The Sketching in Hardware Summit is taking place this July in Berlin, Germany. Though we’re not able to attend this year, we wanted to share some conversations from past Sketching in Hardware events through a series of interviews conducted last summer.
The first is with the legendary Bill Verplank. Bill is well known in interaction design circles for his pioneering work on the design of the original graphical user interface and mouse, research which has influenced all the devices which we use today. Below is a transcript of our amazing conversation, which included a description of his groundbreaking work on the Xerox Star project, the importance of real physical interactions, why forest-fighting robots are bad and how to build Italian hill towns in outer space.
You’ve had a rich career that’s inspired many students and professional designs around the globe. Was there a favorite phase, or stage or project in your career?
I think the project that changed my life was Xerox Star. It was already going when I joined it, so I don’t have ownership of it, except that I was engaged in helping to decide whether there should be one, two or three buttons on the mouse, and what the layout of the keyboard should be, and what the refresh of the phosphor should be, and what the phosphor should be so that you get a more precision phosphor. So for three years we did a series of important tests that engaged me in debates around whether modelessness is more important than consistency. Is visibility more important than modelessness? So if you can’t see that you’re in a mode, it’s a hidden mode. That’s the worst kind, you see, it’s a double negative, but I think we agreed that visibility was more important than modelessness. Larry Tessler argued before he left to go to Apple that modelessness was so important. That’s why there is only one button on the apple mouse because Tessler decided that. He and I are contemporaries from Stanford. We were undergraduates together.
I think that was very formative because it was my first real commercial job and I had been an academic until then and I realized that I knew some things that were useful to industry. We were breaking new ground and doing testing that was based on some mathematical models. There was a history of doing mathematical models at Xerox PARC where Stu Card was proving that the mouse was better than the joystick, for example, and he had a mathematical model, Fitt’s law. So he would say this compares to human performance, look at the mouse, it’s a Fitt’s law device. And if you use a pencil you can get Fitt’s law performance, and you don’t quite get as fast a performance with a mouse as you do with a pencil but the trade off is the same, it’s a logarithmic relationship between difficulty and time, a nice straight line. So there was this academic side that Xerox PARC established, and so all the engineers were scientists and they really wanted proof that their design was the best and we had big arguments because we did the tests and some of them turned their backs on what we showed. We said, “watch this woman” who was a secretary and she was trying to explain how you select things. She said, “Well I click down here and then I move the mouse with the button down and I get to the end of the line and then I go back to the beginning of the next line…” You see, her conceptual model was a highlighter, so was thinking about highlighting the text, we even used the word “highlight”. She knew what a highlighter was, but had never seen anything highlight on the screen, and if you carefully draw a line across a bunch of text then it looks like you’re drawing a yellow highlight. In fact, it wasn’t yellow, it was inverted; all we had were lack and white pixels. And then she’d come down to the next line and she’d go back to the beginning of the line and she’d undo what she’d done before, but the cursor would jump down, and she was going from line to line. She didn’t know that you could go straight down without having to go from line to line. It was clear what her conceptual model was just by watching what she was doing. And I remember one of the designers (we had designers who would come up with these ideas) kind of turned his back on the video and wanted to engage me in conversation and I said, “Look, you’ll see how she thinks it works. Just listen. Watch what she does.” So our job was to make them pay attention to the people we tried things out on.
We were in L.A. and they were in Palo Alto. They thought they were smarter than we were. They had the ideas, but we had the proof from the user advocates, we call them users. We did three years of really thorough tests and we divided responsibilities Terry Roberts did the selection scheme tests. We had a psychologist running the group. He was a PhD, and she was a PhD from Stanford at PARC And I was a PhD from the MIT Man-Machine Systems Lab and had all this experience in information theory models and decision theory models and control theory models of human performance. And we sort of made it up on the fly as to what kind of tests we would do. I did a whole series of tests on what the icons should look like. I had four designers. It was not a scientific test. We were trying to discover principles, but we wanted to compare them in some rigorous ways, so we had four designers complete tests. It was one of the first things we published in ’83, two years after the Xerox Star came out. It was the foundations of the ACM SigCHI conference. So that was my career at Xerox PARC for 8 years (78-86), including 3 years of testing. The last 4 years I was continuing to refine the design. I was much more of a designer and we were looking at tablets and all other kinds of devices.
Bill Moggridge approached me. He had taken on the contract to do industrial design for the follow up to the Xerox Star, so he and Robin Chu, his principal, the best designer he had, did a beautiful redesign of this really conventional beige box. It was a dramatic color, not black and white, not red and yellow, but it was an architectural rendering of a column of an office building if you looked at it that way. And then he designed the keyboard and a monitor display (I think they were separate), but he then sold Xerox on a strategic vision of the future exercise, and that’s when he hired me. He started that before he hired me and I remember watching it go on and I realized that they didn’t know what they were talking about. And it was run by the head of industrial design for Xerox who was famous for these future studies and every year he would produce these visions for the next five years of Xerox products in mockup. So they could go into that room and say, this is the big printer that’s going to be in every copy center, and this is the personal copier that’s going to be in offices. It think they hadn’t quite yet projected copiers that could fit on tops of tables, but printers at that time were big huge boxes.
