The Emergence of the National Science Foundation as a Supporter of Ocean Sciences in the United States
John A. Knauss
Graduate School of Oceanography, University of Rhode Island, and Scripps Institution of Oceanography, University of California, San Diego
U.S. oceanography grew rapidly after World War II, and in the years immediately after the war, the Office of Naval Research, which began in 1946, provided most of the support and much of the leadership. The National Science Foundation (NSF) began in 1950, but for a number of years its support of oceanography was marginal except for biological oceanography. This began to change with the International Geophysical Year of 1957-1958, and by the time the International Decade of Ocean Exploration (IDOE) began in 1970, NSF had in place the organizational structure necessary to become the dominant player. The timing was excellent, because 1970 was the year of the Mansfield Amendment, which limited military support of science in universities to those programs of military relevance. NSF also housed the National Sea Grant College Program for a brief period, from its formation in 1967 to its transfer in 1970 to the newly established National Oceanic and Atmospheric Administration. NSF policies have significantly influenced the course of science in the United States. Two policies specific to oceanography have contributed much to its strength and vitality. They are the NSF policy that assigns and supports ships through individual academic oceanographic institutions rather than through a single organization, and NSF' s development of a support structure that allows for and encourages large, multi-investigator, multi-institutional programs, a type of program that came to flower with the IDOE and continues today.
The long period of growth of American oceanography began with World War II. The war provided a jump start to a field that until then had few practitioners in the United States and little in the way of support. Harry Hess, who skippered a destroyer, used its echo sounder to explore the bottom of the Pacific Ocean and discovered flat-topped seamounts, guyots. Harold Sverdrup and Walter Munk developed the techniques for calculating the strength and time of arrival of ocean swell on landing beaches. Athelstan Spilhaus, then at Woods Hole, held the patent for the development of the mechanical bathythermograph, a device used to determine the range limitations of sonar, but also a device that taught us about the seasonal thermocline, and was later used by Fritz Fuglister and his Woods Hole colleagues for tracking the cold wall of a meandering Gulf Stream.
But as important as World War II was in providing opportunities for those few who were already engaged in marine science, I believe its most important oceanography legacy was that it introduced the study of the oceans to scientists from a variety of backgrounds who found themselves working at either Scripps or Woods Hole. Carl Eckart, Russel Raitt, Brackett Hersey, Allyn Vine, and a number of others never returned to their original disciplines. For anyone interested in this transition (at least for nonbiological oceanography), I highly recommend the Transactions of the American Geophysical Union (AGU). The difference between the AGU Transactions of just before and immediately after World War II is remarkable.
Because of the Office of Naval Research (ONR), the transition from wartime to peacetime science was smooth. ONR began in 1946, the year after the war ended. Many, of course, left Scripps and Woods Hole, the two big centers of World War II ocean research, to return to their earlier careers, but for those who remained, ONR was there to provide a wide range of support. The Bureau of Ships and other naval operations groups would continue to supply significant funds for a variety of research activities related to their military mission, but ONR allowed Scripps and Woods Hole to broaden their agendas.
And ONR ensured that oceanography would be sup-
ported elsewhere. By the time the National Science Foundation (NSF) arrived on the scene four years later in 1950, ONR was supporting oceanography at the University of Washington, whose program was essentially put on hold during World War II, and at a number of new centers: Texas A&M, the Chesapeake Bay Institute of the Johns Hopkins University, the University of Miami, and, when Maurice Ewing ended his association with Woods Hole, the Lamont Geological Laboratory of Columbia University. Later came Oregon State University and the Universities of Rhode Island and Hawaii.
Many have noted how ONR was the template from which many of NSF's policies and practices were formed. What is sometimes forgotten is the wide range of basic science that ONR supported in those early days. Navy support for oceanography is obvious, but in the beginning ONR provided funding for research ranging from cosmic rays and white dwarf stars, to the structure of protein and the biochemistry of muscle,1 and to developments in nuclear physics and high-speed computing. 2 In the days before there was an NSF or a National Institutes of Health (NIH), ONR was the primary source of federal funding for all basic research supported by the federal government, more than a thousand projects at more than 200 institutions in 1948. The total budget? Less than 30 million dollars per year.3
Roger Revelle, still in uniform, was the first head of ONR's Geophysics Branch whose mandate included meteorology, oceanography, geography, geology, and geophysics. Revelle had returned to Scripps as director by the time I came to work in ONR in 1949. I was the sole program of-ricer for oceanography, and my academic preparation was an undergraduate wartime degree in meteorology, which included a single course in oceanography. But little in the way of scientific expertise was required. In 1949 the principal investigator of each of our contracts was the laboratory director. Each proposal might list a number of individual projects, but the director had the freedom to move funds and scientists from one project to another and to undertake new initiatives. All that was required was a brief quarterly progress report. Our two largest contracts were with Scripps and Woods Hole. When I first arrived, each was for $125,000 a year.
