JSTOR

A View of the Future as Seen from the Past1

Received and accepted 2000 January 28

When the editors of these Publications asked me to write a Millennium Essay on how astronomy or a specific area of it has evolved, and where we might expect it to lead us in the future, I first declined. I felt that my track record on forecasting political elections has been so poor in the past that I did not want to take a chance on doing even worse in hard science. However, they persevered, telling me that predictions were not required, and I then agreed to write on a related topic, how observational ground‐based astrophysics has changed in America in my lifetime. Naturally, most of us believe that the future is an extrapolation of the past, so I have tried to include a little in that vein, but I caution my readers to keep my previous failures in mind. I firmly believed that Adlai E. Stevenson would be elected president of the United States in 1952 and in 1956, rather than Dwight D. Eisenhower, who was.

Certainly no one could expect an essay to review astronomy over the entire current millennium and to predict what the next one will bring. We know almost nothing about our science a thousand years ago, except that there was very little of it, much closer to astrology than to what we call astronomy today, and that there certainly was no astrophysics at all back then. Projecting into the future, should we not expect that our descendants 30 generations from now may have a completely different, and better, way of understanding the universe than we do now? It is hard, if not impossible, for us to imagine that, but it was equally hard for Merlin, casting a horoscope in Camelot, or a Native American shaman, scratching a crescent moon on the wall of a cave in the American Southwest, to imagine us.

A century is a much more manageable period to look back. I was born in the first quarter of this century; my parents were born in the 19th century. I learned about the planets, stars, and constellations as a Boy Scout, then learned a little about their nature in a high‐school physics course, and then a lot more when I read several excellent books in the school’s library. My parents gave me even more books, and a small, secondhand reflecting telescope on a pipe mounting, which provided awful but recognizable images of the Moon, planets, nebulae, and M31. At the University of Chicago I received an excellent education in physics and astronomy, on campus for 3 years and at Yerkes Observatory for 3 more. My teachers included Thornton Page (an excellent, challenging, one‐on‐one teacher, Page on one end of a log and Don Osterbrock on the other), Enrico Fermi and Gregor Wentzel in Chicago, and Otto Struve, Bengt Strömgren, Subrahmanyan Chandrasekhar, Bill Morgan, and Gerard Kuiper in Williams Bay. I was there during the post–World War II education boom. In those dear old days there were at most 10 or 12 graduate students at Yerkes, frequently fewer. We all got excellent, hands‐on training for research, with ready access to some of the great scientists of that time. We lived, ate, and slept astronomy at Van Biesbroeck’s boarding house, managed by the wife of a retired longtime Yerkes professor.

Probably boys a century ago learned about constellations and stars much as I did. Whatever astronomy they had in school dealt with the phases of the Moon and planets and a few simplified orbital properties. In colleges of the 1840s and ’50s, a select few students, nearly all of them men, heard lectures and studied textbooks, mostly concerning orbits. They memorized as much as they could, because they were expected to respond with the “right” answer when questioned by an instructor in class. There were no graduate schools or even graduate students in astronomy in America in 1850. By 1900 graduate training existed, frequently in the form of one recent graduate staying as an assistant to the professor with whom he continued to study. A high proportion of the few would‐be professional astronomers, like James E. Keeler, Edwin B. Frost, Walter S. Adams, and Frederick H. Seares, went abroad—chiefly to Germany—for a few years of postgraduate study in Berlin, Munich, or Göttingen. Most of them did not earn doctoral degrees, but returned to America as soon as they learned enough to get a job at home. By the 1900s the University of California (Berkeley and Lick Observatory) and the University of Chicago (there and at Yerkes) had several graduate students at a time, most of them aiming for a Ph.D.

Our training at Yerkes consisted mostly in learning by doing. We all took three graduate courses each quarter, the only ones given except research, and were expected to finish the cycle of 18 courses in 2 years (none were given in the summer, but visiting scientists, frequently from abroad, often gave shorter series of lectures which we were also expected to attend if we were there). Each course met one afternoon a week—Tuesday, Wednesday, or Thursday, leaving Monday for the colloquium and Friday for a follow‐up by any professor who had not had enough time on his regular afternoon to say all that he had planned. I learned to integrate differential equations numerically with a hand‐operated, electrically powered Marchant calculator by working Strömgren’s homework problems and how to classify spectra on the MKK system by doing Morgan’s assigned unknowns.

