Science and the Spectrum of Belief, An Interview with Leonard Ornstein, by Peter J. Denning

Abstract

Editor's Introduction: I recently discovered that cell biologist Leonard Ornstein had, in 1965, written a long and thoughtful essay on information and meaning. Shannon’s idea that communication systems could transmit and process information without regard to its meaning just did not seem right to him. He was particularly interested in how scientists use and interpret information as part of science. He devised a simulation of a discovery process that consisted of sifting through data looking for recognizable classes and finally classifying each individual item into one of the classes. His process was obviously implementable by a computer, although the computers of the day had insufficient capacity to take on large data sets such as those found in his field. I was struck by the similarity with the modern Bayesian “autoclass” programs, which were designed much later and automated the inference process described by Ornstein. I thought it would be fascinating to talk to him and find out more about how he sees science, discovery, information, and meaning. The result is this interview. bBorn in 1926, Leonard Ornstein has had a long career covering cell biology, cytochemistry (first high-resolution methods for esterases and phosphatases), flow-cytometry, electrophoresis, bioengineering, biophysics, electro-optics, optical and electron microscopy (first images of internal structure of mitochondria, cilia and flagella, pores of nuclear membranes, epithelial brush borders, myoneural junctions, electroplax), unsupervised learning, information theory and meaning, pattern recognition and artificial neural networks, automated medical diagnosis, epistemology, agricultural irrigation, and global warming. He received his Ph.D. from Columbia University in 1957 and taught and performed research in their Zoology Department from 1949 until 1964. He joined Mt Sinai Hospital in 1954 and, in 1966, became a professor of pathology at Mt. Sinai School of Medicine. He retired from Mt. Sinai in 1992 and has remained an emeritus professor to the present. He has consulted for numerous companies on medical technology and instrumentation and has been a principal in three businesses (two terminated in 2003, and the last was sold in 2006). He is best known for pioneering a technique called polyacrylamide gel electrophoresis, for the analysis of proteins and nucleic acids, in the 1950s. He has been married for 68 years and has four children, nine grandchildren, and six great grandchildren. (Peter J. Denning, Editor-in-Chief)