The dynamic interaction between epistemological (instruments and techniques) concerns and the kinds of problems radar astronomers seek to solve, which we have seen driving planetary radar astronomy to the present, also in all likelihood will continue to determine its future. Both new instruments and techniques will furnish the means for exploring new targets and for resolving problems, especially those left unresolved or unsatisfactorily resolved by optical means.
Planetary radar techniques developed recently perhaps hint at the sources of future techniques. Three examples are John Harmon's non-repeating code, which he adapted from Arecibo ionospheric research; Dewey Muhleman's use of radio astronomy imaging and arraying techniques at the VLA, as part of the bistatic Goldstone-VLA radar; and the planetary imaging technique developed by Scott Hudson and Steve Ostro. The Harmon and Muhleman techniques reflect the continuing, though diminished, influence of ionospheric research and radio astronomy on planetary radar astronomy.
A surprising number of new instruments may be available, too, many through the grouping of either the Goldstone or Arecibo antenna in tandem with a radio telescope to form a bistatic radar. The Goldstone-VLA radar appears to point the way to additional combinations with the soon-to-be-completed Green Bank radio telescope, or perhaps to the Goldstone X-band radar in tandem with a tracking station in the Soviet Union. Already, the Russian Yevpatoriya tracking station has made bistatic observations of the asteroid Toutatis in conjunction with the Effelsberg radio telescope, though without achieving the impressive results of the Arecibo and Goldstone antennas. Politics and funding will limit what, if any, future bistatic experiments take place outside the United States. Additional bistatic possibilities in the United States include the JPL Mars Station in combination with other Goldstone antennas, as well as an Arecibo-Goldstone link.
The bistatic possibilities are not limitless, however; not inconsequential institutional, political, and budgetary obstacles aside, the elementary technological need for compatible transmitting and receiving frequencies limits many bistatic options. Even more limited is the creation of new radars. Other countries continue to build antennas, such as the Arecibo-size dish planned in Brazil, but none anticipate a radar capability. No facility dedicated entirely to planetary radar astronomy ever has been built; nonetheless, Steve Ostro believes that it is time to build one in order to study asteroids. The cost of designing and building such a radar observatory would approach the modest level budgeted for NASA's Discovery space missions.1 The role of this facility in the Spaceguard project aside, its potential scientific value in a short period of time would exceed that of any one Discovery flyby of an asteroid. Time will tell whether this worthwhile and economical project is realized.
 These are all possible future planetary radar instruments. Nothing, especially not their scientific merit, either guarantees or favors their realization; budgets, not science, will determine their viability. With one exception, planetary radar astronomy always has subsisted on the budgetary margins, either by design (as at JPL) or by fate (as at Lincoln Laboratory). As budgets are trimmed, the freedom to fund bistatic experiments from discretionary funds diminishes, too. The only exception is Arecibo, where a five-year contract stabilizes the research budget, although within the shrinking NASA and NSF budgets. If one can say anything about the future of planetary radar astronomy with certainty, it is that the future is at the upgraded Arecibo telescope.
The Arecibo upgrade, as well as the potentially available novel instruments and techniques mentioned above, will bring new research targets within the reach of radar astronomers. Among the most striking new targets visible to the upgraded Arecibo observatory will be the satellites of Jupiter; Iapetus, Rhea, Amalthea, Dione, and Hyperion. The detections of those bodies very well may lead to radar solving new scientific problems. In addition, the upgraded Arecibo telescope will be able to map Jupiter's Galilean moons at much higher resolutions, perhaps down to 100 kilometers, and uncover fresh facts regarding lo and Saturn's moon Titan.2
Planetary radar may contribute as well to our understanding of the terrestrial planets through analysis of their polarization ratios; however, the greatest amount of research activity will be directed toward neither the terrestrial planets nor the systems of Jupiter and Saturn, but the asteroids The space in which Earth turns abounds with thousands of asteroids. The largest, a kilometer or larger in diameter, number about 2,000, while those 100 meters or more in diameter number 150,000 or more, and those 10 meters and larger amount to some 300,000,000. These are estimates of the asteroid population; the number of asteroids actually observed increases continually. The likelihood that one of those asteroids might approach Earth perilously close has heightened interest in them.