ONR' s role as the sole federal support for oceanographic basic research ended in 1950 with the formation of the National Science Foundation. Alan Waterman, ONR's chief scientist, became NSF's first director, and in time others moved from ONR to NSF. But unlike fields of science less key to the Navy's primary mission, ONR maintained its dominant role in oceanography for a number of years. It was ONR, not NSF, that underwrote the development of programs at Oregon State University and the University of Rhode Island some years after the formation of NSF.4 And it was ONR that underwrote the development of manned submersibles, first the support of Jacques Piccard's bathyscaph Trieste , and later the construction of Woods Hole's Alvin.
When did NSF become the dominant player in support of oceanography? One can look at budgets, and I have, but all who have had any intimacy with the federal budget know that interpretation is not easy. It is difficult enough to assign categories in a contemporary budget. It is almost impossible to reconstruct the actual division of funds after 20 or more years. Major budget items can be tucked away in categories that can be easily overlooked by those doing historical research. For example, one federal report for fiscal year 1969 shows the Navy's contractual oceanographic program 40 percent larger than that of NSF; another shows them essentially equal.5
One must also distinguish between the various types of oceanography. ONR provided relatively little support for biological oceanography, and although the ,Atomic Energy Commission began supporting a wide range. of oceanography, including biological oceanography, in the mid-1950s, as near as I can judge, NSF was the primary source of funds in this field from the beginning. Its total support of biological oceanography was larger than that for all other fields of oceanography in 1962.6
My own sense is that the passing of the torch from ONR to NSF began with the International Geophysical Year (IGY) of 1958 and was completed about the time the International Decade for Ocean Exploration (IDOE) began in 1970.7 ONR
was the dominant player in academic oceanography at the start of the IGY; it was no longer as we moved into the IDOE. However, it is important to recognize that during this period of transition, and continuing today, NSF and ONR supported parallel and joint programs. It is a measure of the skill and common purpose of the program managers in both agencies that those doing the research could mostly ignore the details of the funding as they went about their research, and it is sometimes difficult to remember today which agency supported which parts of which program.
This transition period when primary support of oceanography passed from ONR to the National Science Foundation coincides with the development of the necessary oceanography infrastructure within NSF. The National Science Foundation has from the beginning been organized mostly along disciplinary lines. That NSF did not recognize oceanography as a separate discipline in 1950 is no surprise. The only Ph.D. granting institution at that time was Scripps. With the active assistance of the Office of Naval Research the number of degree-granting institutions began to grow. By 1960 there were a half dozen.8 But most who called themselves oceanographers during this period had earned their degrees in other disciplines, and many continued to question whether a degree in such an ill-defined field as oceanography was the best training. The organizational structure of NSF reflected this uncertainty. As Sandra Toye writes in her administrative history of ocean science in the National Science Foundation later in this volume, "For oceanography, an inherently interdisciplinary field, NSF's early organizational choices created problems that would not be fully rectified for 25 years." A few marine biologists found a home in one or another section of the Biological and Medical Science Division, and it was relatively easy for marine geology and geophysics to find a home in the Earth Sciences Program, but there was no obvious home for physical and chemical oceanography.9
The International Geophysical Year of 1957-1958 brought new money and new prominence to oceanography and the rest of the Earth sciences, and the National Science Foundation structure slowly changed to meet these challenges. The reconstituted Earth Science Section established in 1962 had four programs, one of which was Oceanography under the direction of John Lyman. The Oceanography Facilities program led by Mary Johrde (primarily ship support) was added in 1967, and a Biological Oceanography program headed by Ed Chin was formed in 1968 in the Biological and Medical Science Division.10 As part of a significant reorganization of NSF in 1970, biological oceanography was transferred to the Ocean Science Research Section to join the rest of the oceanographic disciplines. Ship support, polar programs, the Deep Sea Drilling Program, and the International Decade of Ocean Exploration were made a part of a new Directorate of National and International Programs. 11
By 1970 the National Science Foundation had an administrative structure adequate to the challenge of the rapidly expanding field of oceanography. The timing was excellent; 1970 was also the year of the Mansfield Amendment which forbade the Department of Defense to fund projects in basic science unless they were closely related to a military function or operation. The 20-year-long passing of the torch from ONR to NSF for primary responsibility for the support of oceanography was now complete.
