My interest in astronomy developed late. It didn't emerge fully until I had completed my Ph.D in astrophysics (but with a thesis devoted wholly to atomic physics) and was a post-doc at the University of Michigan . It was a reading of Otto Struve's book 'Stellar Evolution' that really ignited my interest in astronomy -- specifically in the question of how double stars came into being.
One summer, when I was thirteen or fourteen, I fell out of a tree and broke an arm. To help me through the unaccustomed period of inactivity, my father, a physician, bought me a book on algebra. He had observed that math was the biggest stumbling block for people pursuing careers in engineering and medicine, and he wanted me to get off to a good start. I enjoyed working the problems, and I read only as much of the text as I needed for that purpose. After that I worked through other math books, so that by the time I came to college, just after my seventeenth birthday, I had more or less the equivalent of an undergraduate education in math. Less in that I had paid very little attention to the formal side; more in that I had had more experience in solving problems than most undergraduates.
I found high school science courses dull and unrewarding. I did, however, attend the Open Nights at the Warner and Swasey Observatory in Cleveland. The Observatory's director, Jason J. Nassau, took an interest in me and gave me math books to read. Though he himself was a practical astronomer, he had begun his career in pure mathematics. One day he showed me a preprint by an astronomer named van Albada, on the origin of the chemical elements. I couldn't believe that this was a topic that someone could actually do research on. I couldn't imagine what that kind of research would be like.
At Harvard I concentrated in mathematics but avoided math courses. Dr. Nassau had suggested that I introduce myself to his friend Bart Bok, who immediately signed me up for his graduate- level course in astrophysics. I had virtually no background in physics and none in astronomy, but Dr. Bok assured me that I would have no difficulty with his course. In fact his lectures were way over my head. I decided I'd better learn something about astrophysics, so, for the first of the three required papers that Dr. Bok assigned in place of an exam, I read Goldberg and Aller's book 'Atoms, Stars, and Nebulae.' My paper came back with a grade of B- and a caustic comment to the effect that this wasn't the sort of paper one was expected to write for a graduate course at Harvard. For my next paper I tackled Chandrasekhar's book 'Stellar Structure' and succeeded in simplifying and generalizing proofs of some of the theorems on polytropes.
After a 2 1/2 year stint in the Army, I returned to Harvard. I had read no math or science during that time. Less malleable than I had been as a seventeen-year old freshman, I became increasingly impatient with math courses and (as it seemed to me) their emphasis on form at the expense of substance. I went to see Dr. Bok, who handed me over to Harvard's resident astrophysicist, Donald H. Menzel. Prof. Menzel immediately set me to reading Condon and Shortley's 'Introduction to Atomic Spectra'. Eventually I wrote my thesis on that subject. I was lucky enough to notice a systematic discrepancy between current theories and experimental data on the structure of complex spectra. In my thesis I set out to show that the effect was real and to construct a theoretical explanation for it. Henry Norris Russell, the grand old man of astronomy and one of the pioneers in the study of atomic spectra, came up from Princeton, where he was living in retirement, to conduct my Ph.D oral examination. He had read and liked the thesis, and I believe it was his voice that persuaded the committee, most of whom had been his students, to pass me. Certainly I knew very little astronomy.
Having been lucky enough to receive a National Research Council postdoctoral fellowship, I chose to go to the Observatory of the University of Michigan, where Leo Goldberg, and early student of Donald Menzel's and the author of he book that had introduced me to astrophysics, was director. During the year I spent at Ann Arbor I finally succeeded in understanding the systematic discrepancies I had discussed in my thesis. I also discovered astronomy, through Struve's book and its revelation that there was no tenable answer to the question of why more than half the stars in the sky are members of double-star systems. After agonizing for some time about this question, I decided that binaries could have formed only when the Universe was far denser than it is now (or was in 1951!). This idea implied that stars were formed before star clusters and galaxies. Pursuing this line of reasoning, I came up with the hypothesis that all self- gravitating systems were formed by a single repetitive process, which I called gravitational clustering. In gravitational clustering, clusters separate out of an expanding cosmic 'gas' of gravitating particles. These clusters themselves then act as 'molecules' in a new version of the expanding 'gas' of gravitating particles, and as the cosmic expansion continues, clusters of clusters separate out. And so on. I presented this idea at a meeting of the American Astronomical Society in December 1951, and was thrilled when the New York Times carried a short account of it.
After Ann Arbor, Otto Struve, whose book had awakened my interest in cosmology, offered me a temporary teaching position at Berkeley. There I developed courses in stellar atmospheres and the physical foundations of astrophysics, and continued to work on cosmology and atomic physics.
I came to Berkeley as a replacement for Louis Henyey, who was spending the year at Princeton working on Project Matterhorn. When Henyey returned to Berkeley, I was asked to take his place at Princeton. Since this meant working directly for John Wheeler, whose work I knew and admired, I gladly accepted. When Wheeler withdrew from the project eight or nine months later, I accepted Donald Menzel's offer to return to Harvard. And with some short breaks, I have been there ever since.
After spending a sabbatical year in Sweden in 1968-1969, I decided to shift the focus of my teaching from graduate to undergraduate education, and from science students to nonscience students. This second teaching career has been more rewarding than I could have anticipated. And it has also influenced the direction of my research. More and more I have focused on questions that interest nonscientists as well as scientists: Why is the future different from the past? What is the origin of the order we see around us? Is quantum mechanics really nothing more than an ingenious mathematical device for making certain kinds of predictions? My second book, Cosmogenesis, scheduled to be published by Oxford University Press in 1988, deals with these and other questions.
I believe that science advances only through the filling in of minute details and the solution of highly specialized problems that only a few other specialists know or even care about. I enjoy doing that kind of science. But the details are interesting and valuable mainly because of the big picture to which they belong. I feel very fortunate in being able to spend a large part of my working life thinking, writing, and teaching about the big picture.