The psychology department at Carnegie Mellon has always had a strong group of researchers interested in cognitive science—encompassing domains such as information processing, problem solving, and artificial intelligence. Many of the faculty that founded the department helped to found the fields of cognitive science and artificial intelligence. Given this history, it is not surprising that part—a large part—of our impact on the field, as reflected in number of citations, derives from our work in the area of cognitive science.

“The juxtaposition of the symbolic and PDP views has created a certain amount of intellectual tension within the department over the years, but the tension has been welcomed (even relished) by all and has had a positive impact on the quality of work that gets done, in both camps.”

This is just a small part of the story, however. Our scientific impact has been enhanced by at least three other noteworthy events. First, historically the kind of cognitive science that was done within psychology had to do with higher-order mental events—emphasizing symbolic processing of information and executive functioning. About 20 years ago, a very different way of thinking about cognition and behavior began to emerge. This new approach provided a direct challenge to the traditional way of thinking about cognition, and it revolved around the operation of what was called a parallel-distributed processing (PDP) system. The idea was that cognitive and behavioral phenomena could be understood and predicted from a system that consisted of low-level, interconnected pathways. These pathways could excite or inhibit each other. They were also sensitive to inputs coming from the environment, and so could learn from that incoming information as reflected in the manner in which the pathways developed and operate over time.

 Is this how the brain is supposed to work?

The idea is to account for cognition in ways that keep in mind the neurological basis of the brain. So these systems reflect or represent neural networks. They take a very simple, basic unit and link them together in various ways into pathways, and the interaction of these basic units gives rise to the complexity that leads from cognition and thought to decision making and higher-order reasoning. They start with a basic unit, and demonstrate through various models and simulations how these systems can give rise to very complicated behavior.

This whole information-processing approach was using the computer as a metaphor for how people think. That approach was top-down, thinking about higher-order thought and how that evolved. This PDP idea starts from the lower end, from basic units that reflect how the brain is organized; single neurons projecting to other neurons. They’re trying to describe the same phenomena but from very different theoretical perspectives.

There was much discussion of what to do about this upstart PDP approach among the faculty within psychology. The answer was very creative. Rather than keep the opposition at a distance, invite them in. Hire the best person who represented this emerging view and interact with him/her on a day-to-day basis, if nothing else but to keep track of what the devil was up to. The department was fortunate enough to make good on this strategy and we did hire a top person working in the area. The department now has a number of faculty who utilize a PDP framework. The juxtaposition of the symbolic and PDP views has created a certain amount of intellectual tension within the department over the years, but the tension has been welcomed (even relished) by all and has had a positive impact on the quality of the work that gets done, in both camps.

In the mid-1980s, which is reflected in the citations. The reason why this department is cited so heavily is that it is full of very bright creative people. All of them. If you had one or two, it would not be a citation impact institution. We can describe how these themes emerge, but unless you have all really good people working with those themes, you’re not going to get the citations. The bottom line is you get so many citations because you have more than your share of bright, creative, hard-working people. How does that work? Who knows? You start off with a group like that and it attracts, like magnets, other people of the same quality.

The ultimate goal was to provide a better understanding of human cognition. Each side more or less challenges the other with their own findings. And right now there is almost an equal number of PDP people and higher-order processing people in the department. And they coexist quite nicely together. One will identify a finding that seems hard to explain in other terms, and they will all then try to come up with an explanation consistent with their own way of doing things. It keeps everybody honest. You don’t know if your worldview is distorted if everybody shares that worldview. Only if somebody starts challenging it do you think maybe you have to think about it. So this diversity within the department allows for this kind of challenge. Once again, it creates a kind of intellectual tension that in turn creates a lot of energy in a positive way.

Another event that influenced our rate of citations was a transition that occurred within the social/personality area of the department of psychology 20-some years ago. Even further back yet, Carnegie Mellon made a conscious decision that, given the small size of the university, it would not try to cover all disciplines and all domains—as a large state research university might. Rather it would try to develop small pockets of excellence. During the early 1980s, the social/personality area was characterized by work on social cognition—the application of a cognitive-science framework to understanding social processes and phenomena. This theme had been chosen in order to play off of the psychology department’s strength in cognitive science. For a variety of reasons, this foray into social cognition just didn’t work. And so the department decided to move the social/personality area into an entirely new direction—toward what’s known as health psychology.

First, the then-head of psychology had an interest in health policy and thought a move in the direction of health psychology would add an important dimension to the department’s activities. Second, several members of the department were already beginning to move their own research agendas in a health psychology direction.

