Just How Far Beyond HCI is Interaction Design?

While reading a new textbook, “Interaction Design: Beyond Human-Computer Interaction” by Jenny Preece, Yvonne Rogers and Helen Sharp, I recognized my own reactions from several recent conversations, talks and “My concern is that the book misrepresents the field of interaction design and hence limits its potential.”papers. The feeling was familiar: the authors were adopting, and adapting, the “design” label a bit too loosely. The academic field of human-computer interaction (HCI) has strong roots in the traditions of behavioral science and engineering. Re-labeling it as interaction design does not in itself move it “beyond.”

Interaction design is a fairly recent concept, albeit growing steadily in terms of professional practice, higher education and even job descriptions. It clearly owes part of its heritage to HCI, even though the turns within established design fields—such as graphic design, product design and architecture—towards the digital material are every bit as important.

There is no commonly agreed definition of interaction design; most people in the field, however, would probably subscribe to a general orientation towards shaping software, websites, video games and other digital artifacts, with particular attention to the qualities of the experiences they provide to users.

The word “interaction” in interaction design captures the time-based, and at the same time, nonlinear nature of the digital, a quality that sets it apart from most if not all other design materials. “Design” is of course a word with multiple meanings, but some typical connotations in more mature design disciplines and in design theory include the parallel emergence of question and answer, the activity of exploring possible futures, the synthesis of reason and emotion, the intervention on many simultaneous levels in a design situation.

The scope of interaction design opens up possibilities for genuinely better user experiences of information technology. HCI has contributed a great deal to the elimination of obvious problems for the users, but its focus on goals, tasks and usability makes it rather limited in terms of positive innovation. My concern is that the book misrepresents the field of interaction design and hence limits its potential.

In the book, there are some excellent interviews with prominent representatives of the interaction design field. In one of those interviews, Gitta Salomon states that (p. 33) “interaction design is a design discipline.” This observation is not taken seriously in the book as a whole. Many issues taken for granted in HCI need to be rethought in an interaction design textbook.

In the preface, the authors define interaction design as “designing interactive products to support people in their everyday and working lives.” But does it make sense to say that a computer game supports people? Even if it “supports” the player’s assumed goals of experiencing excitement or challenge, how does it “support” the player’s boyfriend’s desire to see a bit more of his girlfriend? Is a teenager’s experience of spending time in an online chat community primarily a “supportive” one? Does a piece of techno-critical digital art, such as the work by Anthony Dunne and Fiona Raby, “support” the viewer?

My point is not that “supporting” is necessarily bad, only that it implies a certain ideology: an HCI perspective of goal-driven users whose use should be made as effective and efficient as possible. Interestingly, the same paragraph goes on to quote Terry Winograd’s suggested definition of interaction design as “the design of spaces for human communication and interaction.” The difference between the two definitions would be worth a book of its own.

On the nature of design, the authors (p. 166) rely on “the definition of design from the Oxford English Dictionary [which] captures the essence of design very well: ‘(design is) a plan or scheme conceived in the mind and intended for subsequent execution.’” This notion of design as a plan-then-act activity fits well with the software engineering approach permeating most contemporary HCI, but perhaps not so well with design theory. Planning is rather viewed as performed through acting, expressing, communicating. The execution is the planning; the planning is the execution. Together, they are part of the evolution known as design. The prominent design theorist J.C. Jones gives the following advice:

“First, recognize that the ‘right’ requirements are in principle unknowable by users, customers and designers at the start. Devise the design process, and the formal agreement between designers and customers and users, to be sensitive to what is learnt by any of the parties as the design evolves.”

The implications of this view for practical IT development are substantial and largely unclear. That is precisely why gifted researchers and writers should devote their efforts to it.

There is also the question of what to design. A HCI perspective encourages the view of adapting new technology as painlessly as possible to existing users and practices (p. 5): “one can be more principled in deciding which [interaction design] choices to make by basing them on an understanding of the users. This involves: taking into account what people are good and bad at, considering what might help people with the way they currently do things, thinking through what might provide quality user experiences, listening to what people want and getting them involved in the design, using ‘tried and tested’ user-based techniques during the design process.”

But design is innovative; it is about exploring possible futures, where the users as well as the technology are different from today. In some situations, it even makes more sense to think in terms of designing the users. As Terry Winograd points out in his interview (p. 71): “one of the biggest challenges is what Pelle Ehn calls the dialectic between tradition and transcendence. That is, people work and live in certain ways already, and they understand how to adapt that within a small range, but they don’t have an understanding or a feel for what it would mean to make a radical change, for example, to change their way of doing business on the Internet before it was around, or to change their way of writing from pen and paper when word processors weren’t around. I think what the designer is trying to do is to envision things for users that the users can’t yet envision. The hard part is not fixing little problems, but designing things that are both innovative and that work.”