Nearly all of us learned to observe with the 40 inch refractor, still a useful tool for astrophysical research in the 1940s and ’50s. I worked with Morgan’s spectrograph right out in the open dome, wearing all the heavy clothes I had and doing my best to keep warm on the long winter nights by standing (or jumping up and down to keep the blood flowing in my feet) on an electrically heated mat. We set the telescope on a star by hand, reading the coordinates from the huge right ascension and declination dials far above the floor with a flashlight, and it helped to be reasonably strong or heavy to overcome its inertia. We guided, straining our eyes on faint stars, using a hand paddle to control the telescope’s electrical slow‐motion motors. We took our data on blue‐sensitive, fine‐grain photographic plates, which had low quantum efficiency, a nonlinear response, and no way to read them out except by developing them ourselves in the darkroom, after the exposure had been completed. All the glass in the telescope objective and the spectrograph cut out any violet and ultraviolet light which had gotten through the atmosphere and resulted in spectrograms covering the “blue” wavelength range from 3900 to 5000 Å.

Contrast that way of observing with now. I am retired and no longer observe, but my colleagues on the active faculty do so on much larger and more advanced telescopes, the 3 m Shane reflector (which I used until 1993 January 1), the 10 m Keck telescope, various large national and international telescopes, and the Hubble Space Telescope. They sit in a comfortable heated room, near the telescope at Mount Hamilton or at the CARA headquarters in Waimea, Hawaii, aided by expert technicians, with high‐precision telescope‐setting programs and automatic guiding. They probably have a television set to while away the time, as they need glance at the image of the star or galaxy on the automatic guider only occasionally. Their CCDs have much higher quantum efficiencies than photographic plates, and their sensitivity extends from the atmospheric cutoff around 3500 Å to 9000 Å or more. The data astronomers obtain now is all digital, and they have powerful computers which collect the data, sum them, and display the observed spectrum as it builds up. Largely automatic programs store, reduce, and analyze the data quickly. It is a whole different world from that of my student days, and astronomers today can get much better data than we did back then. They are using much larger telescopes more effectively at better sites, and the results show it.

Let me now try to extrapolate these trends into the future, not for a millennium, but perhaps for a century. But always keep in mind that predicting, while not too difficult for the next year or two, is much more difficult for the next century. The main problem is that there may be so many unexpected differences in what our descendants do, how they live, and what they have at their disposal from our situation today, that we are all too likely to be completely wrong in forecasting their discoveries. A good example is Simon Newcomb (1835–1909), the outstanding American astronomer of his time. He was also a mathematician, an economist, a lecturer, a writer, a leader of and spokesman for American science, and “an officious busybody,” as one of his contemporaries called him in a memorial biography which cooler heads decided was too hot to be published in the good, gray Biographical Memoirs of the National Academy of Sciences. Around the final years of the last century, which Newcomb and all the other astronomers agreed ended on 1900 December 31, just as we know the present millennium will end on 2000 December 31, he wrote and published two separate popular articles in McClure’s Magazine, predicting limited aspects of the future, by implication for the 20th century. One of these articles was “The Unsolved Problems of Astronomy” (Newcomb 1899), the other, “Is the Airship Coming?” (Newcomb 1901).

Newcomb was a very intelligent scientist, and everything he wrote about the astronomy of his time was true. The big problems he outlined included some but not all of those which we and our predecessors have worked on during this century. He was much better in his predictions for orbital theory, his specialty, than for astrophysics, then very new and almost a closed book to him. He also did reasonably well on cosmology, another subject on which he had thought long and hard. But he was limited, as we all are, by his frame of reference.

One of the problems he despaired of solving was whether any of the other planets in our solar system could be the abode of life like ours. Even “the most powerful telescopes we can ever hope to make” could never show mountains or lakes, rivers, or fields, much less any “works of man,” even at the distance of Mars or Venus when they are closest to the Earth, he wrote. But Newcomb could not imagine that within less than a hundred years his descendants would be able to send space vehicles with relatively small cameras close enough to Mars to photograph its mountains, “seas,” valleys, rills, and other features in great detail.

In his second article Newcomb described the two great problems of science as the nature of the “luminiferous ether,” for which we might read “light,” and the “cause” of gravitation. He wrote that he was not certain, but he thought that these questions might be answered in time; if they were, he speculated, their answers might be closely related. He was right, and Albert Einstein’s general theory of relativity, published less than two decades later, fitted his prediction very well indeed. On the other hand, although Newcomb did not quite conclude that “airships” were not coming, he was extremely negative on the possibility that flying machines which could carry humans would be invented. Again, his frame of reference was faulty; his article shows that he was thinking in terms of birdlike machines with beating wings and did not foresee fixed‐wing airplanes with powerful, motor‐driven propellers. Newcomb’s examples have taught me to be extremely wary of any predictions of what will “never” be done.

