09/27/95
CONTACT: Stanford University News Service (650) 723-2558
STANFORD -- Stanford psychologist Roger N. Shepard has been awarded the National Medal of Science for research that has provided major new insights into the nature of mental processes that previously had been considered impossible to study. In particular, his studies have fundamentally altered scientific and popular views of the nature of mental imagery and have had a profound impact on the fields of psychology, philosophy, computer science, linguistics and neuroscience.
Shepard is one of eight distinguished scientists who will receive the medal from President Clinton in a White House ceremony on Oct. 18. More than 320 medals have been awarded since President Kennedy named the first recipient in 1962. Shepard is the 11th Stanford faculty member to be so honored.
Shepard's theories "allow you to understand the underlying geometric structures of thought to which the individual himself doesn't have access. His research has opened up the mind in the way that a surgeon might open up the brain," said Michael Kubovy, a professor of psychology at the University of Virginia who also studies mental imagery.
Shepard received his bachelor's degree in 1951 from Stanford University and his doctoral degree in experimental psychology from Yale in 1955. After a year of postdoctoral research at the Naval Research Laboratory and two years as a research associate at Harvard University, he became a member of the technical staff of Bell Telephone Laboratories, where he stayed until 1966. From Bell Labs Shepard joined the faculty at Harvard as a full professor. Two years later he moved to Stanford, where he subsequently was named the Ray Lyman Wilbur Professor of Social Science.
Joseph L. Young, the National Science Foundation program director who has overseen Shepard's NSF grants for more than 19 years, said that his research has both theoretical and practical implications: "Shepard's elegant mathematical theorizing helps us understand how our mind works. This understanding will allow us to organize control rooms, cockpit displays and even educational programs to be compatible with the way we think."
Shepard's basic research laid the theoretical groundwork for a number of significant applications developed by others, including enhancement of radiologists' ability to diagnose breast cancer and prediction of the performance of prospective airplane pilots.
In his doctoral dissertation at Yale, Shepard tackled a fundamental problem in human and animal learning called generalization. Generalization is the process of responding to a new situation based on past experience. As animals and humans go through life, it is highly unlikely that they will ever encounter the same total sensory input twice. So the ability to generalize keeps organisms from being hopelessly confused by the welter of sensory stimuli that they are continually receiving. The mind can do this only by assessing the similarity of each new sensory pattern to previous sensory patterns.
While working at Bell Labs, Shepard showed that it was possible to find objective measures of the similarity between types of stimuli, such as different colors. Previously, observers assessed such similarities subjectively. In a series of experiments, he asked people whether one pair of stimuli was more or less similar than another pair. He then used a computer method that he developed to convert these qualitative judgments into unique quantitative "distances" between the stimuli when organized into a special "representational" space.
When Shepard applied this new method to spectrally pure colors, for example, the computer automatically yielded a "color circle." The shape represents color in psychological, not physical, terms. When arranged according to physical wavelength, visible light falls along a straight line that moves from red to violet. Psychologically, however, people perceive red and violet as more similar to each other than either color is to an intermediate color such as green. For this reason, when spectral colors are represented according to perceived hue they form a circle.
This computer method, as perfected by Shepard and his Bell Labs colleague Joseph Kruskal, now is known as "non-metric multidimensional scaling."
Shepard then used this technique to discover a universal law of generalization. As the distance between stimuli, like two colors, increases in representational space, the degree of similarity between the two decreases. It does so following a particular curve, which decreases rapidly at first and then drops off more and more slowly. (Mathematically, this is called an exponential decay curve.) Moreover, Shepard found that this same curve emerges for all stimuli, whether simple or complex; visual, auditory or olfactory. This curve also held for all animals that were tested, including humans, monkeys, rats and pigeons.
In 1987, 300 years after Newton published his Principia, which contained the first formulation of the universal laws of gravitation, Shepard published a paper titled "Toward a Universal Law of Generalization for Psychological Science." He argued that the relation described above represents an optimal learning strategy for beings that live in our kind of universe and, therefore, should hold for all sentient life forms.
