Wow, I couldn’t be more honored. I’m really, really glad to be here. I want to thank Rick for that very flattering introduction. I’d also like to thank Maria, Luis, and Rick for inviting me here.
I want to talk about why I believe models are crucial in designing and in research. I want to begin with three embarrassing admissions.
First: Design is stuck. And by that, I mean we don’t know how to make progress as designers. As an example of that I want talk about the AIGA National conference in Boston. The first conference was in 1985. AIGA is the American Institute of Graphic Arts. It’s the main professional organization for graphic designers. Wonderful conference. Milton Glazer came and spoke. Brilliant graphic designer; gave a wonderful talk; showed some really great work. Nicholas Negroponte was also invited, and he came and talked about the work of the Architecture Machine Group and the just forming Media Lab.Twenty years later, the AIGA national conference was in Boston again, and Milton Glazer came and gave another wonderful talk; showed some really beautiful work; and talked about being deeply involved creating change in the world. Nicholas Negroponte came again, and he talked about the OLPC,the one laptop per child initiative, and he showed some really great work, too.
But I came away from that talk surprised by how little design had changed in those twenty years and by how much technology had changed in that time. Perhaps, the main change in design was its adoption of technology.
I think we can do better.
Most disciplines have well-established structures for building and sharing knowledge. Here’s a structure where we have people doing research. The research is reviewed by peers. It goes into journals. There’s a feedback loop, which helps people get tenure, and there’s also a feedback loop, which helps get sponsorship. This is a rough model of something much more complex that many of you are familiar with. Many of you participate in this perhaps.
Design doesn’t have this. What we have is a tradition, which goes back before the academy; it predates the academy. It is a medieval practice still. We have apprentices; we give them critiques. Very occasionally some of us write papers and present our findings, and they go off into the ether or the Ethernet.
Almost twenty years after awarding the first PhD in design in United States, we still have not agreed on what design research is. What design research is is still the subject of conferences.
We don’t agree on what design knowledge is either. In fact, many people question the very notion that there is such a thing as design knowledge.
But if being stuck weren’t bad enough, design is stuck in a bad place. This is the second embarrassing admission: We don’t know how to make successful products. John Cain says we know how to make successful products, we just don’t know how to make them regularly.
What I mean is that at a deep level, we don’t know how to do this. Even Apple and Steve Jobs are not always successful. Anyone remember the Apple III, Lisa, Newton, the Apple QuickTake? That was the first consumer digital camera ever released. I worked on these last two; so I know what I’m talking about when I talk about failure. E-world, anyone remember that? It used the technology on which AOL was based. Apple had it; sold it to AOL; and licensed it back. There’s a business model. [laughter] Steve Jobs has made a series of cubes; he has a thing about them; and he doesn’t give up. [laughter]
Product Management, the art of making successful products, is rarely taught in design schools or business schools. It’s a little bit like a fly ball dropping between the center-fielder and the right-fielder.
The people who make products don’t agree on how to do it. Who manages the schedule and the budget? How do you determine requirements? Do you need requirements? Do you start with requirements? Do you start somewhere else perhaps? Who owns design? Who owns the spec? Who can say, ‘No’? Who can say, ‘Yes’?
Often, the official process isn’t what actually happens. The product requirements document (the PRD) is barely begun, but the engineers already have a prototype. That’s called being agile.
Agile processes work well in small start-ups, building products for people like themselves. For example, 37 Signal’s Base Camp. Great stuff. The 37 Signals folks are big advocates of agile development. Curious story. You know who the major investor in 37 Signals is? Jeff Bezos, CEO of Amazon. You know how Amazon develops product? A different process; far from agile. This seemingly contradictory behavior—investing in the advocates of agile but using rigorous planning—points to confusion not just at Amazon but across the industry.
What’s not clear is how to achieve coherence and scale—how to build platforms or interlocking systems—without rigorous planning. This is a religious debate: How to build large, complex systems and be more agile.
We must also consider the research side of the product development process, using research to help understand people and their contexts. Design schools and consulting firms promote this idea. A few forward thinking corporations—many of you work for some of these companies—are promoting these kinds of best practices. And yet upfront research remains rare for most new products.
The value of design research is in doubt. How many of you who have seen Don Norman’s article in Interactions?
“Design research is great when it comes to improving existing product categories, but essentially useless when it comes to breakthroughs… Major innovations come from technologists who have little understanding of all this research stuff.” And that’s what your friends are saying. [laughter]
Skeptics often cite Apple as making great products seemingly without formal res el. This tacit to tacit movement is the apprentice watching the master and tacitly picking up knowledge. I’m suggesting that we also need to be moving from tacit to explicit; that is, explicitly articulating what we have learned tacitly. The next step is to move from explicit to explicit; that is, to build on what we have learned. And then—and this is perhaps the difficult part—to move from the explicit back to the tacit—to embody the new knowledge we have created.