I think it is safe to assume that astronomy will remain an observational science, as it has been from the beginning. Planets, comets, planetary nebulae, variable stars, novae, supernovae, white dwarfs, pulsars, H ii regions, galaxies, quasars, clusters of galaxies, and gamma‐ray bursters were first discovered by observers. Some of these objects’ properties were measured, if only crudely, and then theorists began analyzing and explaining them, providing insights for further measurements and searches. Just as optical telescopes’ apertures grew by a factor of 10 in diameter from the Yerkes 40 inch in 1897 to the Keck I 10 m in 1994, I feel completely safe in predicting that they will grow by another factor of 10, to a 100 m telescope in 2100, if not earlier. I know that at least one group is now planning a 30 m instrument, and though its size may be scaled down before it is built, that will surely be only a temporary slowdown.

It seems to me very probable that by the end of the next century the largest optical telescopes will be erected on the Moon rather than on the Earth, to get away from light pollution as much as to escape atmospheric absorption and seeing. Probably the largest observatories will be in the lunar hemisphere opposite the familiar one we see, so that the astronomers of 2100 at least will not have to contend with the full Earth blocking part of their field of view and spilling bright light into any instruments that have not been adequately shielded and baffled. I imagine that telescopes will be operated remotely from the Earth, with commands and data relayed back and forth via artificial satellites near the Moon, but perhaps a colony of telescope, instrument, and data‐handling specialists will be stationed at a lunar base, each spending several months or a few years there before returning home to Earth. If the observational teams are not there, surely they will operate the telescope from an astronomy center on Earth. Research astronomers will submit programs but will have even less to do with the actual observations than at present; they will receive highly processed data obtained with instruments much more powerful than any we have today, built around detectors with high quantum efficiencies in all wavelength or energy bands.

Observational research will be done by relatively few large teams, which will include theorists, “observers,” image‐ and spectra‐simulators, long‐range planners, gofers, librarians (of huge electronic storage systems), statistical experts, and no doubt several other types of specialists in fields we do not even know of today, any more than Newcomb knew of aviators or Merlin knew of railroad engineers.

Looking further and further out into the universe, or back in time, requires bigger, better, and hence more expensive telescopes and instruments. They must be used extremely effectively to justify their cost; this means that large groups of experts must be assembled to work together, each doing one task or operation as efficiently and productively as it can be done. This trend has already gone further in physics than in astronomy, but we are close behind. In my youth I could not conceive of a paper with five or more authors; today there are only rare papers with one, two, or three authors. I am glad I entered astronomy when a single grad student, or a single postdoc, or a single junior professor, or a senior professor with a postdoc and two or three graduate students could make real contributions to knowledge, based on data they had taken, reduced, analyzed, discussed, and understood themselves. However, those days are over, and though I regret it at a personal level, I understand and applaud it as the way to do scientific research today. All those trends will surely continue and grow stronger in the coming century, and our descendants will understand the universe better in 2100 than we do today, because they will build better equipment and use it more effectively than we can now.

How many astronomers will there be by then, and how will they be trained? It is plausible to assume that the present structure of graduate schools will persist; faculty members will still want grad students to learn everything they learned themselves a generation earlier, plus all the knowledge that research has added since then. There will be more professors, more students, more courses, longer stays in graduate schools, and more years spent as a postdoc on the path to a “real” academic or research position. Astronomy is an expensive luxury which only wealthy countries can afford, but America is rich and its productivity is continually growing. I think our country will continue to fund astronomical research, which is much the most interesting, attractive, and understandable science to the general public. The number of astronomers will continue to increase.

As a graph published by David DeVorkin and Paul Routly (1999) shows, the number of professional astronomers, measured reasonably well by the membership of the American Astronomical Society, increased from 113 charter members in 1899 to about 650 half a century later. Joel Stebbins, then the grand old man of the Society, projected linear growth of its membership at about 10 per year from those two numbers, which would have led to about 1150 members in 1999 (Stebbins 1947). In reality the growth was exponential, not linear, and the doubling time shortened rapidly around 1957, a consequence of the post‐Sputnik boom in astronomy and space, so the actual number of professional astronomers in 1999 was about 6500. If we project the overall growth for the next century by the same factor as the past one, we can expect astronomers by 2100! Perhaps there will be some saturation, decreasing the rate, as about half the population now enters college by age 20, but many drop out, students will stay in school and in college longer, and researchers will live and work longer, if perhaps less productively. Even if we assume the slow exponential growth rate of the first half‐century of the Society’s existence will return for the next hundred years, there will be about astronomers by 2100. In the words of Robert Browning, “Come, grow old along with me, the best is yet to be.”