Interview with John Carroll

John Carroll is another pioneer who ‘accidentally’ ended up with two degrees (because he took so many courses to satisfy his diverse interests) and whose career was majorly influenced by having a year to kill. That one years…spent at Watson Research Center…turned into 18. How many other people, who describe themselves as having the ‘soul of a psychologist,’ can tell you stories about the very first times the ‘user’ became an interesting topic of conversation at IBM? How many have participated in the development of the HCI field from inside a major corporation and academia and the birth of a professional association? How many have worked with Noam Chomsky only to decide they didn’t want to be a linguist? Well, at least one. And today, he’s interested in all of the interesting challenges of teaching, which is awfully lucky for the students at Penn State. Confused about what you want to do with your life? Don’t worry…according to John, it will all look simple once you get there.

EXCERPTS FROM THE INTERVIEW

I actually don’t like the term ‘user’ very much at this point. I’ve gotten tired of it. It is homophonous with a “drug user”. And it seems to make people passive and one-dimensional. I tend to try to pinch myself and replace it with person or something more interesting… But oddly, the term “user” with all its flaws, was very important early on. It conveyed to the technology people that the arcane software they produced eventually had to be USED. It was a focus for the emerging professional identity of CHI people.

There’s a rough side to interdisciplinary – maybe you already know this – everybody disagrees about everything – they all came from a different place, they’re trained in a different way; they have different values and so forth, so it’s not like it’s living in heaven. But it is very interesting. It’s very rich. And it’s not so narrow and alien the way it would be, I believe, if I had gone into that engineering college with my colleagues.

AN INTERVIEW WITH JOHN CARROLL

Conducted by Tamara Adlin on October 3, 2008 08:56 AM

John reminded me that “everything you do changes the rest of your life.” And in his case, it can also help build a whole new field.

Tamara Adlin:  Hello, everybody. Today I am very excited to be talking to John M. Carroll. He was a founder of Human Computer Interaction, which is the youngest of the nine core areas of computer science identified by the Association for Computing Machinery.

He served on the Program Committee of the 1982 National Bureau of Standards Conference on the Human Factors of Computing Systems. That was particularly interesting because it, in effect, inaugurated the field and was the direct predecessor of the field’s flagship conference series, the ACM CHI Conferences.

For the past two decades, John Carroll has been a leader in the development in the field of human computer interaction. He has published several books, including Designing Interaction: Psychology at the Human-Computer Interface and Making Use: Scenario-Based Design of Human-Computer Interactions .

Today he is the Edward M. Frymoyer Professor of Information Sciences Technology at Penn State.

Hello!

John Carroll:  Hello.

TA: As you know, I am collecting the stories of the folks who were instrumental in this field before the field was even a field.; I always start by asking about the first thing you can remember that really fascinated you.

JC: I think that I always had two interests in was trying to accommodate, and I guess it makes more sense looking back than it did at the time, trying to look forward. I was always interested in human sciences – psychology, sociology, humanities, the arts, philosophy, but also in math, that interest later came to include computing when computing got invented.

Even as an undergraduate, I struggled with these two interests; majoring in a lot of different things. I ended up as a math major but really just as a compromise among many interests. Then I went to graduate school in psychology and linguistics, so I swung back the other way.

I think this tension was resolved for me eventually by getting into human computer interaction. But it was largely accidental. I had been a graduate student at Columbia, and I got my PhD in Psychology. But as I was explaining, I always liked math too so in psychology, I specialized in the Psychology of Language – at that time it was called Psycholinguistics. This meant I took a lot of linguistics courses, worked on transformational-generative grammars, and so forth.

Anyway, when I finished graduate school, I interviewed extensively in psychology and linguistics departments but I didn’t find anything suitable. I went to the IBM Watson Research Center for Post Doc year. My girlfriend at the time was one year behind me at Columbia. I thought, well, I’ve got a year to kill. I’ll just take the Post Doc that will pay me the most, hang out for a year, and have a better interview season next year.

TA:  One second; before we continue, I just want to ask you a couple of things about the earlier days. You said you were thinking about a lot of different things in college.

It is very interesting to me that one person could be so interested in both math and psychology. What were the other things you were considering too?

JC: I started college as a major in electrical engineering. I then got very interested in philosophy and English literature. I have always loved writing, and analyzing text. Then, I got interested in psychology and information science, which at that time at Lehigh University, basically meant information retrieval.

So in fact, I wound up with two bachelor’s degrees; one in Information Science and one in Mathematics. One of the things that can happen when you don’t know what you want to major in, and you just take a lot of courses, is that you end up with enough courses to get two degrees.

When I got the Information Science degree, no one seemed to know what it meant. But ironically, the department I am in at Penn State now is called Information Sciences and Technology.

TA: Your parents must have had an interesting time when you came home for holidays. I imagine every time you came home, you had a different set of interests.

JC:  I must say, they were a little tired of it. They were happy when I went to graduate school. They thought I had committed to something.

I also should say, I’m really not a flakey person; in fact, I would say despite the fact that I ended up living my professional life with software engineers, designers, and computer scientists, I always have conceptualized things as a psychologist.

There’s something about the apprenticeship of graduate school, maybe – or maybe psychology just really was the right identity for me – but I feel I have the soul of a psychologist; that’s really what I am.

TA:  Yes; and flakey is not a word I’d use. Somebody who comes out of college taking lots of classes with two bachelor’s, is not flakey. The person in college having taken all those courses with NO bachelors degree -

JC: (laughs) OK that’s flakey. I guess I was indecisive, anyway.