"Multidimensional scaling has become one of the most widely used analytic techniques in a number of scientific and applied disciplines," said Carol L. Krumhansl, professor of psychology at Cornell University and one of Shepard's former students.
Robert Nosofsky, professor of psychology at Indiana University, has applied Shepard's work on generalization and multidimensional scaling to the field of learning. "Shepard's impact has been monumental. It's difficult to summarize his contributions and accomplishments as they are so diverse and far-reaching," Nosofsky said.
A classification scheme related to that developed by Shepard and Nosofsky and a variant of multidimensional scaling developed by Shepard's Bell Labs coworker J. Douglas Carroll, now at Rutgers University, have been adapted for the diagnosis of cancer. The adaptation was done by David J. Getty and John A. Swets at Bolt, Baranek and Newman Laboratories in Cambridge, Mass., together with a number of medical coworkers. They have demonstrated that the method yields significant improvements in the X-ray detection of breast cancer and in the use of magnetic resonance imaging to determine the operability of prostate cancer.
At Stanford, Shepard began working on another fundamental psychological problem. Just before waking one morning he had a spontaneous vision of three-dimensional shapes rotating rigidly in space. This vision sparked a new line of research that revolutionized thinking about the basic nature of mental imagery. At the time, the field of cognitive psychology was dominated by theories of artificial intelligence that assumed all thinking was based on the mental manipulation of discrete mental symbols. But Shepard thought it likely that some significant mental processes were more like continuous simulations of external events.
Shepard and his students -- initially Jacqueline Metzler and, most extensively, Lynn A. Cooper -- embarked on a series of experimental studies of "mental rotation." They showed pairs of differently oriented objects to people and asked them whether they were the same or different in shape. They found that the time it took to answer this question increased directly with the difference in the orientation of the two objects. This strongly suggested that people made such comparisons by mentally rotating one of the two objects into the same position as the other. What is more, under the conditions used in the first experiment, they did so at a constant rate of about 60 degrees per second.
"When these findings came out in the early '70s, they electrified the cognitive community," Kubovsky said. "Previously, the notion of imagery being analyzable by methods that are so rigorous was simply unimaginable."
In 1984 Shepard and Cooper, now professor of psychology at Columbia University, published a paper in Scientific American summarizing this research, "Turning Something Over in the Mind." After its publication, they received many letters from lay people who said that the article described the kind of mental process that they used themselves. "One of the most interesting ones came from the president of the Canadian snooker society saying that this is exactly what snooker players have to do to anticipate the trajectory of balls when they are hit," Cooper said.
Increased understanding of these sorts of mental manipulations has proven to be of considerable practical importance as well. For example, Thomas Carretta, who works for the Human Resources Directorate of the U.S. Air Force, had students who were entering undergraduate pilot training take a mental rotation test. Those who made quick, consistent and accurate responses to a mental rotation test were more likely to receive recommendations for fast-jet training after completing their instruction, he reported.
In his career, Shepard has received a number of awards and honors. He is a fellow of the American Association for the Advancement of Science and the American Academy of Arts and Sciences, and is the William James Fellow of the American Psychological Association. In 1977 he was elected to the National Academy of Sciences. He has received awards for scientific achievement from the American Psychological Association, the Society of Experimental Psychologists and the New York Academy of Sciences, and an honorary doctorate from Rutgers University. Shepard's most recent book, Mind Sights, illustrated with his own hand- drawn visual illusions, now has been published in five languages. Last fall he delivered the prestigious William James Lectures at Harvard University.
-dfs-
950927Arc5087.html
This is an archived release.
This release is not available in any other form.
Images mentioned in this release are not available online.
Stanford News Service has an extensive library of images,
some of which may be available to you online.
Direct your request by EMail to newslibrary@stanford.edu.