In 1966, while NSF was still grappling with how to integrate oceanography into its organization, Congress created the National Sea Grant College Program and placed it in the National Science Foundation. Senator Pell, who introduced the first Sea Grant legislation, was not certain NSF was the best home. For a time he even considered the Smithsonian Institution. Noting that the Smithsonian had served as the nineteenth century launch pad for both fisheries and the weather service, he thought it might serve a similar role for Sea Grant until such time as a better fit could be found within the administration. However, the Smithsonian did not rise to the challenge during legislative hearings, and Sea Grant went to NSF almost by default.12 Sea Grant was about applied research, and it included research in economics and the other social sciences. It had an educational component, and perhaps most critically, it had a significant public outreach program patterned after the very successful agricultural extension service. Sea Grant was not an easy fit in the National Science Foundation of the 1960s. The 1969 report of the Commission on Marine Science, Engineering and Resources (the Stratton Commission) recommended the bringing together of various ocean-oriented agencies within the federal government into a new National Oceanic and Atmospheric Administration (NOAA). NOAA was established in 1970 and the Sea Grant program was a part of it.
Patterns of support established by National Science Foundation during this period have done much to shape the development of oceanography in this country and the way it is practiced today. The most obvious, and probably the most important, is peer-reviewed science proposals, which is prac
ticed throughout NSF. Elsewhere in this volume one can find examples and discussion of oceanographic achievements resulting from such support. In the remainder of this paper I wish to concentrate on two NSF policies that I believe have done much to shape in a very positive way the structure of oceanography within our universities. The two are ship operations and the large, multi-investigator, multi-institutional program. The first was set in motion by ONR and later bought into by the National Science Foundation. It is now the UNOLS (University-National Oceanographic Laboratory System) (see Byrne and Dinsmore paper later in this volume). The second began in 1970 with NSF's International Decade of Ocean Exploration. There were multiship, multi-investigator, multi-institutional programs before the IDOE (e.g., Operation Cabot, the multiship study of the Gulf Stream in 1950 and the International Indian Ocean Expedition that began a dozen years later), but I expect most would agree that this type of program gained full expression with the IDOE. These multi-investigator programs have long outlasted the original source of funding. The IDOE folded its administrative tent in 1981, but IDOE-like programs continue to constitute a significant share of the NSF oceanography budget.13
Nowhere is cooperation between the National Science Foundation and the Office of Naval Research better seen than in the construction and support of research ships. And nowhere, I suspect, have the two agencies had to work harder to bring in line their differing modes of operation. Prior to World War II, only Woods Hole had a research vessel, the Atlantis, that could venture far from shore for any length of time. After the war the mark of an academic oceanographic research program was an oceangoing ship. Many were surplus World War II vessels, modified with varying degrees of success to perform oceanographic research. Their names are familiar to all of this period: first Crawford, Horizon, Vema, Spencer F. Baird; later Chain, Argo, Trident, Pillsbury, Yaquina, Alaminos, and others.14 It was not until the Navy made available to Lamont one its first AGOR vessels, the Robert D. Conrad, in 1962, and NSF built the Atlantis II for Woods Hole in 1963, that the academic research fleet began to acquire ships designed for the task.15
I firmly believe the ship support practices that evolved in the United States after World War II were critically important to the development of oceanography in this country. The system that has developed, whereby the oceangoing research fleet is operated by the university research establishment is unique. I have sometimes wondered how this decision was made. How did it come about that each of the major oceanographic institutions operated one or more research vessels capable of operating far from home port? Was it a conscious decision, the pros and cons carefully weighed and thoroughly thought through, or did it just happen? I believe it was the latter. Woods Hole had a ship, the Atlantis , built and supported before World War II with funds from the Rockefeller Foundation. If Scripps was going to carry out the Marine Life Research Program designed by Sverdrup before he returned to Norway at the end of World War II, it would need some vessels to conduct the monthly surveys. Among its first acquisitions was the 143-foot, 900-ton seagoing tug Horizon, capable of working in the open ocean for weeks at a time. One reason Maurice Ewing left Woods Hole to found the Lamont Geological Observatory was that the Woods Hole director, Columbus Islen, could not guarantee the ship time Ewing wanted for his worldwide geological and geophysical surveys. He got it with Vema, a former 700-ton, 200-foot yacht built in 1923.16 These early decisions at Scripps and Lamont were made before there was an NSF. Perhaps this is all it took; once Lamont, Scripps, and Woods Hole had their vessels, the pattern was established. If you were going to be an oceanographic research institution, you needed a research ship.