TA: Yes. So this was in 1972 when you got your Bachelor’s.

JC: Yes.

TA:  Then you went on and finished your PhD in 1976.

JC:  That’s right. It was the Spring of ’76 when I was doing a lot of academic job interviews, and finally decided to stay in the New York area; partly for personal reasons, and partly to give the interviewing thing another whirl the following year.

As everybody who has pursued an academic career knows, there are so many universities, but for any given PhD student, any given scholar or scientist, there is a far smaller number that are appropriate, and might actually be looking for a faculty member with your particular skills and interests.

So if you wait a year, you really get a chance at a whole different set of job opportunities. It wasn’t an irrational thing to do, and it also made personal sense because I would have had to settle for less, and also move away from my girlfriend.

I went to the Watson Research Center, which was an absolutely excellent place in linguistics, and also had a very good research staff in cognitive psychology. However, going directly from being a psycholinguist at Columbia to an industrial lab in 1976 was not the normal career path. It surprised my Columbia colleagues. But it really was not that shocking. There were lots of wonderful colleagues at Watson from the start. It was actually quite awesome to ponder the quality of the scientists that worked there. They certainly covered all of my interests.

You have my bio, so you know I stayed there for 18 years, not just one year, as originally planned.

TA: But it originally started out as being a year to wait for other potential academic positions?

JC: Yes. I went there to stay for just a year, to wait for my girlfriend. But a year later, I wasn’t in that relationship, and I switched from a Post Doc to a research staff member. Everything you do changes the rest of your life.

TA:  It’s amazing; some of the folks I am talking to – there are personal stories locked in there – like an ex girlfriend that really influenced what ended up being some influential career choices.

JC: Right. But of course in everyone’s life there are things like that. I’m feeling this keenly as a parent. My daughter is a freshman at Penn State. We just had a long conversation last night about her career  plans. I was actually remembering some of this same stuff with her.

She knows this story, of course, and has heard it too many times, but she heard it again last night. She’s asking herself what should she major in, and I said, well, of course, you’d like to pick the thing that 20 years from now will be the “right” choice, but how can you do that. There’s no way we can ever do that. Too many unpredictable things will happen.

TA: So what is your advice to her?

JC: My advice to her was to follow her heart; to major in what she’s passionate about and interested in, because I think she’ll put more time into that, and it’ll pay off better for her.

TA:  Yes; and it certainly helped you. You wouldn’t have thought, like you said, at the very beginning, looking at your academic career, that it would have ended up making a straight line; mathematics, information sciences and psychology.

JC: No, but things have a way of working out. They don’t work out for everybody. And you can’t just stand around, and expect it will work out, but I think people can try too hard to steer, when in fact we’re only in control of our fate to some extent.

TA:  I also like what you said to your daughter, which is you’ll end up working hard in something that you really like, which in and of itself is putting you in a better position because you’ll do well.

JC: Yes. I think that is good advice. She’s a good kid; she’ll do well. I think she had actually thought all this through on her own before she talked to me; she’s just trying to humor me. She knows I like to give advice.

TA:  Well, my dad actually helped me too when I was struggling because I was in a weird place and didn’t know what career it would end up with – he said the same thing to me and it worked out for me too -

JC: Good advice!

TA:  So you find yourself at Watson the first year. You said you really liked it; there were some really smart people there doing some interesting stuff.

What did you do and love in the early years you were there?

JC: I got to work with some really gifted psychologists who were there; people from top graduate programs who, for their own quirky reasons, had also taken this dive from the academic-default trajectory into an industrial lab, but really, an outstanding staff: John GouldJohn ThomasClayton LewisLance Miller, Don Nix, Steven Boies; first rate scientists.

Plus, on the linguistics side, Watson had some outstanding people too; David Johnson, Paul Postal, Warren PlathStan PetrickJohn Sowa – so between the two of these groups, I had a wonderful collegial support.

I spent the first years of my IBM career working on psycholinguistics, on the one hand – I studied referring expressions, and, on the other hand, studying software design as a kind of boundary case of problem solving; what everybody now calls ill-structured problem solving or wicked problem solving.

Those are both really interesting questions; academically, scientifically, in their own right, and very sustaining things to investigate. It really helped me grow a lot, I think, as a professional and as a scientist.

Nevertheless, I wasn’t completely grown up and committed to what was becoming HCI yet. Because when I had the chance in 1980 to go to MIT as a kind of a senior Post Doc, and to work for a year with Noam Chomsky, I did it.

TA: Oh, boy.

JC: I did it because, even though I liked my life at Watson, I liked what I was doing, I still felt a tension. I liked the applied psychology work we were doing on software design as problem solving, and other work I was doing on how people refer to things, and make up names, but I felt it wasn’t really linguistics or cognitive science. I was not sure I wanted to become an applied psychologist

TA: Let me ask you; why did Watson want a bunch of psychologists? Why did they want a bunch of linguists at that time? That sounds awfully pure research to me.

JC:  Yes; it’s an interesting thing. I think you have to understand this historically.

In the 1970s, industrial research institutions were fundamentally different than they are now. In the context of the Cold War and Sputnik, and American global economic dominance, a far greater proportion of GNP was being directed to basic science than is the case now. I won’t be so indelicate as to ask you how old you are – but I’m guessing you are younger than me. 