Consequently, I’d like to make three assertions:

First, interaction design is a design discipline, which means something other than the science-and-engineering perspectives of HCI.

How, then, can we approach the question of quality in interaction design? How can we tell good from bad, and how can we devise development processes and cultures of practice to increase the chances of reaching good designs?

The HCI answer (p. 19) is to express quality in terms of measurable usability goals and to address “the rest” as we please: “Usability goals are central to interaction design and are operationalized through specific criteria. User experience goals are…less clearly defined.” But taking the approach suggested by the authors to address these less clearly defined user experience goals results in equating a desktop video conference system for distance learning with an online community that provides support for people who have recently been bereaved (p. 20).

A more well-grounded design approach would offer better ways of approaching the quality issue. Some useful starting points may be to view design as knowledge construction within a community of practice, to consider design-theoretical approaches to articulating the languages and meanings of products (so-called product semantics), and to consider product or use genres as knowledge-organizing systems.

An issue of particular interest is the possible role of critics in interaction design. One can imagine a field of interaction design criticism in analogy with more mature design fields such as architecture or graphic design. This appears problematic from a HCI perspective (p. 182): “Finding measurable characteristics for the user experience criteria is even harder, though. How do you measure satisfaction, fun, motivation or aesthetics? What is entertaining to one person may be boring to another; these kinds of criteria are subjective and so cannot be measured objectively.” However, it is possible to talk about good and bad interaction design also in broader contexts. A few examples exist of more relevant interaction design criticism, but there is clearly room for much development.

Secondly, the notion of quality in interaction design is not well developed. Neither are the social structures needed to develop and sustain the notion. A HCI perspective is not the most appropriate starting point.

The emphasis on usability and task support also carries a notion of aesthetic qualities as equivalent to visual embellishment. On the topic of simplicity in web design, the authors state (p. 27) “A certain amount of graphics, shading, coloring and formatting can make a site aesthetically pleasing and enjoyable to use… The key is getting the right balance between aesthetic appeal and the right amount and kind of information per page.”

But if aesthetics concern the sensual and perceptual qualities of experience as a whole, how can there be a tradeoff between aesthetic appeal and the right kind of information? “Right” is an aesthetic judgment in this context. In another interview in the book, Gillian Crampton Smith makes the following observation (p. 199). “Obviously there’s the aesthetic of what something looks like or feels like but there’s also the aesthetic of how it works as well. You can talk about an elegant way of doing something as well as an elegant look.”

Separating the usability of a system from how it affects its users, factoring out ”the aesthetics” in terms of “pleasing shapes, fonts, colors and graphical elements,” is problematic. Every interaction involves feeling as well as intellect; aesthetic qualities reside in the overall interaction, which is determined above all by the functions, structures, social action spaces and temporal qualities (the dynamic gestalt) of the system.

Third, it makes sense to talk about aesthetic qualities of interaction, even though we lack an adequate language as yet to do so. But the language of HCI is not the best place to look for inspiration.

To conclude, I think the book discussed here represents a larger movement within HCI towards design. My point is that this movement involves a shift in philosophical foundations as well as professional practice. The book is a very capable presentation of contemporary HCI, covering relevant technological developments as well as the growing insight that HCI “users” in fact do their (effective and efficient) work in social environments. It is pedagogically well structured and presented, and would be an obvious recommendation for any foundational HCI class.

But it does not go beyond HCI. More specifically, it is not a book about interaction design. Taking the philosophical and practical shifts more seriously will involve dealing with the issues I’ve raised if we truly want to go “beyond HCI.”

For more information:

Posted in Big Ideas, Workplace and Career | 8 Comments »

8 Comments

  • frank elbert

    April 23, 2002 at 3:19 am

    I quote Jonas:
    “I think what the designer is trying to do is to envision things for users that the users can’t yet envision. The hard part is not fixing little problems, but designing things that are both innovative and that work.”

    Well, isn’t this exactly what, for example, IDEO and Cooper does for a living for a long time?