Anyway, that was my first exposure to industry, my first exposure to doing human factors testing, and in a sense defining a new field which was centered around human factors studies of human computer interfaces.
When I went to work for Moggridge, I knew we weren’t just going to design computers, but we would design all kinds of things that had interfaces, so we called it “Interaction Design”. I think that’s a good term. Unfortunately most of the people doing that kind of work are graphic designers and it’s mostly graphic designs that are dominating interfaces now–how do you design web pages.
So much of business today is focused on screen-based interfaces, and people like myself and the other folks at the Sketching in Hardware conference are passionate about the intersection of the physical and the digital. So much of business is in screen-based interfaces. Do you think the business of physical interaction design is keeping up with the screen-based design?
There are a lot of people making careers of doing interaction design. In a talk I did a couple of years ago for IxDA, I started with a film that my mother took when I was eight and in it I’m trying to saw. And my dad comes along behind me and he says, “Don’t do it like this” and he took my hand and over the back of my shoulder he said “do it this way” a beautiful straight stroke. Which is hard. It’s easier to do it the way your arm wants to do it. And then he said, well before you saw, you need to mark it, and he pulls out a big metal square and marks the cut, and puts the pencil just like I do in my pocket after you mark the cut. It’s actually how I work with my saw to take my disk drives apart. I even mark it with a straightedge and a pencil. It was a wooden saw, a small one he had for me, I still have that one, too. And I still had the bench until about last year when it fell apart. It was 60 years old! We used it all the time. I have new blades on the hacksaw. We didn’t use the little saw enough. And I learned to sharpen saw blades. I learned a lot from my dad, and I think that interest in the craft of doing things with your hands was very important to me, so what I talked about was craft. Among the graphic designers what I said that you not only bring art and design, but you bring craft. And I had a nice quote from a calligrapher named Hella Basu. It was from a greeting card made in Asheville, SC. In all object making there are three things that are important. One is art, which is about the meaning of something. Not weather or not it’s artistic, but what does it mean. And then there’s design which is about the purpose of the thing, to what human purpose is it built, why did you make it. So art and design go together. There’s the meaning and the purpose, for example to impress someone, or to hold in your hand. And finally there’s the third thing which is craft, and that’s about the quality of the production of it, whether you made it by hand or made it by a machine, there’s craft in every kind of object-making. And art and design are part of it, but then there’ the doing of it and I think that’s what the people in the audience responded to, that I’m a mechanical craftsman and I appreciate the mechanical quality of things. That’s why I was attracted to tangible user interfaces and still am.
So that was important to me based on the way I was raised, as the son of a craftsman… who went to Stanford and became a manager. He was a carpenter in his dad’s business. They were building homes, back in the 40′s and 50′s.
I really enjoyed the Plank project that you showed. [Plank is an ongoing research project and workshop activity around haptic feedback.] It’s an interesting example of exploring the nuances of physical interaction design beyond the off-the-shelf buttons and dials that have dominated design by necessity up until recently. Those nuances however can be hard to communicate. What you’ve built might look like an ordinary joystick in a video. Do you have any advice for today’s designers regarding how they can explore those nuances?
I think you need dramatic demonstrations. You need to have the before and after. If you can turn the force off and show that you can’t control it like you can with the force feedback on. If there’s something you can hear, or see, some kind of a line you can draw with it, some kind of a start-stop. You need these dramatic demonstrations to show that there’s a rhythm you can get because you’re hitting something. It must be really frustrating for a conductor. He has feedback because he can hear the music, but he has to plan way ahead. He’ll say, okay, we’re not ready start yet, but just watch me. There’s a lot of anticipation and he moves his hands in just the right way, and then, boom. Now at least a quarter second later you begin to hear the music coming out of the performers. And if they’re good, they know what the conductor meant by his gesture, but then he’s kind of following the orchestra rather than leading it. Once he’s given them the downbeat… I’m very impressed by conductors because they have a horrible instrument to play because of the terrible feedback! And there are a lot of tricks to what they do. if you ever watch Seiji Ozawa, he’s a real dancer, dancing around the orchestra, conveying with his whole body what the story was that he was telling. Others are technicians in getting the exact tempo and amplitude. So I’m really awed by conductors, so I wonder why it is that when you get up to about ten musicians there are very few times in a jazz orchestra that someone needs to stan up to say go. Or you just point over there and they know that when they get up to the twelfth bar it’s time for the horn solo, so the horn soloist might stand up, so that leading goes from one instrument to another. I’m quite interested in how jazz orchestras work and how they listen to each other,and they riff off one another. My son is a jazz musician.