TA: I am 40.

JC:  You are quite a bit younger than me. The Russian’s Sputnik satellite in the mid 50s scared the hell out of the west. The U.S. in particular, as everybody knows, is very susceptible to fear mongering. And back then, we still had the economic engine to respond. There were labs like Watson, like Bell Labs, and like the original Xerox Park, that were very basic; that were doing fairly basic research.

The companies only expected a certain portion of that to really end up in their product plans, ever. That’s kind of an alien idea today, isn’t it. Today everything is much more ROI.

TA:  It’s interesting. What you just said is that a lot of science got supported out of a certain amount of fear.

JC: Well, during the Cold War, geo-politics was all about fear. And of course it is now too! There’s always a motivation; everything is done for at least one reason. The science boom of the Cold War had a lot of very positive effects.

President Kennedy’s goal of getting to the moon was intended to send a Cold War message to the Russians and the rest of the world. But it also was clearly all about big ambitions, human achievement, and hope. If we think about our own industry – computing, the 1960s were the years when the computer as we know it was being invented. Lots of big things were happening; cognitive psychology and modern linguistics were invented in that decade.

TA:  I suppose the analogue today – I worked for the Department of Defense for a strange little while – and I was really amazed that how much we did in the labs was really farmed out money to places likeGeorgia Tech’s Visualization Lab, and funded a lot of academic research.

But that’s not a company doing it; it’s the government straight to academia.

JC: Right; and that paradigm also was established in these years; the big funding programs that universities like Penn State now depend on. Research universities to a considerable extent are an integral part of a distributed government research institution. Those large applied research programs are essential to both the government and to the universities.

TA:  What an amazing opportunity for scientists at that time.

JC: It was. It was an amazing time.

Anyway, I decided to take a break from Watson, and to go to MIT and refocus on linguistics and cognitive science. Honestly, a big part of the attraction was the chance to study with Chomsky. His impacts on more or less everything I had studied or worked on was huge. He was my PhD advisor’s advisor, so indirectly, for me, he was the source.

I went to MIT for the academic year 1980-1981. It had an unexpected effect on me. I thought and I hoped it would help me decide what to do, where to direct myself,and it did. Working with Chomsky, and being at MIT made me absolutely certain I did not want to be a linguist.

TA:  Really!

JC: Yes; people are always surprised when I put it like that, but think of it this way: if you work with the best person in the world, if you work with the person who you’ve been reading for your whole life up to that point, and worshipping, really, if you then get to work face to face with that person, and you really get to see what you might be if you could succeed as well as humanly possible – that’s a very strong test.

It didn’t interest me as much as I thought it would. Many of the key issues in transformation-generative grammar seemed remote from any empirical consequence.

TA:  Interesting. It totally makes sense.

JC: This is not to slight Chomsky, because in fact he impressed me enormously. He’s still probably the most impressive individual I’ve ever worked with, ever talked to – he’s really amazing. Amazing work ethic, amazingly smart, amazingly quick. He is in a singleton set.

But I did not want to be a linguist after that year. I am still interested in language, still fascinated, but I wanted to do something different as my own science. It was wonderful timing for me, because, immediately before I had left IBM for MIT, we had held a corporate task force, of which I was a member. That task force recommended that IBM should be much more ambitious in the areas of cognitive science, cognitive psychology, and use interface design – we didn’t yet have the term human computer interaction. It set a new strategic direction for IBM Research. When I returned from MIT, the recommendations were being implemented.

These new directions were also directly supported by changes in the business environment and throughout the corporation. In the 1970s, of course, IBM was all about big iron and professional programmers, but as the 80s came, we were getting positioned to become the PC company, and everybody else was too.

This was a very fundamental change for IBM. It changed who was the user and who was the customer. Some of these changes are still being assimilated in our industry. I had been studying programming as a kind of problem solving before I went to MIT, because when we talked about people and computers, we automatically thought of programmers the problems of professional programmers. That was the 70s.

The task force I served on in 1980 wrote a report, as task forces often do, and then followed through by presenting the results and recommendations to IBM executives in the Research Division and elsewhere. By the time I came back from MIT, there clearly had been a sea change at IBM with respect to psychology; different departments, different groups, different managers. We were hiring a lot of people. We were rapidly getting focused on the problems with naïve, novice users, general users, business employees learning and using computers to do regular work, not programming.

TA: In the beginning, the users really were the programmers.

JC: Yes. There’s an abrupt change right around 1980. In fact if you read Shneiderman’s first book, which I believe was published in 1980, it is essentially about the problems of programmers. It was called Software Psychology.

psybook

Figure 1: Shneiderman’s Software Psychology book, published in 1980.

But by 1981, when I returned from MIT, suddenly it was a whole different set of problems. It was computers and kids; computers and secretaries; computers and office principals.

I made my personal decision, and I did not have worry about wallowing in self doubt. There was way too much to do. As you mentioned in your introduction, the spring of 1982 was the Gaithersburg Conference – at the National Bureau of Standards, which is now NIST. That was essentially the zero’th CHI Conference.

TA:  So you remember what you talked about then; how did that turn into CHI?

JC: The key thing about that conference was that it gathered a sufficient critical mass of dispersed people, and turned them into a community. When we arrived at that meeting we were a shapeless group, people who sort of knew one another, had read a paper or seen a citation. It was not yet a community. I remember waiting in line for lunches and things like that, at that conference – I met many of the people that I have known very well for the last 25 years.