Finally, there was a movement afoot at the University of Pittsburgh to develop their own presence in the health psychology domain. By developing an alliance with the emerging health psychology group at the University of Pittsburgh, the department of psychology at CMU might be able to leverage that association into a critical collective mass large enough across institutions to attract other high-quality researchers of interest into the local community. For these reasons, the decision was made here to go after one of the top researchers in health psychology in the early 1980s. This effort was successful and that person remains a member of the department and also one of the most-cited contributors to the field. With this early acquisition, the department was able to hire several other outstanding health psychologists over the years, resulting in the establishment of a world-class program.

This is another event that helped to shape CMU’s success in the field of Psychology/Psychiatry: the creation of the Department of Social and Decision Sciences and within that department the development of a cadre of researchers interested in decision-making and risk assessment. This was another part of the university’s effort to develop "pockets of excellence" and to build in areas in which it has a competitive edge. Work on decision-making and risk assessment shares a broad border and considerable overlap with work being done in cognitive science. So the creation of the new department and the decision to build in the areas that it did was not totally coincidental. The group of research scientists at CMU studying decision-making and risk assessment is first rate and has contributed much to the scientific impact that the university has had.

Yes, very, very strong, both within the Humanities and Social Sciences College and across the university. This university has always prided itself on the collaborative arrangements made. And Pittsburgh is a very interesting place in that light, more generally. We have these two institutions here—CMU and the University of Pittsburgh—and not only do collaborations occur within the institutions but across institutions. I can’t think of any other institutions, so close together, that are so collaborative.

Not really. The main papers are from cognitive science, and that’s partly because it’s the oldest tradition represented in the department, but that group also comprises the largest number of faculty members. We have 25 faculty members, and the cognitive-science group is probably three quarters of us. Only about 20 to 25 percent are faculty in social/personality/health psychology. So we’re most cited in cognitive science, but it’s also clear that the social/personality/health area contributes enormously to the citation impact of the department. At least three of the most-cited papers are from folks in this social/personality/health area. This place will always be strong in cognitive psychology, given its history and other themes throughout the university. But moving into the health psychology area, or trying to orient the social/personality group to health psychology in the early 1980s, really helped add to our impact.

Sure. But, as I said, the university has dealt with this in part by creating what they call "pockets of excellence." They don’t try to cover all bases, only the bases they think they will be good in. Then you try to build on those areas in various parts of the university. So if somebody happens to work in an area that we have strength in, then we’re very competitive. But there will always be people we might want to hire, who won’t be interested in coming here because there’s not a critical mass in that area and so they won’t be comfortable.

It’s a very social department. Routinely groups of five, six, seven people are going to lunch together every day. People come from all our three main areas, developmental, cognitive, and social/personality/health. Everyone knows what everyone else is doing. People socialize a lot together. There was a work ethic established here a while ago by some of the founding faculty in the department that continues to permeate the place. There’s a little circular stairwell in the department. On the top is a stained glass window, and there’s a quote on it from Andrew Carnegie: "My heart is in the work." It really sums up this place.

It’s clear that one major thrust, which has already begun, involves the integration of cognitive science with neuroscience. There’s an interest in developing an ever deeper understanding of the neural basis of cognition; figuring out how neurological structures at various levels of complexity and integration give rise to emergent psychological functioning. Specification of such structure/function relationships will help to understand how the brain works, but will also help in the development of applications like prosthetic devices of various sorts that might enhance the functioning of people with different disabilities. It is clear that the field of cognitive science will assume an increasingly greater neuroscientific flavor as time goes on. Down the road, it will be necessary to have access to the relevant methodologies (magnetic resonance imaging techniques, for instance, or single-cell recording methods) that allow this integration to occur. In this regard, it’s worth noting that not only is CMU a citation leader in the Psychiatry/Psychology domain, but in the domain of Neuroscience & Behavior as well.

One final comment here is the importance of interdisciplinary research. The hot areas of research today necessarily involve participation of individuals from very diverse backgrounds. As I just noted, the evolution of cognitive science is giving rise to increasing interactions between cognitive psychologists and neuroscientists, with contributions from computer scientists and people working in robotics as well. Work in health psychology requires the collaboration of psychologists, physicians, immunologists, pulmonary specialists, and so on. To be at the cutting edge of research seems to demand diverse interdisciplinary input from many sources. And so universities will have to have the flexibility to incorporate structures into their organizational base that provide for and facilitate such cross-disciplinary work. One route that CMU is pursuing is the increasing reliance on centers that integrate work being done in diverse departmental units. Whatever the solution, it’s clear that interdisciplinary research is here to stay. Universities, in order to be intellectually successful in the future, will have to have some way of embracing that interdisciplinarity.

Dr. Michael Scheier
Carnegie Mellon University
Pittsburgh, PA, USA