By Susan E.F. Chipman
Members of APS may find it interesting that I chose to enter psychology specifically as a science, among other sciences that I might have chosen at a critical point in my life. Like many research psychologists, I came to the field via a convoluted path. I majored in mathematics at Harvard, also studied physics, and took a small but choice sample of psychology courses to meet general education requirements: George Miller and Jerome Bruner’s Psychological Conceptions of Man (actually a very early cognitive science course), and Erik Eriksson’s Psychological Study of Life Histories.
The early days
Losing interest in mathematics and unable to decide on a new academic direction, I would need to find a job, but at that time even women who had graduated from Harvard had difficulty escaping the secretarial ghetto. I decided to apply to the Harvard Business School, which had recently opened to women, was admitted and even awarded a fellowship (see Chipman, 2009, for an account of being one of only 10 women among 700 students). Women students today may have difficulty imagining what that era was like. In an interview with a major consulting firm, I was told (despite a Presidential order forbidding sex discrimination) that they might hire a woman, but they would certainly never promote her! I was not asked if I could type.
The HBS experience helped me decide I really wanted to do a science; against that background, all sciences came to seem alike. I chose psychology because it seemed to me that the distance from the beginning of one’s studies to the frontier was very short and that the immature state of the science meant it was still possible to contribute to the fundamental structure of the field. For personal reasons, I applied only to the Harvard Psychology Department. My backup plan was to go into physics. A first-year research project with Roger Shepard (implementing his first idea for research on mental imagery) resulted in my first publication (Shepard & Chipman, 1970), but he soon left for Stanford. For obvious reasons, I explored mathematical psychology and took a special prelim, but decided against it. Miller had left Harvard, and Bruner had focused the Center for Cognitive Studies on pioneering studies of infants, also not my choice. Bruner nevertheless supervised my education in cognitive psychology, but was transitioning to Oxford. I also explored language research with Roger Brown. Deciding on a thesis topic was not easy. Eventually, I selected visual pattern perception because I considered it a very important unsolved problem: Success might make it possible for a woman to get an academic job, which I was beginning to realize might be a problem.
Memorably, S.S. Stevens asked me what I was doing and said, “Many great minds have worked on the problem of pattern perception and have failed. You, too, will fail.” At one point my nominal advisor advised me to write up some of my research and submit it as my dissertation because a newly instituted deadline for completion was approaching. Although no defects were found, it was rejected for “lack of sufficiently novel and interesting results,” a standard that I did not believe had been universally applied. Submitting a second dissertation (on the perceived complexity of visual patterns) in 6 months made me something of a legend in my own time among the graduate students. The results were more novel and interesting, but I was totally exhausted.
Now I needed a job. My final advisor, E.B. Newman, wrote a letter on my behalf to the chairman of the Psychology Department at the University of Colorado, where my husband was already working as an astronomer. The response was that the department had a faculty member’s wife who could teach a course or two if needed! By asking challenging questions at colloquia, I attracted attention. Bill Battig offered me some lab space, and Peter Polson gave me access to the departmental computer, enabling me to do research. In a state with only 3 million people, there were very few relevant jobs, but I applied for several, and remember a few bizarre interview experiences: “He (the chairman) just wanted to see a real live specimen of a Harvard graduate.”.
After a year I went on the national job market. I ended up as a visiting Assistant Professor at the University of Michigan, under their mathematical psychology training grant. Although it did not lead to a permanent job, my year at Michigan was very important. David Krantz, Amos and Barbara Tversky, Clyde Coombs, Judy Reitman, Jim Greeno, John Anderson, Lynne Reder, David Kieras, and Clayton Lewis were all there that year. I also applied for and received an NIMH small grant to continue my developmental work, and then applied for an NIH postdoctoral fellowship that would have supported me doing my research at Colorado. Due to budgetary problems at the NIH, by the time I was awarded the fellowship, it was too late, and I had to decline it.
Sometimes I wonder about the road not taken, which would probably have led to being a professor specializing in pattern perception. (Contrary to Stevens’ prediction, I made significant progress in this area; Chipman, 1977; Chipman & Mendelson, 1979). While waiting to hear from NIH, I responded to a National Institute of Education (NIE) ad for a cognitive psychologist. I left my husband behind, but after 3 years he joined me in Washington.
Quite soon after joining NIE, I unexpectedly took over the management of a large grants competition for reading comprehension, mathematics learning, testing, and teaching research. I was then asked to be the NIE program officer on the joint NIE-NSF program, Cognitive Processes and the Structure of Knowledge in Science and Mathematics (topics in which I had high interest, given my background.) There were only two competitions, but several famous research efforts were supported. In addition, I recruited all the young cognitive psychologists I knew to serve as reviewers, thus attracting some to research of this kind, notably Susan Carey. After these successes, in a reorganization I was made an Assistant Director of NIE, heading a division responsible for research on mathematics education, general cognitive research, social influences on learning and development, and computers in education.
Thus, for a while, I was in a position that integrated my wandering educational history: managing largely psychological research on the learning of mathematics. Inheriting a set of grants on the issue of women (or girls) and mathematics, I accidentally became an expert on that topic (Chipman, 2005). In addition to conventional grants competitions, the NIE issued some RFPs for larger research projects that had to be defined and sold to upper management. These provided staff members an opportunity for significant and creative, if anonymous, research contributions. RFPs issued by my division included: Research on the Perception of Graphs and Charts, the Cognitive Demands and Consequences of Computer Learning, and a review of programs purporting to teach general thinking, learning and problem solving skills, a start to a new program area (Chipman, Segal & Glaser, 1985). I found I enjoyed research management.