TA:  Like who?

JC: Like Thomas Green; he got one of those awards from the CHI Conference recently. He was a leader in Britain. I’d heard his name, seen his paper. He was a little bit like me because he’d been a psycholinguist before he got into studying command languages.

I first met Tom Moran and Stu Card, and Tom Malone at Gaithersburg. I knew Ben Shneiderman from before, but I met a lot of his software psychology associates at Gaithersburg. Wow; it was an exciting three days.

It wasn’t an accident that this event was held in the Washington, D.C., area. That was a hub for the software psychology interest area, which was one of the most important constituencies of the early CHI group. It’s interesting and ironic; you look at CHI today, and there’s not a lot of focus on the psychology of programming. The field has evolved.

It’s evolved toward the end user, towards interactions and the experiences of interactions. The problems of software professionals, or even non professional programmers are not really central now. Gaithersburg helped to start a rescoping of the early research focuses of CHI: cognitive user models – like GOMS, interaction with menus and commands, games as a framework for learning. 

TA: So by now, you did have a little bit of language around this; about the idea of human computer interaction as something to talk about.

JC: Yes; I don’t know where that term came from; human computer interaction. The Card, Moran and Newell book was called The Psychology of Human-Computer Interaction, I think, but that was 1983. Of course, they were working on that book for several years.

TA: I’m not sure. I’m going to be talking to them.

JC: Yes; you can ask them. I can tell you one thing for certain: I did not coin the term human-computer interaction. I read a reference to me, I cannot remember exactly where, but someone attributed the term to me, and that’s wrong.

TA: It’s interesting that CHI became CHI. Computer human Interaction isn’t something you hear a lot but I guess if it had been HCI it would have been the “hickey” conference which would have been a lot odder.

JC: The other way I think about that, is that if you are ACM, you always put the computer first!

TA:  Right; that makes sense.

JC: The other odd thing about CHI as the name for the conference is that the name is really human factors in computing systems; it is not computer human interaction. That’s always been the title, Human factors in computing systems. I guess no one can say HFICS.

Another interesting footnote in the history of CHI is that role of the Human Factors Society and the many human factors people who are interested in computers. Today, we identify professions like usability engineer and interaction designer, even user experience designer. But in the early days, it was human factors engineers, and documentation designers. That human factors constituency was critical in the early days. I was a part of it.

People like Dick Pew, Alphonse ChapanisRobert Williges, and my colleague at Watson, John Gould were very established human factors guys who got interested in this CHI project and gave it a lot of support early on. Like the psychology of programming, the human factors influence in CHI has fallen away over the years. I don’t know if that’s good or bad; I guess I’d be inclined to think it’s bad, but it’s true.

TA: It’s how the definition has changed – or maybe it hasn’t changed.

JC: Yes; I think it has changed. I think that CHI is not really human factors any longer; it’s got its own tradition in terms of evaluation, and it’s got a much stronger interest in design and aesthetics than traditional or even contemporary human factors. So it’s a different community now. Maybe that’s just the way things go; human factors did not become extinct. It remains an important community in the wider spectrum of professions that contribute to making technology that people can use. They have their own conference and their own community.

TA:  I think of Human Factors people as more interested in the more physical piece – like how big fingers are and whether -

JC: Ah, the ergonomics parts; yes.

TA:  But it sounds like then they were thinking a lot about mental processes and interacting intellectually with systems.

JC:  Yes; they were going through a change too. Human factors has its own roots in interactions with physical materials and devices, but as displayed information and touchpad interactions became typical, human factors changed.

I would say the paradigmatic problem in human factors now is as contrasted, say, with CHI, is more that they are more interested in control systems; like air traffic control or semi automated factories or piloting planes and boats – those are the things they seem to focus on or gravitate toward more than, say, digital camera interfaces, composing and posting photos on Flickr, etc.

You do not hear much about control system issues in CHI.

TA:  Maybe that has something to do with inherent physicality in each of those things.

JC: That’s true; there’s a physical operation on one end of a control system.

TA:  Right; and there’s the exhaustion factor which is physical, or digital -

JC: You’re right; maybe there is more continuity there with the roots of human factors which was pretty much motor behavior aspects of work. Yes; maybe that’s right.

TA:  That is really interesting. Computer human interaction people now certainly have a lot of involvement in medical systems, which you could argue the same things, but I personally haven’t heard of folks who are doing things like air traffic control systems or stuff like that. It’s interesting how these many invisible lines start to appear.

JC:  Yes; it is a fascinating thing about professional communities, or maybe about communities in general. They are warm and fuzzy with respect to defining “us”, but that always also designates others as “them” – as left outside. Some of the people I know in human factors are a little bit bristly about CHI because they feel that they are not seen as “us”.

You can’t blame a community for having a finite scope. A professional community would lose coherence if it tried to keep too many things in focus. And CHI has many, many things in its focus! As time goes by, some topics, and the people who study them break away to form their own communities. Human factors slipped out of the focus for CHI, as did the psychology of programming.

In those cases it is just especially ironic because human factors and the psychology of programming were so central and so important to the early CHI community.

TA:  Yes. So now we’re back at IBM and we’re starting to get into the mid-80s now.

JC: I don’t know how far you want to go but after the first years of the 80s, I think those were very chaotic and exciting years where every CHI was not just a conference, but a new manifestation of the community. Things were changing very fast. For example, this thread we were pursuing of different communities playing roles over time was pretty salient during the 80s. It seemed to many of us that we were tuning the original constituencies to create a sustainable and effective balance.

