From AI to Piaget
Posted on December 10, 2007
David Klahr (bio)'s career route to cognitive science started with engineering and business.
I started out, I went to college to be a civil engineer. I wanted to wear a blue work shirt and stand on a hill and build a hospital on that hill. I thought that would be cool, and then everyone said, "No, no. Civil engineering is passé. What you want to do is be an electrical engineer," and this was back in the late '50s, so I went into electrical engineering and I did okay, but I didn't like it much, and then somewhere around my junior year, I got interested in computers.
Now this is, dawn of the computer age. We weren't quite using the abacus then, but it was a long time ago, and MIT was pretty far along and so I started doing computer programming. I thought that was cool, and I had an advisor, who had read some stuff by some guys at Carnegie Tech who were saying computers can think, and why didn't I read some of that stuff. So I wrote a little program that learned how to play a game, and it was really a little piece of artificial intelligence, is what people would call it now, but I didn't know what I was doing.
Then I decided I wanted to have a career as a programmer, because I really thought computers were cool; so I started working for a small place. Actually, almost across the street here, in Copley Square. It was a small outfit that started, and had some government contracts to do computer programming for the Air Force mainly, but one of the things they wanted to do was study what they called an "adaptive machine," and this was the early name for things like artificial intelligence.
It was a machine that was really a program that would change parameters and learn something. At that point I was strictly a computer hacker, a programmer, but I had this interest in intelligent "thingies," but I didn't even know what they were. I'd never had a psych course in my life at that point. I was out of college.
Someone said to me, again — just the luck of having smart people around me. They said, "Why don't you go to this place called Carnegie Mellon. They have this business school where this guy does this thinking machine stuff is. That's where Herb Simon is and that's where Alan Newell is, and why don't you go? You could study business, but you could also do other things."
So I went to business school, and while I was there I quickly decided what I wanted to study was organizational decision-making, and then I had a few courses in grad school that had to do with organizational decision-making, managerial decision-making, information flow in organizations.
But you can see it's starting to move into cognitive science little by little. And for my dissertation, I studied the decision-making processes of college admissions officers, and looked at the way they weighted at different factors, and I had this big complicated mathematical model to do it all.
Somewhere along the line I went out to a summer conference. I went along as the kind of flunky to one of the artificial intelligence faculty, and there are people there who are junior people who just got their PhDs, big shots and really big shots, so Herb Simon was there and so was Bob Glaser, and so was Leo Postman, and so was Lee Cronbach, and then there were a bunch of low-level people.
And here's the defining moment in my life. I met a guy who had just gotten his PhD in education from Scotland. His name was Ian Wallace, and he said, "What do you do?" and I said, "Oh, I do complex information processing on computers," and I went on forever, just blathering away about how great this was, and then I realized that the rules of social discourse suggest that you should sometimes ask other people what they do.
So after about an hour of this, I said, "Oh, what do you do?" And he said, "Well, I'm a Piagetian," and I said, "What's that?" and he said, "What do you mean, 'What's that?'" I said, "What's a Piagetian?" and he said, "Oh, someone who studies Piaget," and I said, "Who? Who's Piaget?"
There's a watch called Piaget, but ... So I knew nothing about cognitive development. Zero, and he started to tell me what Piaget did, and Piaget was interested in cognitive structures and he had this very complex mathematical language for doing it. And I said, "That's the wrong language, because it's static, and computational models have a dynamic language, and wouldn't that be cool if you could study conservation and class inclusion and transitivity; if you could study that writing little computer programs for those kinds of models?" and so, that was the big deal.
I was riding this crest of the business school world. This had nothing to do with Piaget or psychology, so the trick was, "How can I do this stuff with this Ian Wallace guy that has to do with cognitive development when I'm already on a business school track?"
So, then I started to look around for agencies that funded research like that and I went to Chicago's business school and I taught there at the business school and in the math department, some of the AI stuff, and I started to apply for research grants. I applied for a Spencer Foundation thing.
I think I didn't get it, and then I applied for a Fulbright and I got the Fulbright, and then I wound up going to England, Scotland actually, to work with this guy, and my whole career shifted at that point, and then I came back to Carnegie Mellon and went into the psychology department.
So, I think the moral there is if you want to shift fields a little bit, make sure you have a solid base, but look around and try to be innovative, and you have to take some risks.
When I told people at the University of Chicago, I was an assistant professor in my second year, and I said, "I'm going to take a year off and go to England on a Fulbright." And everybody said to me, "That's the stupidest thing I can imagine. You haven't even been reviewed yet for promotion and you're already leaving, and you're already changing fields, and you don't have a track record, and you've only published a little bit in your own field. What's the matter with you? Are you nuts?" and I said, "No, I really want to to do this."
When I was still in Scotland, I got a letter from Carnegie Mellon asking me if I wanted to come back. I was lucky, because I was around a person who was a giant in the field, so Herb Simon was my advisor and he said, "I think that cognitive psychology knows enough now that we could start to engineer better education in business school; so would you like to come back and be," he called it, a "learning engineer." "Would you like to take what we know about learning and forgetting and memory and get the guys who teach accounting and economics and all the rest of it to use that stuff in a better way? Be the learning engineer?"
And I said, "That's kind of interesting, but I really want to do developmental psychology also, so here's the deal. "Suppose we do that and I really make that my teaching load? I don't have to teach anything, and then my research would be over in the psych department," and they said, "Okay. Come back." So what's the answer to that one?
The answer is some institutions are much more flexible than others and sometimes it pays to figure out the institutional and administrative history of a place before you go there if you want to shift fields. Some places are very rigid and these guys don't talk to those guys, and it's really bad. Other places you hear just the opposite story.
There's a rich intellectual interchange in a place, and for people who are going on the job market and who have some options, I think that's something you want to look into and you have to read between the lines when you're being recruited, because everybody will tell you how rosy it is, because they want you, but you really have to look for some evidence that there's a lot of interdisciplinary work. Carnegie Mellon was very easy to move around and change boundaries and things like that.
Excerpted from interview with researcher at the 2007 SRCD Biennial Meeting in Boston, MA.
Please note that the feedback is viewed only by 4researchers staff and is not intended for communication with individual contributors.
Use the form below to submit feedback about this article. If you would like a response, please be sure to include your e-mail address.