    Cooper about interaction design: “It is a synthesis, however—more than a sum of its parts, with its own unique methods and practices. It is also very much a design discipline, with a different approach than that of scientific and engineering disciplines.” http://www.cooper.com/newsletters/2001_06/so_you_want_to_be_an_interaction_designer.htm

    IDEO: http://www.theartofinnovation.com

    In my perception interaction design is a design discipline like any other design discipline (architecture, graphic, interior, fashion, etc.). The academic insights of HCI are part of the interaction designer’s toolbox. The interaction designer uses those insights when designing his creations. (this way we can say that both http://www.amazon.com and http://www.natzke.com or examples of good interaction design)

  • sherlock_yoda

    April 25, 2002 at 4:40 am

    I think the notion of ‘interaction design’ has been useful in bringing to light the fact that many HCI professionals don’t actually know how to create excellent or innovative products (sensing impending outrage, I better qualify this…)

    Many HCI professionals I know and have know are excellent in contextual requirements analysis, in creating scenarios and testing existing designs or prototypes. However, put them in front of a blank piece of paper and they really have little better ability to do an initial design than a lay-person. They end up using the ‘throw it at the wall and see if it sticks’ methodology discussed by both Nielsen and Cooper. This might create an adequate solution that would be better than solutions produced by people with a less user-focused perspective, but it will never create a great solution.

    I think one strong reason for the fact that many HCI professionals cannot design, is because HCI courses rarely teach design, especially in a practical way. Winograd discusses this eloquently in his book ‘bringing design to software’. He feels that HCI or interaction design needs to be taught in a similar way to other design – i.e. using a ‘workshop’ methods where students do lots of realistic projects which are then subjected to peer/tutor review. This is very different to academic/theoretical learning. It is doing not reading. It is also building up experience and learning to question oneself.

    Sherlock

  • Andrew H Otwell

    April 26, 2002 at 6:54 am

    Agreed, Sherlock. Most HCI folks learn nothing about “good visual design,” or the marketing aspects of design. HCI is essentially a research field, building a continually evolving body of knowledge through incremental, experimentally verified tests. It’s evolutionary, and rarely revolutionary or innovative. Some of that work is by definition opposite to some necessary aspects of Design, for example the perfectly legitimate design requirement of “make it look cool.”

    While the results of that work are useful to designers, that work *is not* design. There shouldn’t be any pretense that HCI practicioners are designers any more than Art Historians are Artists. Designers work for clients, most research HCI does not. Designers must take into account the needs of their clients’ businesses, not just the needs of the client’s users.

  • George Schneiderman

    April 26, 2002 at 11:45 am

    As long as everyone else is picking on HCI professionals, I’m going to pick on design professionals. In my experience, most graphic designers and even many web designers don’t know or care about users or their goals. They see themselves as artists out to create beauty, not as craftspeople out to create useful tools. It seems to me that in many respects HCI provides a much more solid foundation on which to build an interaction design profession than does, say, graphic design. (Which isn’t to say that graphic design doesn’t have important contributions to make.)

    The techniques that I bring to a problem as a software interaction designer are all oriented around providing users with tools to accomplish something—tools which will be easy to learn and remember, and efficient, pleasant, and powerful in use. I have nothing against games or digital art projects, but their design has little to do with my practice as a software interaction designer. My professional practice is all about designing (mostly) digital tools that help people accomplish goals—I call that “software interaction design”. Maybe “interaction design” isn’t the right term for that, and should be defined more broadly as Jonas proposes. But I believe that the term as I use it does define a coherent professional discipline which needs its own label.

    As far as innovation: I may design products to meet goals that users don’t even have yet. But to do so successfully I still need to understand my potential users well enough to envision what goals they might form once we provide them with enabling tools.

    PS I just got back from CHI, the annual meeting of the premier academic/professional conference for HCI people. Interestingly enough, during one of the panel discussions (“CHI at 20″), Don Norman made the assertion that for HCI to move forward as a professional discipline it needs to figure out how to apply itself to product design.

  • Robert Reimann

    April 26, 2002 at 4:06 pm

    Well, to quote myself (since Frank was so kind as to include a link to my article in his response), I proposed that interaction design is the design of the BEHAVIOR of artifacts, systems, and environments (and is thus also concerned with how form supports that behavior). The fact that interaction design focuses on behavior requires it to use a new set of methods and practices, because traditional design disciplines and methods address concerns of form and content, but not those of behavior. Interaction design is considered most often in the context of digital systems because such systems generally exhibit behaviors complex enough to be worthy of design (a toothbrush has little behavior that isn’t evident by inspecting it; inspecting the physical shell of a computer tells you nothing of its very sophisticated behaviors).