I remember discussing CHI ’86 with Stu Card, and talking about whether we needed to make an adjustment that year through the program committee process to be a little more solicitous to the cognitive people. I don’t think we need to do that as deliberately now. The community has its own momentum and it will go on.

Maybe it will be a little less or more cognitive but I don’t think people think they need to manage the balance. But in those years we did.

TA: Was there a big increase in attendance each year?

JC: It was growing pretty steadily, yes. It has been on a wonderful upward trajectory. There were some years that were slight setbacks. I think the first year they went to Europe in ’93, was a down year in attendance. But it was a big point, I think, to make; that the community was international, and that that conference could go to Europe, and that that was going to be normal.

I feel it was a pretty good growth trajectory.

For me, the next big thing that happened was in 1984 which is when at Watson, we formed the User Interface Institute. This was a big thing professionally for HCI and usability and user interface people in IBM. Up to that time, some of us were human factors people, some of us were user interface programmers but there wasn’t an appreciation that this was its own multifaceted community that spanned both of those things, and that it was an integral part of the IBM professional community.

TA:  We talk about where terms from in our field – it occurs to me that I would really love to know who first uses the term “user.”

JC: Yes; I don’t know that one either. I actually don’t like the term very much at this point. I’ve gotten tired of it. It is homophonous with a “drug user”. And it seems to make people passive and one-dimensional. I tend to try to pinch myself and replace it with person or something more interesting.

TA:  In my presentation, I have a slide that says user is a four-letter word.

JC: But oddly, the term “user” with all its flaws, was very important early on. It conveyed to the technology people that the arcane software they produced eventually had to be USED, It was a focus for the emerging professional identity of CHI people. Forming the User Interface Institute at IBM, provided a huge boost to HCI people of all sorts throughout the corporation.

It also helped the HCI area grow at Watson. In one particular year, I remember we got ten new head count, and we became a second level organization which meant there was a senior manager of user interface research – for the first time ever in IBM.

TA:  Who was that?

JC: It was me.

TA:  It was you! Okay. That was a great position to be in.

So you were in the position, even that early, of trying to find people who would be really good at this. For people who have been in that position before, I like to ask how did you -what kinds of questions did you or would you ask to figure out whether somebody would be good at this.

JC:  I’m not sure my record is that good that my personnel practices should be recorded! I’ve hired some duds; I’ve hired some excellent people. I hiredWendy Kellogg. Maybe she is my greatest personnel achievement in recruiting, but she was quite easy to recruit!

My philosophy is to always hire the most talented people. Good people find a way to succeed even if to do so they have to change the organization they are in. Good people make intractable problems merely challenging, and they make challenging problems interesting.

One of the most interesting pieces of my managerial wisdom I got fromSteven Boies back at Watson. He said that the most powerful determiner of success in research is tolerance for ambiguity. That really stuck with me. I definitely look for that in people I hire.

TA:  That is a great way of putting it.

JC: It does work but it doesn’t always work; probably nothing you can do will always work. We have the same problem now. We’re hiring faculty here at Penn State – I’ve hired some good ones but not always.

TA:  And choosing students too, I suppose.

JC: Yes; same thing. I think it’s always a good thing to go with talent. If you have somebody who’s really bright, they might fight back, you might just not convince them, and maybe if you fail, maybe they’re right. I don’t know. But that’s been my philosophy and I’m too old now to have another one, I think.

TA:  Fair enough, although I don’t know if I would agree with that last statement.

JC: So now you are at this institute which is very exciting and new, and actually getting a lot of political clout at IBM.

TA:  Yes well the next 10 years were some of the best years of my life. I managed that group for less than two years and I stepped down and replaced myself with one of my managers, which is a good thing to do!

So I got to manage, for the last eight years of my IBM career, a very small group of really good people doing basic research in HCI.

TA: Did you do that just because you enjoyed the work more than becoming a administrator of an institute?

JC: Yes. At IBM Research, philosophically, first level manager really is a technical person; almost a team leader. You do personnel but you are supposed to be a technical contributor. That is your identity. But a second level manager is really a mission manager and a personnel manager. If you do technical work, that’s okay as long as it doesn’t jeopardize what you arereally doing.

I identified more with the technical role so I stepped down. I stayed a manager, though, until 1994 when I resigned from IBM.

TA:  What sort of projects were you really interested in doing at that time?

JC: The main areas we worked in during that period were programming environments – the psychology of programming wasn’t gone yet; it was still pretty big then. IBM was facing a crisis in transitioning from the old procedure languages, like PS1 and Fortran, to Smalltalk and C++. We were interested in helping the company cope with that re-training problem of learning object oriented design principles and techniques.

TA:  So this was still really dealing with programmers.

JC: Yes; definitely it was. In that work, we worked primarily with IBM employees.

I was also interested in learning and information design for new users. I did that work more in the context of office systems and more general purpose systems; PC applications and so forth.

TA:  When I talked to Mary Beth, she said that she was doing stuff on experts and you were doing stuff on novices and that was some of your first work together.

JC:  That’s right. One of the first things we did together involved trying to identify a principle that could explain both expert and novice behavior patterns.