In many ways, then, planetary radar astronomy has come full circle. It began with the study of large populations of meteors, and the observation and analysis of the large and varied asteroid population is carrying it into the future. Forty years ago, the forte of radar lay in its ability to determine accurately the radiants and speed of meteors and to ascertain unambiguously that they orbited around the Sun. Today, the value of radar is its ability to fix asteroid orbits with an accuracy and certainty that no other method can match.
Current planetary radar techniques, however, can do much more with asteroids than the earliest radar investigators at Jodrell Bank and the Canadian National Research Council were able to do with meteors using their pioneering techniques and far less sensitive radar equipment. Today's planetary radar astronomers can characterize asteroid composition, size, and shape and can provide a unique imaging ability. The ever growing number of asteroid targets, combined with this wide range of epistemological tools and the present societal interest in a potential "killer asteroid," guarantees that the future of planetary radar astronomy will be asteroid research.
Asteroid literature and funding have grown markedly over the last fifteen years. In the past, similar rapid growth in ionospheric and radio astronomy research carried forward planetary radar astronomy. This growth has not yet reached radar studies of asteroids, however, and despite the expanding observational opportunities created by asteroid studies, the number of radar astronomers probably will not increase significantly. Steve  Ostro remains the sole full-time asteroid radar astronomer. Once the upgraded Arecibo radar becomes available, the number of radar investigators studying asteroids probably will increase, or rather, must increase, if an adequate number of observational opportunities are to be seized. Already, the Arecibo Observatory has hired a planetary astronomer with an interest in asteroids3 who will also take part in radar observations of asteroids. Don Campbell will participate in those observations, as too may Dick Simpson of Stanford. The growth in asteroid science, then, may shift current radar researchers into the field, rather than provide a basis for expanding planetary radar astronomy.
As a scientific species, planetary radar astronomers have tended not to reproduce themselves. Hiring individuals from other fields yielded planetary radar astronomers in the 1960s and 1970s, but none in the last 15 years with the exceptions of Marty Slade at JPL and the recent hire at Arecibo. The number of planetary radar astronomers created through paid employment, therefore, may remain small and relatively stable. Being small yet may have its advantages in a future certain to be shaped by budget cuts in NASA and U.S. scientific research in general.
The other traditional career path into radar astronomy, university training, may furnish fresh practitioners, though. Gordon Pettengill at MIT and Don Campbell at Cornell directed many radar astronomy dissertations, although certainly not all of those students entered the field. The MIT-Cornell axis has supplied planetary radar astronomers since the 1960s, but the last Ph.D. to enter the field through that route (Steve Ostro) graduated in 1978. Moreover, with the retirement of Pettengill at MIT and the approaching retirement of Campbell at Cornell, who will train future planetary radar astronomers at Arecibo?
Outside of MIT and Cornell, only Caltech appears equipped or willing to train them. There, Dewey Muhleman has graduated one student, Bryan Butler, in 1994, who did doctoral research in radar astronomy. Although interested in pursuing radar research, Butler is at least equally excited by the prospect of planetary radio studies at the VLA, where he has taken a position. Muhleman is not interested in training additional radar astronomers.
In contrast, Steve Ostro at JPL teaches a course on radar astronomy at Caltech. That position gives him the ability to both recruit and train future radar astronomers. The key to training future radar astronomers in an academic setting like MIT or Cornell remains the master-disciple relationship. Replacing Pettengill and Campbell, then, is Ostro, who is in a unique position to carry the MIT-Cornell alliance one step further by linking Caltech, JPL, and Goldstone to it.