Coming from a basic research background, I was surprised to discover how politicized both educational research and practice were. There were right-wing and left-wing approaches (more ideological than evidence-based) to teaching both reading and math. When Ronald Reagan was elected President, the Science Education Directorate budget at NSF was promptly reduced to zero and virtually the entire staff was dismissed. NIE’s demise was more gradual. NSF staff served “at the pleasure of the director”, although often for a very long time, whereas NIE staff were on short-term contracts. As these contracts expired, the jobs became rewards for people who had worked in the Reagan campaign.
The Office of Naval Research
I escaped shortly before my contract expired by moving to a research management position in the U.S. Office of Naval Research (ONR) Personnel (that is, testing) and Training Research Program, later renamed Cognitive Science. I had known and very much admired the research program. The ONR’s style of research management is very different from that of agencies like NSF or NIH. Each year, the ONR publishes an announcement of interest in science and engineering. Projects typically begin with negotiative discussions to establish mutual interest; by the time a full proposal is submitted, the probability of funding is very high. Individual program officers have real power of decision but are held responsible for producing usable results from basic research in a reasonable amount of time, perhaps 30 years, and further expected to promote the movement of those results into more applied research and even into practical application. There are many ways to manage research programs effectively. Even within the NSF, for example, different programs operate very differently. A strength of research funding in the United States is that there are many different centers of decision and many different ways of making funding decisions so that truly good ideas are likely to find funding somewhere.
Over the years, I refined the broad but spotty coverage of the ONR program to emphasize what made ONR unique in the federal funding scene: computational theories of human cognitive architecture and artificially intelligent tutoring systems (ICAI). I supported several different approaches to computational cognitive architecture and am proud of having brought them to a state of useful maturity. Spending special program money that one of my predecessors had won, I brought ICAI from being a basic research enterprise in AI to the beginnings of application.
As the special program ended, Navy Systems Commands began to contract for the development of practical ICAI. At the basic research level I turned to the problem of true natural language interaction for ICAI, although an external advisory committee considered it too difficult to attempt. This became a very exciting cognitive science research enterprise. Discoveries came rapidly, as for the first time, multidisciplinary teams focused on human tutorial interaction. Martha Evens, a computational linguist, produced the first usable tutor with natural language interaction capability — teaching medical students cardiac physiology (Evens & Michael, 2005). These were proud accomplishments of my ONR years.
More details of research progress over the years of my career can be found in Chipman, 2010.
A very important and challenging aspect of my job, largely invisible to the general research community, was maintaining, increasing, or replacing program funds — by making mega proposals in internal competitions (financed by shaving funds out of the apparently stable long-term funding lines), seeking Department of Defense-level funds for large Multidisciplinary University Research Initiatives (MURI), or competing to issue SBIR topics that created opportunities for the few small businesses with research capabilities in psychology and cognitive science. A significant fraction of these proposals failed, but some succeeded. One was an early neurocognitive program (before fMRI), Neural Constraints on Cognitive Architecture. MURI projects I sold included Tutorial Dialog and Automated Skills for Cognitive Readiness.
There were very positive aspects to my research management job: Working very closely with outstanding, innovative researchers and having an opportunity to shape the direction of cognitive science. There were also negative aspects: Little control over one’s time, the need to respond instantly to often critical threats, and almost unlimited work demands. The role of Defender of the Faith — in psychological and cognitive science research — could also be very stressful in an agency dominated by physical scientists and engineers who made all the important funding decisions. Eventually, these factors led me to retire. I am now trying to define a more relaxed intellectual life for myself — writing, reviewing and perhaps even developing proposals. My husband and I now have a second home in Boulder, CO, where I enjoy being a peripheral participant in the life of the Department of Psychology and Neuroscience, which now has many women professors.
References and Further Reading:
Shepard, R. and Chipman, S. (1970) Second-order isomorphism of internal representation: Shapes of states, Cognitive Psychology, 1, 1–17.
Chipman, S. Complexity and structure in visual patterns. (1977) Journal of Experimental Psychology: General, 106, 269–301.
Chipman, S. and Mendelson, M. (1979) Influence of six types of visual structure on complexity judgments in children and adults. Journal of Experimental Psychology: Human Perception and Performance, 5, 365–378.
Chipman, S., and Segal, J., and Glaser, R. (1985) Thinking and Learning Skills, Volume II: Research and Open Questions, Erlbaum.
Chipman, S. (2005) Research on the Women and Mathematics Issue: A personal case history. In: A.M. Gallagher and J.C. Kaufman (Eds.) Gender Differences in Mathematics: An Integrative Approach. New York, NY: Cambridge University Press, 1–24.
Chipman, S. (2009) Expertise in the Management of People: A New Frontier for Research on Expert Performance. In K.A. Ericsson (Ed.), Development of Professional Expertise. New York: NY, Cambridge University Press. 470–493.
Chipman, S. (2010) Applications in Education and Training: A Force Behind the Development of Cognitive Science, Topics in Cognitive Science, 2, 1–12.
Evens, M. and Michael, J. (2005) One-on-One Tutoring by Humans and Computers. Mahwah, NJ: Erlbaum.
Leave a comment below and continue the conversation.