Our biggest impediment was getting a personnel accommodation. When she was hired in 1982, she joined a different group from me. We were dating pretty much from the day she was hired, but that was not an issue since we were both non-managers and in distinct groups. But in 1984, when I became the manager of the whole user interface area, she in effect reported to her husband, because we had been married in ’83. So we established a sort of non-linear reporting structure where she went directly to my manager for personnel issues. I guess this kind of thing is more common now with so many couples meeting at work, and working together.

TA: Which is only fair.

JC: Of course; you have to do it that way but it was awkward. I think it worked out okay. We tried initially to have nothing to do with one another’s work. We thought that would be better for both of us career-wise, but it just didn’t work for us. So now we don’t bother trying.

TA:  You’re sort of, as she described it, a package deal for universities.

JC: (laughs) Yes; we’ve been very lucky. Really one way you could look at us, is we’re one faculty member for the price of two.

TA:  (laughs) That’s pretty good. What else can you tell me about the rest of the years at IBM?

JC: It was a great period in life for me. We had a lot of continuity; worked on a set of challenging problems for a number of years, long enough to get to the experience of feeling we knew what was going on and how to control it through design! We didn’t get much management interference and we got good support, great colleagues -

TA:  Did your work end up in products such that you could see it?

JC: Not really. I didn’t really work for that, though. In those days, IBM scientists tended to have either an academic-oriented career or a company-oriented career. The key goal was excellence. Our management preferred that you do one of those well.

My technical career at Watson was academically-oriented. To the extent, I contributed to the company’s business interests, it was through management and through indirect technical influences. For example, I remember when I first published a paper on the minimalist information design model. The next morning I had calls from two IBM product development labs. That led to consulting interactions, and in one case to a joint study. But – oddly in a sense – those relationships with IBM product groups were mediated, indeed they were enabled by publishing in the open scientific literature!

When I managed the User Interface Institute, we got people on 1 or 2-year temporary assignments from the IBM development labs to visit Watson as Fellows. This was a very nice way to influence development groups. When these folks went back to IBM development, they knew our work inside-out.

TA: Right. From a research perspective, from where you were, was the internet starting to raise its cute little head up?

JC:  Sure. IBM in the late 70s had well developed email and network services. IBM had its own internal network called the VNet that was a worldwide network. We used it routinely to collaborate; usually just send email but we had more advanced ways to have more collaborative sessions with people – back in the 70s.

As time went forward, one thing that happened in IBM, was that we used IBM internal stuff not really less and less, but just less exclusively. I remember my bouts of Mac envy in the early 80s. I was one of the few people who had a Mac because I studied new user problems and I didn’t want to just study PC applications so I had Apple equipment too but basically no one in IBM used a Mac in those days.

TA:  That kind of makes sense.

JC: I don’t know; I’m not sure about that. Ultimately they became much more common and IBM opened up more to the general internet services other people were using; it wasn’t just Vnet. – and so forth.

TA: Were you talking to some of the folks at Apple at that time – wereDon Norman and Larry Tesler there already?

JC: Yes.

TA: I’m wondering how other people in our field started to come into your world.

JC:  I knew Don; he was another of my heroes from early on when I first got into psychology. I don’t remember exactly when I met him; probably around 1981. We were closely in touch with him. He had the group at UCSD.

I knew Larry Tesler from when he was doing the Lisa. I published a paper called LisaLearning in which I critiqued the Lisa training and user interface as far from a design panacea. My point was not that it was a bad system, but that it was so innovative that it was scientifically important to thoroughly and empirically deconstruct it – which I think Larry appreciates. Larry has an argument that I should have studied people who knew LESS about computing and information, that more naïve people could learn the Lisa better than more sophisticated people. I find that argument a bit strained.

Figure 2: The Apple Lisa
Source: en.wikipedia.org

Alan Blackwell has recently written a paper on metaphors that appeared in ACM Transactions on Computer-Human Interaction (TOCHI). He was talking about the emergence of the idea in HCI, and of course Lisa is one of the early desktop interfaces in a commercial system. He interviewed Larry and me as background for that, and inadvertently stoked up this little skirmish again. 

TA:  Oh my; you think he’s still mad?

JC: I saw an email; I know he’s still a little mad. Great systems never die, we just reconstruct them in our minds!

TA: HCI Tips; I think that’s the first one I’ve heard.

JC: He was the designer so he was invested in it, and he thought that I was a little too – well actually, he didn’t take issue with anything specific I reported or found – but he had an explanation of why it was this way and I guess it irked him that that wasn’t better known. Maybe that’s what it was. Maybe it is now because I think Alan put a footnote in his paper.

TA: Good. So now, where are we in your career.

JC: Let me just give you the highlights; it’s a long and winding road.

In ’94 I left IBM. I left because we got the offer from Virginia Tech. It was one of these ‘one faculty for the price of two’ deals. We thought well, there aren’t going to be a lot of offers like this where we’re both going to get a regular faculty position at a school with outstanding strengths in HCI, and this seems like the right time to leave, so let’s do it.

After having gone to IBM for a year and stayed for 18, I left. I was stunned that I ever did it, because I had been there long enough finally to think I was never going to leave but looking back, I’m really glad we did leave because the Watson Center, after that period of time, almost right at that time, changed dramatically.

It’s not so much for the better or for the worse; the way it changed, is it got much better, more effectively, more closely integrated into product. So now the normal way that a project is structured there, there’s funding from a product, there’s a plan to integrate the work with the product, there’s guidance from product organization.