Ostro, a graduate of MIT who conducted his doctoral research on Cornell's Arecibo instrument, and a former member of the Cornell faculty, found Ray Jurgens and Marty Slade, graduates of the Cornell and MIT programs, respectively, when he began work at JPL. Ostro's arrival at JPL signalled a joining of the JPL and MlT-Cornell research groups. His opportunity to teach at Caltech and recruit radar astronomers, coming near the retirements of Pettengill and Campbell, assures the continuation of the master-disciple relationship as the source of future radar astronomers, but within a larger institutional (MIT, Cornell, Caltech-JPL) and instrumental complex that joins the Arecibo and Goldstone radars. The centering of Ostro within that complex also positions him to direct the future of radar astronomy.
Like the number of practitioners, the radar astronomy literature will remain at a low level as a result of both the small number of researchers and the nature of the science reported in those publications. The discoveries to be made on the terrestrial planets and the moons of Jupiter and Saturn will not generate a substantial number of articles, because those discoveries likely will not merit that level of scientific attention. The results  of asteroid research, moreover, will be described in articles that discuss the characteristics of a substantial population of asteroids and not the properties of just one or two asteroids. Consequently, the number of asteroid-related publications will remain limited.
Indeed, virtually the entire history of planetary radar astronomy has been one of limits. The number of practitioners has been limited, if not declining. As we saw, after the initial "explosion" of planetary radar activity in the early 1960s, as measured by the number of experimenters, publications, and instruments, the field of planetary radar astronomy assumed the manpower and publication dimensions of Little Science. After further shrinking during the 1970s, leaving a handful of researchers utilizing a single radar instrument in 1980, planetary radar astronomy stabilized at this lower level (the Arecibo radar being down for the duration of the upgrade). The circumscribed number of opportunities to train future radar practitioners in academia, as well as retirements (most current practitioners are at or near retirement age), will keep manpower levels low.
The practice of planetary radar astronomy as Little Science in the instrumental and institutional context of Big Science will likely continue into the future. A number of factors integral to the field have confined planetary radar astronomy to its existence as Little Science. To begin with, the field generally has operated at the limits of the technology (the instrument hardware). As soon as an instrument became available, radar astronomers sought to discover what new targets it could detect. Once the farthest target was reached, and the spatial limits of research defined, radar astronomers had insufficient sensitivity to achieve more than a detection. Imaging planetary surfaces always involved pushing the instrument's signal-to-noise ratio and resolution capability to the limit. These restrictions in turn prompted radar astronomers to continually press for hardware modifications that provided incrementaI increases in sensitivity. In the end, though, what could be done was limited by the capability of the instrument.
Another growth-limiting factor inherent in planetary radar astronomy is the availability of targets, a factor intimately linked to instrument capability. The planers and their moons cannot be detected unless they are within radar range. The sensitivity limits of planetary radars, such as the Arecibo telescope, prevent investigators from observing targets except when they approach Earth. At other points in their orbits around the Sun, they are too far away for radars to detect them. Thus, Venus is observed at inferior conjunction and Mars at opposition.
A related problem is that of declination. Although most planets rotate around the Sun more or less in the same plane, called the ecliptic, they are not visible in the sky at all times because of the Earth's motion about its own axis. A further complicating factor is the ability of the radar antenna to "see" a portion or all of the visible sky, that is, the so-called declination window of the antenna. The declination window of the Goldstone DSS-14 dish runs from 40° South to 80° North, while the Arecibo telescope is limited to solar system objects that pass within the far narrower band from 40° North to just below the equator.4
The combination of declination window and radar sensitivity restricts observational opportunities, so that planetary research demands only about five percent of total antenna time. The finite number of planetary targets and the narrow observational windows also tend to limit the number of radar researchers. Thus, the tendency at Arecibo was to establish a given target as the terrain or turf of a particular researcher. The number of radar investigators at JPL was always too small for such a division of targets, although  during the 1970s Ray Jurgens "specialized" in Venus and asteroids and George Downs in Mars, with Dick Goldstein continuing to do a little of everything.
The considerable and expanding number of known asteroids is too large for a single investigator. This will be especially true in the near future, once asteroid detection relies on CCD imaging and the Arecibo upgrade reaches completion. Then joining Ostro in radar observations of asteroids (many asteroid scientists hope) will be virtually all members of the small club of radar astronomy practitioners. Again, the expanded program of asteroid radar research that will take place throughout the remainder of this decade will not lead to a transformation of planetary radar astronomy into Big Science.