What was surprising to me, was finding myself helping a graduate student with a paper, reviewing his paper, and he said, “Of course you know about the SECI Model. You’re the model guy. How could you not know about this? And I said no, don’t know about this.” This was a Brazilian business information student who was writing a paper for Touch Point, the journal of the Service Design Network. What surprised me was that these models are isomorphic. They have the same structure. They are essentially the same model.
Now this is a long way to go for a small point that most designers understand intuitively—designing is learning.
And yet, there is something profound in that point. We can see a structural similarity. We can make it explicit. We can tell a story about it. And that is more powerful than simply saying, “Well of course design is learning.” We can, in telling the story, say how designing is learning.
I want to finish by talking about models as a form of boundary object. Models are artifacts that bridge the gap between disciplines. I’m sure many of you know Susan Star’s paper where she talks about boundary objects. I want to read this, pardon me. “Most scientific work is conducted by extremely diverse groups of actors. Simply put, scientific work is heterogeneous. At the same time science work requires cooperation—to create common understandings, to ensure reliability across domains, and to gather information, which retains its integrity across time, space, and local contingencies.”
…boundary objects are produced when sponsors, theorist, and amateurs collaborate to produce representations of nature. Among these objects are specimens, field notes, museums, and maps of particular territories. Their boundary nature is reflected by the fact that they are simultaneously concrete and abstract, specific and general, conventionalized and customized.
Scientists have made headway in standardizing the interfaces between different worlds … by reaching agreements about methods, different participating worlds establish protocols, which go beyond mere trading across unjoined world boundaries. They begin to devise a common coin, which makes possible new kinds of joint endeavor.
Let me repeat that, “They begin to devise a common coin, which makes possible new kinds of joint endeavor.”
We need new kinds of joint endeavor. We need to build bridges between research and design. We need to build design knowledge. We need to create great products. We need to build systems and services. We need to build more models.
I want to specially thank Shelley Evenson, whose thinking permeates this talk; although I’m responsible for the errors; Michael Gallagher, who did the graphic design; my long-time collaborator Paul Pangaro; John Cain for his thoughtful critique, and of course the inimitable Rick Robinson.
Thank you very much.
QUESTIONS
Moderator: So if we have a question or two for Hugh right now, we can take that or you can consider questions, and after one more short interlude, we’ll go to the break. Question? Can you go to the microphone please?
Speaker: Hi, I’m a PhD student working actually measuring boundary objects and visuals to tap into a discipline, multi discipline.
Hugh Dubberly: This is what I was afraid of, that there would be experts here.
Speaker: No actually I follow your papers. So what I was wondering is how do you use the component of visual literacy, because that is something I run a lot into when I meet with other disciplines, and they just like, yeah you are designers. You make graphic designer stuff; so you understand visuals. We don’t have like,… I know all the cognitive value that models have for example common efficiency, retaining. We are more exposed everyday to tools like visual models, but how would you answer the questions. Are other disciplines ready to work with models in a visual way? Thank you.
Hugh Dubberly: Absolutely. Other disciplines can and do use visual models.
I can offer this thought. A list is not a model. But it’s great to start with lists. When I teach students to make models, I give an assignment called the baseball problem. I ask them to make a model of baseball. We begin by making a list of all the elements of baseball. It’s important to move quickly to understanding the relationships between the items in the list. It’s important to think about how they form a kind of network. That network may have some hierarchies in it. There are multiple possible structures. Representing those is important.
I often tell design students that you will go into the business world, and you will find yourself in a meeting, and you will think of the movie, Ground Hog Day, that I have been in this meeting before, many times, and you should take action. And the action that you should take is to jump to the whiteboard and say, “I hear you saying this.” And then you should draw out the nodes and the links so that people can see them. And you should say, “I think this is probably wrong. But help me debug this.” And by doing that, you will take the conversation from Ground Hog Day to making something that we can see and discuss and iterate. You’ve captured the idea; it is no longer fleeting.
So that’s my advice.
Break?
Moderator: Yes.
Hugh Dubberly: Thank you.
Hugh Dubberly is a partner in Dubberly Design Office (DDO), a San Francisco–based consultancy that focuses on making hardware, software, and services easier to use, more effective, and more fun, through interaction design and information design. At Apple Computer in the late 1980s and early 1990s, Dubberly managed cross-functional design teams and later managed graphic design and corporate identity for the entire company. While at Apple, he co-created a series of technology-forecast films beginning with Knowledge Navigator that presaged the appearance of the Internet and interaction via mobile devices. At Netscape, he became vice president of design and managed groups responsible for the design, engineering, and production of Netscape’s web portal. In 2000, he co-founded DDO. Dubberly also served at Art Center College of Design in Pasadena as the founding chair of the Computer Graphics Department. He has taught courses in the Graphic Design Department at California State University—San Jose, the Design Department at Carnegie-Mellon University, the Institute of Design at IIT, and the Computer Science Department at Stanford University. He edits a column “On Modeling” for Association of Computing Machinery’s journal, Interactions.
Ethnographic Praxis in Industry Conference Proceedings 2011, pp. 2–11. © American Anthropological Association, some rights reserved.