There’s a group here called the Stanford Laptop Orchestra, SLOrck. Princeton was the first place with a laptop orchestra and that was called PLOrk, and Ge Wang whose office is right here came from Princeton where he studied with Perry Cook who came from Stanford so it goes back and forth. And he’s now famous, not just for the laptop orchestra which he brought from Princeton, but with Chuck, a language he wrote for his PhD, and he went from the laptop orchestra to the iPhone orchestra and now the iPad orchestra, so there’s been a nice succession. And he’s done various ways to make speakers so you can hear the sound come from each of the musicians with a dome speaker. It’s quite a development. He has started a company called Smule. With his Ocarina instrument you blow into the microphone of the iPhone and convert that into loudness. So for the last three years he’s been hiring the best students and teaching them here. He’s built quite a business. Perry Cook retired from Princeton and is on the board at Smule and lives in Oregon.
So my last question: You’ve been very involved in visions of the future throughout your career. In thinking about those visions, have things panned out the way you envisioned?
In 1975 I was finishing up my first four years teaching at Stanford. I hadn’t quite finished my PhD, so I was going back to MIT, but that summer I ran a study for the design division at Stanford and NASA, so for five years they had done a summer study for the American Society of Engineering Education and it was about how to teach systems design. The way we taught design here at Stanford was by doing design, so the professors would come and they would all participate in a group design exercise. Three years before I did my study, they had done a study about how to detect extra-terrestrial intelligence, and they designed an antenna array called Cyclops, and that array was eventually built. It’s out in central California where they don’t have too much smog. It’s run by the SETI project. That was a success, and I ran two summer studies. The first one was on how to fight forest fires in California, and that ended up as a fight between the roboticists who wanted to build big tele-operated robots that would fight the fires and put them down with big cannons of water and flying robots–they were very technological–and the systems analysts who said hey, we don’t need big machines to fight forest fires, because we can prevent forest fires… we just need to let them burn, we need to do forest management. They did a whole cost-benefit analysis and determined that it wasn’t worthwhile to spend money on firefighting, that instead we should spend money on fire prevention. That doesn’t mean putting them out. You’ll hear stories this year about fires getting out of hand because for 15 years they haven’t been cutting down the undergrowth, and the forests were 20 years old since they’d had the last fire, and, boom!, he right conditions set them off. So that was very interesting because I had set up a while speaker series and got all kinds of people in. Even the students who were here were involved. We set it up so that two or three people were on call, so that if there was a big forest fire happening, we’d get in a car and drive with the US Forest Service or the CDF (California Division of Forestry), the principal firefighters here in California. That was a really exciting project because I got to know about something that I didn’t know before, and I got experience running a summer study. But we ended with this debate that I couldn’t solve, so we issued two volumes: Volume 2 was about fighting fires, and Volume 1 was about how to prevent them, and a cost-benefit analysis. And they didn’t see eye-to-eye, but managing that was part of the learning to me.
The big summer study was space colonies. It’s a long story! At the end of the summer we proposed what was known as the Stanford Torus. It came out of 1975 Space Colony Study at NASA Ames Research Center. It was a beautiful big book, because NASA put a lot of money into producing beautiful illustrations, and we had two of the professors editing the final book, and so their names are on it and my name is on it in the back. There was a Princeton researcher who was publishing popular science articles saying that in the future we were going to live in great cylinders that were 16 miles long and two miles in diameter so they would have a 6-mile circumference, so that’s a big cylinder, if you an imagine it. So up there in the sky are clouds, and atmosphere, and way off in the distance you could see people walking upside down on the far wall. They were amazing illustrations. And one of the questions was how fast would it rotate? What rotation rate at one mile radius do you need to get one gravity? And it’s really slow. It’s about one rotation per 10 minutes, but it’s enough to stick you to the ground. And he had these visions of how wonderful it would be would be like to live that way, and he would arrange them in pairs, because he wanted them to be aiming at the sun so that they would collect sunlight, so they would have to be rotating around the sun. And cylinders want to continue rotating in the same direction they started, and you need big torque to get them rotating in the opposite direction; cylinders want to continue rotating, but if you have two cylinders rotating in counter to one another and you connect them together then they will be generating very little force on one another, and have a zero net rotation, so you could set them rotating at one revolution per year around the sun. Big mathematics! But we decided that our first colonies would be for 10,000 people, and we calculated how many square feet we needed, and he agreed that the density of an Italian hill town would be enough, so we calculated around that, and we designed and Italian hill town that didn’t rotate people at more than 1 RPM. It’s published in a report and they’ve continued to study it. Every summer they run a Stanford torus design competition.
Bill Verplank is currently a visiting scholar at Stanford’s CCRMA and is involved with Stanford’s D-School. You learn more about his thinking in his interview in Bill Moggridge’s Designing Interactions project online.