As ONR began the practice of supporting a growing number of oceanographic research institutions, it also did its part in providing these institutions with supporting research vessels. In time, the Universities of Hawaii, Miami, Rhode Island, and Washington, along with Texas A&M University and Oregon State University, all had major oceangoing facilities. NSF bought into the practice and slowly became the dominant player in terms of determining how these vessels were to be used and supported.
It did not have to be this way. One could imagine NSF and ONR following the route NSF established in the support of the meteorology departments of this country. It could have established the ocean equivalent of the National Center for Atmospheric Research in Boulder, Colorado which, in the beginning at least, supported the airplanes and other large field equipment required by academic meteorologists. For many years the U.S. practice of having its major academic oceanographic institutions operate an oceangoing research fleet was unique. Even today, the practice is rare.
I am convinced that the spawning of a number of major university oceanographic institutions, each operating one or more large ships and other facilities, has been a major reason for U.S. world leadership in oceanography. Managing a large research vessel can expand one's intellectual horizons. Someone on your staff has an interest in the flux of ions across the air-sea interface, which requires taking your ship to a variety of locations; then comes the question of what happens when these chemical constituents leave the boundary layer; and five years later, as happened to me at the University of Rhode Island, you suddenly find yourself with an atmospheric chemistry program larger than that in any meteorology department in the United States at the time.
It can also add another dimension of experience. Finding ways to repair an engine in a foreign port some 4,000 miles away, or getting a crew member out of jail in the same port, broadens your range of experience and increases your confidence. I expect there was never any doubt in the minds of the directors of Scripps, Woods Hole, Lamont, and the University of Miami (at that time the big four of oceanographic institutions) that they were quite capable of overseeing the work of Global Marine and the conversion and operation of its oil drilling vessel the Glomar Challenger for the Deep Sea Drilling Program. Later, the informal consortium was formalized as the Joint Oceanographic Institutions (JOI), Inc. and participation in the ocean drilling program became international, but major operational responsibility still rests in the different JOI institutions. I know I am not alone when I claim that the Ocean Drilling Program (as it is now called) is one of the most successful, as well as one of the longest-running, singularly focused international oceanographic research programs. The leadership, including the formal direction of the program, has come from the major university oceanographic institutions of the United States.
If somehow the decision to operate major oceanographic facilities had gone the other way—if the responsibility for ship operations had been vested in a central organization, for example, the marine equivalent of NCAR—then I expect buoy arrays, submersibles, and similar facilities would have been housed there also. If this had happened I believe ocean science would be weaker in the United States than it is today.
However, having said this, I believe it is the National Science Foundation, and not the major oceanographic institutions, that deserves much of the credit for ensuring that the academic ship operation program works as well as it does and continues to be acceptable to the ever-growing group of scientists who need to find a way to work at sea. In the beginning, each institution ran its ship program differently. What might be available as ship support or a ship' s scientific equipment at one institution would not necessarily be available at another. One needed to know the details. As a consequence the scientist in charge of a seagoing program generally came from the same institution as the ship. As the family of seagoing oceanographers grew, not all made their careers at ship-operating institutions. The UNOLS program does much to guarantee uniformity of support among the ships of the fleet. The ships are different and some programs can only be accommodated on certain vessels, but the number of unpleasant surprises concerning vessel furnished facilities and support has been significantly reduced, if not entirely eliminated. You no longer need be a part of a ship-operating institution to conduct a major research program requiring a research vessel.17
If the early ship support practices of ONR and NSF that evolved into UNOLS did much to define the early development of academic oceanographic institutions in the United States, the NSF support of multi-institutional, multi-investigator programs through the International Decade of Ocean Exploration did much to define the practice of oceanography in the 1970s. These large programs, each with its own catchy acronym, established a pattern of doing oceanography that continues today. How the IDOE came to be and was organized is described in the paper by Jennings later in this volume.18 I wish to note the contribution of these programs to what might be called the sociology of oceanography. Just as I believe the operation of oceangoing ships did much to define the oceanographic institutions for the first half of NSF's 50 years, so I believe NSF's sponsorship of large multi-investigator programs has done much during the last 25 years to develop a level of cultural sophistication among the oceanographic community found in relatively few others fields of science.