    Note that “behavior” does not necessarily imply support of goals in the narrow sense of “enhanced productivity” or the like. Having fun is also a human goal, and a game’s behavior is what creates a sense of fun. Game designers don’t often call themselves interaction designers, but I would propose that this is precisely what they are doing in their specific domain. While I think there are many distinct principles and patterns that come into play in the design of interactive entertainment vs. productivity applications, I believe that the broad *process* of designing behaviors could be generally applicable even to the interaction design of games.

    We need a broad a definition for interaction design, because interactive systems do not stop at tools and productivity applications, but extend to entertainment, community-building, and potentially outside the digital arena altogether (design of human-organization interaction for example). Digital systems are the most prevalent, but certainly not the only entities that could benefit from the process and results of interaction design.

    Robert M. Reimann
    Director of Design R&D
    Cooper | Humanizing Technology

  • Roberto Bolullo

    April 27, 2002 at 3:42 am

    I think we have to make a closer look just to the name itself to have a deeper understanding of the discipline.

    Interaction: interaction is a process of continual action and reaction between two parties (whether living or machines). It is debatable whether or not a computer is capable of actually initiating action rather than merely reacting through its programming. (Nathan Shedroff (2001) “Experience Design” , New readears)

    Design: The design etymology goes back to the Latin of + signare and means making something, distinguishing it by a sign, giving it significance, designating its relation to other things, owners, users, or gods. Based on their original meaning, one could say: “I design is making sense (of the things)”. (Krippendorff, K (1989) “On essential contexts of artifacts or on the proposition that “Design is making sense (of things)” Design Issues, 5(2), pp. 9-39)

    Interaction design: we could extract then that an Interaction design is dedicate to define partier’s behaviour with “sense”. Then we could say that thru the understanding of the context, the designer defines the behaviour of products, services and environments.

    And with this definitions we can talk a lot about interaction design and its practice: new methods and practices to gain specific needs, integration of all features that affects the behaviour from the business, user, social and production-consumption perspective, design discipline, etc…
    For me the most important issue for Interaction design is to try to separate definitely form the idea that is a digital issue because if not we are just envisioning the 10% of its full potential.

    I wrote an essay in my school years (not long ago) about what is interaction design and give some examples for inspiration. You can see at:
    bolullo.com/roberto/projects/id.pdf

    Roberto Bolullo
    bolullo.com/roberto

  • sherlock_yoda

    April 30, 2002 at 12:57 am

    I just wanted to clarify one of my earlier points – especially with regard to George’s comments.

    When I was talking about design, I wasn’t referring to graphic design, artistic design, visual design or anything creative or ‘wacky’. I was talking about product design – i.e. the practice of solving problems within a particular context with particular constraints. The closest analogies being either industrial design or perhaps architecture.

    George said,
    My professional practice is all about designing (mostly) digital tools that help people accomplish goals—I call that “software interaction design”
    This is what I was also referring to.

    My criticism was that many people from standard academic HCI backgrounds don’t have much practice in this form of design. They may know some theory and they may know some requirements and testing methodologies, but they don’t have much practice in going through the complete design cycle. This is what traditional design courses are good at (be that visual design, graphic design, industrial design) – they ensure that all graduates leaving their courses have designed 100s of small design projects, which are then subjected to peer/tutor review. In a nutshell, I think HCI or interaction design education should be at least half practical (intense practical at that), as well as academic. Winograd in ‘Bringing Design to Software’ makes these points much more eloquently than I do.

    I am not pointing the finger at HCI professionals. I am pointing the finger at our current education practices. We do not teach this stuff properly. Ideally, there would also be some elements of apprenticeship/mentoring for developing interaction designers – this is currently very hard to find.

    As The King said, “A little less conversation, a little more action..” :0)

    Sherlock

  • sherlock_yoda

    April 30, 2002 at 7:01 am

    I have recently been reading Norman’s book ‘The Invisible Computer’, where he discusses the idea of gadgets or computer technology becoming seen as products (like TVs, radios, fridges, etc). The computer part will be invisible – we will stop thinking of things in terms of technology, computers, operating systems, etc.

    Perhaps this is the goal of interaction design – to stop people worrying about interfaces/computers and to see things as products – simple devices/tools that help us get things done. Maybe we should call ourselves ‘product designers’. This term also hints at a broad notion of design including the design of visuals, interactions, functional elements and maybe even manuals, boxes, marketing etc. I think this may be a better term than ‘user experience designer/director’…this is too ‘techy’ or ‘.com’ a term.

    I think MicroSoft went this way a long time ago. I think they have Product Managers who have overall responsibility for the direction/quality of a product – the overall user experience.

    Just some thoughts popping into my head…

    Sherlock

Sorry, comments are closed.