Q&A with Rick Johnson and Clif Pollock
Professor C. Richard Johnson, Jr. has changed his career focus several times by bridging gaps and figuring out what people need—and then making it happen. He recently met with ECE Director Clif Pollock to explain how he has progressed from adaptive control to telecommunications to art history, and how he succeeds as an electrical engineer in the art world.
Pollock – Your teaching and research have always been distinguished by the level of interaction with experts and by the dexterity in which you change research directions. You started your career in control and today are working with art museums. How did this evolution start?
Johnson – This whole idea about transferring between two areas started when I was in graduate school. A classmate, Tom Mitchell, and I arrived at the same time and served as TAs in electronics for a circuits course. Tom was doing artificial intelligence (AI) and machine learning systems and I was working on signal processing. We hung out together and I noticed a lot of similarity in our fields. I have one of the early publications with him pointing out how AI has a lot of the same structure as signal processing.
Pollock – So you recognize similarities between fields and then zero in on these.
Johnson – Yes, and it’s because I’m connecting two things that I know something about. The thing is at some point you need an expert to help you finish. Sometimes it’s the person inside the application. Sometimes it’s the theory wiz who is a hundred times better than you. I’ve always thought and I’ve always told my students that sometimes the easiest thing is to put together two things that you are familiar with.
Pollock – But this seems to come naturally for you. Is there something that motivates your changes?
Johnson – With me, it really comes down to circumstance. With adaptive control, everyone looked up and realized that the theory was done. I could no longer attract the best students to my research group, or even some of the best—they didn’t want to do control anymore, they all wanted to do telecommunications.
I saw opportunities in telecom to combine control and signal processing, and with the help of another college friend, I built a research and teaching program in communications.
The next thing came along and the physical layer communications course collapsed. Again, the best students weren’t available anymore and I thought, “For once I’m going to work on a problem that I care about instead of just using it as a vehicle for doing what I really enjoy—doing fun things with images, interacting with Ph.D. students, and teaching them how to do research.”
Pollock – So is this how you got started in the art world?
Johnson – Yes. For years Tom Mitchell and I had been trying to find a way to work together. I called him up and said, “Hey Tom, I’ve got this idea. When we were in grad school, I did that art history stuff. All of the art exams are about attribution: the professor puts a slide on the wall and you have to give its name, date, and the artist that painted it, and write a short essay.
“So I though, with a little bit of machine learning or even a little bit of image processing, I should be able to somehow put this authentication or artist identification on some sort of steroids. Wouldn’t that be fun?”
He said, “I know someone in the Netherlands who’s actually trying to do something like that with the Van Gogh museum.” So I contacted this guy and told him I have a real interest in the problem area they’re working on, trying to identify brush strokes for fakes and such. I was in a good place because I can speak the language of engineering and I can speak to the museum because I have some background and an interest in art history.
Pollock – Did you feel uncomfortable stepping so far away from traditional electrical engineering subjects? How did your colleagues feel?
Johnson – As academics, our rewards are traditionally based on specializing narrow and deep. That’s where research support comes from, that’s what attracts students, that’s what generates results that are considered typical for publication. But that’s not what motivated me—I wanted to do something I was passionate about. I was lucky to be at Cornell, where people appreciate cross-disciplinary work. On every side, I was supported in moving in this direction.
Cornell has a culture of attempting to lower every potential barrier to cross-disciplinary research. That’s just part of the philosophy here, and everything’s in place to do that. And because of this environment, I feel emboldened.
Pollock – So, you were an electrical engineer knocking on the door of an art museum. How did you gain traction in
Johnson – In the last 15 years or so, a subfield has emerged called technical art history that needs forensic evidence to support the work. An object can tell you an enormous range of things—how it was handled, where it came from, what it was intended for—and these things can be pieced together. However, most of the evidence is based on chemistry because that’s the background of most conservationists. There’s no injection of computation to speak of. So now, we’ve written software programs that assist conservationists with this type of
Pollock – Cool! Are you turning conservationists into geeks?!?
Johnson – I teach conservationists how to use these programs and we’re trying to get them to establish a procedure, a standard format—this is completely foreign to the field. For instance, I’ll ask a group of conservationists what standards they follow and they readily admit that everybody does it differently. But no one knows why, really. We’re learning about the culture of the field so when we do find these cultural gaps we can say, “Well, if you did it like this, then this other thing would happen, and you’ll get this extra bonus. How great would that be?”
Pollock – How is your research process different now than it was when your focus was control or telecommunications?
Johnson – What’s interesting is the way research is done in museums. They mount exhibitions and do things to bring in visitors, but they also do research because that’s what maintains their status in the museum world. And research at museums usually happens by the curator saying, “I’m doing something on this particular thing. It has a theme of this, so I need five examples of this.”
I come in saying, “Give me the X-rays for everything. Can we get all 700 of them tomorrow?!” And they look at me and ask, “Why would we ever do that?” As engineers, we have a way of measuring everything, then seeing what we have, and only then figuring it all out. And the museum conservators are just horrified by this! I often get asked, “Why would you waste all of your time on that? You’re looking for this, so you just need to look at the three examples of this.” So the process completely changes.
Now both methods are needed because if you’re not paying attention you’ll miss the thing you need most. With almost every experimental thing I’ve ever been involved with, halfway through we said, “We should have connected this over here as well.”
Pollock – How do you envision the future connection between engineering and the humanities?
Johnson – Honestly, I think digital humanities is going to be huge, another one of these gigantic things that we can play a role in. And it’s not just in art history—there are many more opportunities we haven’t thought of yet.
I’m lucky enough to do work in art history as an engineer. It’s a passion of mine and I hope others will end up doing the same thing—merging their skills as engineers with something they’re passionate about outside of engineering. Then we’ll really strengthen this connection between engineering and
Pollock – What advice would you give to a student wanting to merge ECE with a field that is completely foreign to engineering?
Johnson – It’s a whole lot easier if you can find the part of the orchard with low-hanging fruit. You don’t need a ladder, but you’ve got to look a little bit.
I fully expected this to go nowhere, I just thought it would be a great ride—if I had a chance to be inside the museum, to learn this world, then it would have been enough. But it’s turned into a big deal now. Everyone that has ever been involved, especially those first green-lighted at the Van Gogh museum, they all agree that it’s gone beyond anything we would have imagined in terms of the reach, the success, and the impact.
It’s been a great ride and I’ve learned a lot. Their world is different, but there are many things that are similar.
by Jackie Swift Around 50 years ago, Gordon Moore, chief executive officer of Intel, predicted that every two years the number of transistors a microchip could contain would double. Known as Moore’s... Read more about Keeping Computer Hardware Fast and Furious