I think IBM was never as ineffective, say, as Xerox Park at impacting the real products of IBM but after my time in IBM – it was coincidence – they reorganized to become much more effective, and I think they have become much more effective.

The way I reconstruct this looking back, is that if we had stayed, it would have been a new job anyway, a new IBM that was reinventing itself, so we may as well just get a new job.

TA: Yes; and stay in pure academics.

JC: Right. We were able to stay in pure academics. I don’t think staying would have been bad at all, the Watson HCI group is still one of the top research groups in the world in our area. It just would have been probably as big a transition as the one we made in moving to Virginia Tech. And it was a big transition.

When you go to academics, you become an intellectual entrepreneur. At IBM, if I had an idea and I wanted to pursue it, I just had to talk to one guy; my manager. If I want to do something as a professor, I have to write grants, pitch ideas to potential partner organizations, recruit and train students, and all the other stuff we do here – like teach! It’s much more cumbersome; it’s much more slow but I was older when I started this, so it’s fine for me.

When you’re younger, you want to strike fast when you have an idea, you want to act on it immediately. It’s certainly better to be in industry where they have resources within the organization to pursue their own venture activities.

TA: Yes. You said ’94; did Virginia Tech – the year after I started looking for grad programs, I brought it up with a couple people, that at that time there really wasn’t any place to go if you were interested in this.

Did Virginia Tech already have a department? What department did they put you in? 

JC: I joined Virginia Tech as the head of the Computer Science Department. I know this sounds a bit strange, since I was just holding forth about being a confirmed psychologist. But in the early, almost all HCI programs in North America were in Computer Science departments.

It may also seem odd that a university would hire someone who has never been a professor at all to be a full professor and a department head – in charge of the safety and security of discipline’s worth of students and faculty members.

I was already 44 when I started my college career, and with respect to research I was very well known. So, Virginia Tech wanted to bring me in to lead their HCI area, and it is hard for a university to bring people in at senior levels without some sort of extraordinary argument. In my case, it was the “outside head” argument. The Computer Science department went to their dean and said we need a new leader; we have to go outside the university to find a new leader. They got me. 

TA: Was anybody in the Computer Science Department teaching HCI related courses?

JC: Yes. In fact Rex Hartson – do you know him?

TA:  No, I don’t.

JC:  Rex is retired now, but he was there – Bob Williges was in the Industrial Engineering Department; he is also retired now. There’s also a woman there who was a long time research scientist; she was not a faculty member but was a researcher at Virginia Tech, Deborah Hix. Those three were well established, HCI researchers. That’s a lot for one university!

Virginia Tech was already on the map for HCI. So, when Mary Beth and I went there, for an academic group, we had a pretty big group. And over the years, we hired a lot of people – Doug BowmanScott McCrickardManuel PerezChris NorthDeborah TatarSteve HarrisonAndrea Kavanaugh. And after I left, they hired more! They have maybe ten people.

TA: Is it a full department?

JC: No; HCI is not a department. They’re mostly in computer science and the rest are mostly in industrial engineering; same thing as before.

In 1996, we formed a university-wide Center for Human Computer Interaction at Virginia Tech and to help to bring people together, and also get other people – like people from the education and psychology departments; sociology, electrical engineering; to get them involved too.

TA: Was that tied to a degree program?

JC: No. It was tied to something called a certificate program where you could either enroll and just get a certificate, or you could take it as a graduate minor, a sort of specialization add-on to any graduate degree Virginia Tech offers. For example, you could get a PhD in Communications but have a graduate minor in HCI. Quite a few people did that. That actually became pretty popular and in fact I proposed a similar program here at Penn State.

TA: What decided you on coming to Penn State?

JC:  Change. Again. Plain Change. I was very happy in Blacksburg. It’s a great place to live; great school.

In 2003, Virginia Tech decided to move its Computer Science Department into the College of Engineering. Prior to that time – and during the whole time that I had been Department Head, they were in the College of Arts and Science. We had proposed to the university to form our own College. We wanted to form a college of computing and information – what is sometimes called an i-school.

Historically, computer science departments either grew out of math departments, in which case they ended up in the college of arts and science, or they grew out of Electrical Engineering departments, in which case they ended up in the college of engineering.

Virginia Tech was a math-based department. We were in this college of arts and science. Universities are very futile and unbelievably living out of date a couple of centuries ago. If you are in this college of arts and science, it’s a real resource mess. Our department was always trying to deal with this and it was a struggle the whole time I was there. 

TA:  I would imagine, computer science is a very awkward stepchild for folks who are in pure arts and sciences.

JC: Absolutely. We definitely stood out as an oddball in a group of oddballs.

TA:  I’m sure some of those professors didn’t know what to do with you in the staff meetings.

JC: Exactly. So as part of a university reorganization – yes, universities have reorgs too, just not as frequently, we were offered a deal. It was one of those things like the deal you can’t refuse – and that was to join the college of engineering. It seemed to me that this was what the department had to do to ensure the best possible future.

Nevertheless, it was not really good for me personally. Although I can live with computer scientists, I am not a computer scientist. I am really a psychologist. I did not want to have to put up with the hassle of dealing with learning to live in the engineering college of Virginia Tech.

To me it felt a lot like the move 10 years before from IBM. Virginia Tech was changing, and even if I stayed I would have had to readjust. Thus, the costs of moving were reduced.

It just so happened, much as it did ten years before, that that year at a conference, I met a guy from Penn State. They had made me an offer the year before and I had turned them down. I said I am interested now. So they came back and here I am.