Oceanographic field work is expensive and often frustrating. Many of the most interesting problems are best attacked by a multiprobe program. Those who succeed at sea with their observational and experimental programs soon learn that it can be dangerous to put all of one's effort into a single approach or a single instrument. We also learn that serendipity often plays a larger role than we wish to admit in whatever successes we achieve. By 1970, many of those with extensive seagoing experience were ready to embrace the concept of a multifaceted, multi-investigator approach to oceanographic field work. We believed it was the most cost-effective way to attack problems that were often not as well defined as we liked to suggest in our proposals. Multi-investigator programs followed naturally, and several were in the planning stage at the start of the IDOE: the Geochemical Ocean Sections Study (GEOSECS), the Mid-Ocean Dynamic Study (MODE), and the Climate: Long-Range Inves-
tigation Mapping and Predication (CLIMAP) program.19 These large multi-investigator programs that reached full flower in the IDOE have continued to constitute a significant fraction of NSF's ocean budget.20
One reason these large programs continue is because they generate interesting, often exciting, science, and they appear to be cost-effective. I also believe that an unanticipated positive contribution of these programs is that they teach a degree of cooperation, mutual respect for different approaches and personalities, and an understanding of institutional complexities. Scientists who participate in—some might suggest those who survive—these large multi-investigator, multi-institutional programs generally come away with a deeper understanding of their colleagues and the various ways of achieving their own goals.
I do not wish to imply that scientific cooperation is unique to oceanography. Joint investigations and joint papers are the rule, not the exception, in science; and much organization and cooperation is required to gain the maximum effectiveness from such large pieces of equipment as telescopes, satellites, high-energy machines, or deep-sea drilling vessels. Committees of the National Research Council and others are well equipped to outline important problems that require attention. But I do believe it is relatively rare in science to not only have a problem first defined by a committee, but then to have a committee outline the approach, determine what kinds of scientific specialists are required to successfully implement the approach, provide a steering committee to ensure that there are no significant holes in the combined proposals, make certain that deadlines are met, and do all of the other chores necessary to ensure that the whole of the multi-investigator program is greater than the sum of its parts.
Joint programs of this kind are not to everyone's taste, nor do I expect many to make a career of participating in such programs one after another, but I do believe that large multi-investigator programs contribute to the education of those who participate and that oceanography is a stronger field today because so many have been associated with at least one such program. I have no proof, and others may disagree, but it is my sense in talking to colleagues who have been so involved, that they have a deeper understanding of what it takes to mount a successful science program (whether it be single or multi-investigator, small science or large science). They know how to go about it, and they have a confidence that they can succeed. I believe NSF's multi-investigator, multi-institutional programs, which started with the IDOE and continue today, have contributed significantly to all of oceanography, both big science and small.
In summary, oceanography in this country was jump-started during World War II, and the Office of Naval Research was there immediately after the war to support and to expand the field. The formation of the National Science Foundation in 1950 began a slow, 20-year passing of the torch of primary support from ONR to NSF that was completed in 1970 with the start of the IDOE and the passage of the Mansfield Amendment. Both agencies deserve major credit for ensuring that the transition went as smoothly as it did. For those of us who were supported during this period it was sometimes difficult to remember which agency was responsible for which support. There were few hiccups as we transitioned from one grant or contract to another.
Fifty years of partnership between those who practice science and those who support science result in cultural patterns that we sometimes take for granted. As I reflect on how oceanography has been supported and is supported in other parts of the world, I see several examples of NSF's way of doing things that I believe have been important to the development of oceanography in this country. The first, of course, is the peer-reviewed grant proposal system, which pervades all of NSF and has been widely discussed for many years. Two others are peculiar to oceanography and may be less obvious, but I believe each has played an important role in the development of both oceanographic institutions and oceanographers in this country. They are institutional support of ships through the UNOLS program and the sponsorship of multi-investigator, multi-institutional programs that began with the IDOE. As NSF looks back on 50 years of support for oceanography, it has much of which to be proud.