Planetary radar astronomy has been and likely will remain Little Science embedded in the matrix of Big Science. John Krige's study of British nuclear physics research in the period right after World War II provides a different case of Little Science being conducted with Big Science instruments. One can find another parallel example in the telegraph networks of the nineteenth-century United States.
The Western Union telegraph company, formed by the merger of several separate companies, absorbed both of its principal rivals in 1866 to become one of the nation's largest companies and thereby created the largest electrical communication network in the world. While not Big Science, this was Big Business and Big Technology. By the very nature of their position in the company, telegraph operators had access to the large-scale technological laboratory formed by the telegraph lines. Just as access to technology led to the emergence of planetary radar astronomy, so access to the telegraph network led these operators to perform electrical experiments on the lines. Out of those experiments came numerous inventions, many of which were patented.5
It is not going too far to draw this parallel between Little Science (planetary radar astronomy) and Little Technology (telegraph inventors), for several reasons. For one, most radar astronomers were trained as electrical engineers, not scientists. Also, radar astronomy was a science driven by technology, namely, the availability of radars capable of planetary exploration. It was through these instruments and their associated techniques of analysis, not through direct sensory observation, that radar astronomers conducted their experiments. They analyzed not sensory experience, but wave patterns of electromagnetic signals which analysis by computer software made "visible." Thus, not only were the instrumentation and techniques of radar astronomy dependent on technology, but so was the very content of the science.
Planetary radar astronomy historically has remained at the intersection of science and engineering. Attendance of radar astronomers at both IAU and URSI meetings during the 1960s reflected the dichotomous nature of radar astronomy, perched between radio engineering (URSI) and astronomical science (IAU). The dichotomy arose from the fact that radar astronomy is a set of techniques (engineering) used to generate data whose interpretation yields answers to scientific (planetary astronomy and geology) questions. Also as a result of this dichotomy, planetary radar astronomy concerns itself with two different but related sets of problems (in the Kuhnian sense discussed in Chapter Five). One set of problems is epistemological, that is, it deals with how radar astronomers know what they know and relates to the radar characteristics of the planets, such as surface scattering mechanisms, dielectric constants, and radar albedos; these problems arise out of  the engineering side of radar astronomy. A second set of problems, such as planetary orbits and spin rates, arises out of the science side of the field.
The rooting of Little Science (of Little Technology) within large technological systems, such as the Western Union telegraph network or the Deep Space Network, suggests that it may be in the nature of large-scale "technosocial networks" or "systems" (to borrow the terminology of the social construction of technology mentioned in the Introduction) to sustain Little Science (or Little Technology). Large technological systems form a unified set of relations among individuals, objects, and ideas. As tightly "constructed" as these technosocial networks may be, the magnitude of the resources they encompass is of a sufficient extent to allow small-scale entrepreneurs (be they scientists, engineers, inventors) within the system to "socially construct" smaller technosocial networks within the larger.
Without the larger technosocial network, then, the smaller network is unthinkable. Planetary radar astronomy simply would not have existed without the enormous, powerful, highly sensitive radars on which the experiments were conducted and which were called into existence by the demands of the Cold War and Big Science. Another requisite, of course, was the radar experimenters themselves. The linking of research groups at MIT (Lincoln Laboratory), Cornell University, and (most recently) JPL (Caltech) has provided a means by which the Little Science planted in the interstices of large technological systems can perpetuate itself despite declining resources and limits to growth. For example, as planetary radar activity ceased at Haystack, it continued at the Arecibo Observatory. Given the symbiotic relationship between Big Science and the Little Science which depends on it, as well as the nature of that dependency, funding cutbacks intended to reduce Big Science also will diminish, or perhaps even eliminate, Little Science. Future research will have to determine how vast (and by what standard(s) that vastness is measured) a technosocial network must be in order to sustain Little Science.