TA: How has it been?

JC: It’s very good. The department I am in is not computer science. It’s called the College of Information Sciences and Technology. We have some computer scientists but from my point of view they are mostly the ones I want to have near me; like people who are software engineers or search engine guys or people interested in geo-spacial information or something like that.

We also have people interested in organizational issues in systems; how organizations change or government policy in the information technology (IT) sector or law. We actually have a lawyer as one of our faculty; law in IT or IT economics; we have two economists on the faculty.

We have a sociologist. I really like the fact that it’s under one roof that I have all this mixed bag of IT people with a broad range of perspectives.

There’s a rough side to interdisciplinary – maybe you already know this – everybody disagrees about everything – they all came from a different place, they’re trained in a different way; they have different values and so forth, so it’s not like it’s living in heaven. But it is very interesting. It’s very rich. And it’s not so narrow and alien the way it would be, I believe, if I had gone into that engineering college with my colleagues.

TA:  It sounds like the problems are much more vibrant for you than for lots of people. Having problems and everybody is disagreeing is kind of fun, I would imagine, too.

JC: It is for me. I have a thick skin and disagreement is stimulating to me. I am constantly stimulated here and I’m learning stuff and that’s the way I like it to be.

TA:  Can people get degree programs there?

JC: Yes; we have a bachelors, masters, and PhD in Information Sciences and Technology. That can mean a lot of things. We have a bachelors in Security and Risk Analysis. We are developing other programs.

I think our programs tend to be flexible and individualized. Students can take a lot of initiative in pursuing skills and concepts they value. In computer science there is a mentality – I call it the interchangeable part mentality: Computer scientist, especially at the bachelor’s and masters’ levels, should all be the same – the same skills, the same courses. 

TA: It’s almost like a trade, I suppose.

JC: Yes; exactly. That’s not my philosophy and I just don’t think that’s possible. I think the further out you are from the core of computer science, the more limiting; the more uncomfortable that is.

TA:  I wonder if our field will ever feel like that or if it’s going to stay rich, the way it is.

JC: I don’t know. I’m actually starting to wonder – this school I am in – they call it an I School – have you heard that term?

TA:  Yes; they are starting to pop up.

JC:  I am actually pretty positive about this as a model for where HCI really fits into the academic landscape. Some of the other examples of I Schools would be Michigan – very strong HCI group there in the I School there – University of Washington; same thing – Irvine, California – University of California, Irvine, same thing.

There are several examples of it and I think there may be more coming. To me it is a good deal for HCI.

TA: Let me ask you one final question. It’s the question I like to end up with, which is, looking forward now, what do you find fascinating? What really floats your boat in terms of what you are thinking about or researching today?

JC: I’ll give you a couple of answers to that.

One, is when I went to Virginia Tech, they had this project starting. It just happened to be starting the year we got there, called the Blacksburg Electronic Village. It was an experiment in community networks. They didn’t invent community networks but they produced one of the ones that’s one of the best known ones.

I got involved in computing in the home, let’s say, or in the community – civic computing. I’m still very interested in that and I am still pursuing it.

Now the frontier in that area is ubiquitous computing so it’s community networking, infrastructures implemented using broadband wireless. So that’s something that fascinates me and interests me.

I like the idea of more closely integrating information services with ordinary life and asking how can this empower not just our ability to shop but our ability to connect with other people and maybe even do worthwhile things in our neighborhood.

So that’s one answer.

The second thing that really excites me, is, oddly enough, teaching. Of course when I went into the university, I had to teach. Teaching was part of my responsibility. I have really gotten interested in teaching. I’ve made a teacher course in usability engineering that has basically been an experiment that’s been going on for seven or eight years now.

I’m very interested in new kinds of teaching materials and teaching activities; injecting different kinds of project work and case study analysis and different kinds of techniques to learning usability engineering.

I see both of those things I just mentioned, as technical projects almost without end because I don’t think there’s a question I’m trying to answer, rather, it’s direction I’m going, and I don’t think there is a final goal because I don’t think there’s a limit to what we could do.

Finally, my third answer would be I am feeling open to new areas and questions. I was always a very focused researcher, at least in my own mind. But I am 56, and I have enough success in life up to now so that I don’t feel like I must stay on the straight and narrow.

I have enjoyed letting my students take more initiative in determining what we investigate. In a couple cases, this has been extremely rewarding – students have brought me into new problem areas I would most likely not have gotten into on my own. I really enjoy letting my students drive, if you know what I mean.

TA: Yes.

JC: I think that is a great luxury when you can afford it. Of course, if you’re a young professor trying to make your own name, you probably can’t afford it but I can. So I am enjoying that.

I have some really good students. They’ve got really good ideas. They’re taking me to some interesting places.

For example, one of them is interested in how information technology can evoke and sustain creativity for people working collaboratively. That’s an area I can honestly say, pretty surely, I would not be working in if I hadn’t had this student.

TA:  And I am sure they love it; getting the support and the interest from you.

JC: Yes, and they like to drive too. People like to feel they are in control and not just following the professor.

TA:  On that note, I want to thank you for talking to me today, and sharing all those stories. I had no idea about a lot of that. I’m sure most people who read this wouldn’t have either.

It’s fascinating to hear how all the work you did in different fields ended up creating a straight line although who would have thunk it in the beginning.

JC: Things look simple once you get there.

TA:  Exactly. Thank you.