's because the most valuable contribution of "user-centered design" may turn out to have been its insistence that the designer's responsibility to the user does not end with idea generation and production. Usability engineering, for better or worse, provides qualitative and quantitative methods for the evaluation of the effectiveness of a given design before your company makes a costly commitment to it. You might prefer that tiny text and fancy drop-down menu, but it's easy to tell when customers don't.

Redström writes "We argue that the coming ubiquity of computational artifacts drives a shift from efficient use to meaningful presence of information technology." Use is pretty easy to evaluate: did someone use it correctly, on the first try, and did she learn to master it in a reasonable amount of time? Has her productivity increased? These questions might have no real meaning when applied to products subsumed into the "digital ground." Should you measure how little a customer notices the product she's purchased, for example? What are the right questions to ask to measure the effectiveness or value of a design? This is a very serious issue once design leaves academic spheres for market-driven ones. It may be that, as McCullough writes, "at the heart of a new economy is the proposition that finance and the means of production are not the only forms of capital," but we still need ways to measure the appropriateness of given design solutions before they're placed into the world--Jakob Nielsen won't be banished easily from the next economy, either.

(Redström, by the way, addresses questions of evaluation at length in Chapter 9 ("From Use to Presence"), although his answers certainly won't satisfy a product manager or channel marketer looking for hard data. This seems to be an attempt at quantifying embodiement, although it's bafflingly specialist.)

It's also nice to see that McCullough's writing voice has evolved in a positve direction. Digital Groundis much more readable and enjoyable than his previous book Abstracting Craft, which really is quite tough going through long stretches. Although I'd prefer a few more concrete examples of pervasive interaction designs and designers in the body of the text, McCullough manages the tricky balance of academic and practical material nicely. Digital Ground should be required reading for anyone interested in where interaction design is headed.