The technological dependence of radar astronomy, and the availability of that technology within large technological systems, thus accounts for the emergence of radar astronomy within Big Science settings. The technological dependence of planetary radar astronomy, however, does not explain its utilitarian proclivity, namely, the tendency of radar astronomers to justify their research by its usefulness to space exploration. Nor does the training of most radar astronomers as electrical engineers, who must think in both theoretical and practical terms at the same time, illuminate that tendency. The rise of radar astronomy concurrently with the creation and rapid growth of NASA was perhaps not coincidental.
Although the space agency did not build research instruments outside NASA laboratories during the 1960s, its very existence from 1958 suggested the future availability of funds for instruments and research activity. The Endicott House Conference reflected those funding hopes. After 1970, when NASA funding became a reality, radar astronomy quickly began participating in NASA space missions, such as Viking, until radar astronomy became a space project, the Magellan radar mission to Venus. This close relationship to NASA space missions certainly amplified whatever utilitarian bent radar astronomy already had.
This utilitarian bent also arose from the very nature of conducting Little Science within the context of Big Science. Doing Little Science requires that scientists constantly defend the pragmatic value of their research. A good example is radar astronomy at JPL; it lived off the budgetary margins of NASA space missions until the 1980s. Because obtaining antenna time depended on securing the approval of a NASA mission, radar astronomers had to argue the value of their research on practical, mission-oriented terms.
 In contrast, obtaining antenna time at the Arecibo Observatory depended on the sci-entific value of the radar experiment; its value to NASA was far less important, although research directly related to NASA space missions was carried out there. The primary dif-ference between the JPL and Arecibo facilities was the official recognition granted radar astronomy at Arecibo from the start. The Arecibo telescope always had radar astronomy as one of its prime research objectives, while the JPL Goldstone antenna served mainly to track NASA launches, not conduct scientific experiments. NASA recognition for the scientific value of the Goldstone radar dish has yet to be realized fully or even established on a permanent foundation, though some preliminary steps have been taken.
We can conclude briefly the following about planetary radar astronomy. After a brief initial burst of activity, radar astronomy quickly developed the characteristics of Little Science in terms of manpower, instruments, and published literature. The field continued to shrink throughout the 1970s, reached a low plateau of activity around 1980, then rose slightly in the middle 1980s, as the Goldstone radar once again became available for research.
A number of factors kept planetary radar astronomy a Little Science. Radar sensitivity and target visibility within the declination window limited observational opportunities. The shortage of observational opportunities in turn restricted the number of investigators who could pursue radar astronomy on a full-time basis. Close ties to NASA space projects intensified radar astronomy's utilitarian tendency. The need to justify Little Science within a Big Science setting played at least an equal part in shaping that tendency. The case of the Arecibo Observatory, though, demonstrates the importance of securing institutional recognition for the conduct of Little Science from the outset. Finally, the subsistence of Little Science within Big Science niches and their symbiotic relationship may be a function of large-scale technological systems, whether they be the Western Union telegraph network of the nineteenth century or the big dishes of twentieth-century radio astronomy and space communications.
1. Ostro 25 May 1994.
2. Ostro, "Benefits of an Upgraded Arecibo Observatory," pp.238-239: Ostro 25 May 1994; Campbell 8 December 1993.
3. Harmon 15 March 1994.
4. Renzetti, Thompson, and Slade, "Relative Planetary Radar Sensitivities: Arecibo and Goldstone," TDA Progress Report no.42-94 (Pasadena: JPL, April-June 1988): 292.
5. For a careful scholarly study of telegraph operators as inventors, see Paul Israel, From Machine Shop to Industrial Laboratory: Telegraphy and the Changing Context of American Invention, 1830-1920 (Baltimore: Johns Hopkins University Press, 1992), which is based on the dissertation of the same title, Ph.D. diss., Rutgers University, 1989. For a discussion of the role of the entrepreneur in channeling the resources of large-scale organizations, specifically, the introduction of radio and radio research within the French military by Gustave Ferrié, see A. Butrica, "The Militarization of Technology in France: The Case of Electrotechnics, 1845-1914," paper read at the joint meeting of the American Historical Association and the History of Science Society, Cincinnati, December 1988.