Major Discoveries and Biomedical Research Organizations: Perspectives on Interdisciplinarity, Nurturing Leadership, and Integrated Structure and CulturesRogers Hollingsworth and
Ellen Jane Hollingsworth, The presentation made by Dr. Hollingsworth at the Building Research Success symposium was based on this paper. A related version of this paper has been published in Peter Weingart and Nico Stehr, eds., Practising Interdisciplinarity. Toronto: University of Toronto Press, 2000. pp. 215-244. Introduction This paper is concerned with the structural and cultural characteristics of research organizations which influence the making of major discoveries in twentieth-century bio-medical sciences, especially characteristics of research organizations which repeatedly make major discoveries across time. Although most of the empirical analysis for these findings is based on research organizations in the United States, we also make some reference to research organizations in other nations. This paper is part of a larger study involving structural and cultural characteristics of bio-medical research organizations in four countries. Why do research organizations vary in their capacities to make major discoveries in biomedical science? Science, especially in the twentieth century, has been very dynamic and has grown in unpredictable ways. Because most research organizations experience considerable inertia and change rather slowly, they have considerable difficulty in adapting to the fast pace of scientific and technological change. Time and time again, a research organization has been a world class leader in an area of science, but because of organizational inertia and failure to adapt to new trends, it has lost its leading edge. This paper argues that organizations require distinctive structural and cultural characteristics if its scientists are to make major discoveries repeatedly. It is to the identification of these characteristics that this research is addressed. The questions posed in this research have their bases in the sociological literature concerned with how the structural and cultural characteristics of organizations influence the making of radical innovations. There is a vast and excellent literature in the history and sociology of modern science about performance in the scientific community — e.g., scientific discovery, the creative process, and more generally, scientific productivity (Ben-David 1960, 1971, 1977; Merton 1961; Pelz and Andrews 1966; Zuckerman 1977; Allen 1978a,b; Olby 1979; Fox 1983; Allison and Long 1990; Shapin 1995), and about the organizational contexts within which science occurs (Lynch 1985, 1993; Shapin 1995). In recent years, an increasing number of studies (Latour and Woolgar 1979; Fujimura 1987; Latour 1987; Shapin 1995; Rheinberger 1997) has focused on the importance of the research lab/department as the site of discoveries. These and other studies have emphasized the importance of tacit knowledge and have demonstrated that knowledge is highly differentiated, unequally distributed, and richly sited in local contexts (Lynch 1985, 1993; Polanyi 1966; Dasgupta and David 1993;). Whereas much of the recent literature has focused on the research setting of a single organization, this paper differs in being more comparative and historical in nature. It specifies a series of research organization and laboratory and/or department level variables and then uses brief case studies to analyze the pattern of relationships among these variables, especially as they relate to the making of major discoveries in bio-medical science. Concepts,
Data, and Methods The concept "major discoveries" Conceptually, this project defines a major discovery as a finding or process, generally preceded by numerous "small" advances, which solved a particular problem and in turn led "to a number of smaller advances, based on the newly discovered principle" (Ben-David 1960: 828; Merton 1961, 1973; Rosenberg 1994: 15). Historically, a major discovery might have been a radical or new idea, the development of a new methodology, a new instrument or invention, or a new set of ideas. It need not have occurred all at once; it might have extended over a substantial period of time, involving a great deal of tacit knowledge (Polanyi 1966; Latour 1987). To implement the concept "major discovery", we rely heavily on the scientific community, using criteria the scientific community have created to recognize major discoveries. Using a diverse set of strategies to operationalize our definition, we include discoveries which led either to winning or near winning of a major prize. While we rely heavily on discoveries associated with major prizes as a strategy for defining major discoveries, we are very careful not to rely on any single prize. These discoveries are:
We have access to the Nobel Archives for the Physiology or Medicine Prize at the Karolinska Institute and to the Archives at the Royal Swedish Academy of Sciences in Stockholm prior to 1946, but for reasons of confidentiality, we do not have access to these archives for the past 50 years. To capture the variety of major scientific discoveries during this period, we also use several other criteria. We include:
Recognizing that not all major discoveries can result in a Nobel Prize, we have emphatically worked at making certain that this is not a project about Nobel Prizes. Once having identified a major discovery, we then have determined when and in what research organization(s), department(s), and lab(s) the discovery occurred. In some instances, the research organization did not have departments. The result of this process has been to identify organizations associated with major discoveries, and to award to them "credit" for the major discoveries with which they were associated. In some cases, scientists made their major discoveries by conducting research first in one and then in another organization. All organizations in which scientists conducted research directly associated with major discoveries have been credited with the discovery. The research takes note that all scientists who were engaged in making major discoveries were not always recognized by prize committees. However, this is a study about major discoveries and the properties of the research organizations where they occurred. Thus, the omission of "unrecognized" individuals by prize committees does not significantly bias our results. Our method of conducting in-depth studies of the organizations, departments, and/or labs where there was the occurrence of a major discovery permits us to identify those scientists who were involved but did not receive recognition by a prize committee. This research is not a history of scientific ideas or a study of creativity of individual scientists, although it acknowledges that discoveries were made by individuals and that creativity occurs in individuals. The concern of this paper is with how the context of the research laboratory and/or department and organization influenced the making of major discoveries. However, major discoveries did not occur at random in organizations, labs, and departments; rather, there were regularities in the characteristics of organizations and labs and/or departments where they occurred. Structural and cultural concepts The analysis of research organizations and labs and/or departments revolves around seven basic concepts (see below for the concepts, and for each concept, indicators or examples). They are the degree of
It should be noted that depending on whether the unit of analysis is the organization or the laboratory and/or department, there are modest differences in the appropriate indicators (or examples) of the concepts, which are noted below. Key Concepts:
Indicators and Examples The organizational level Scientific Diversity:
Depth:
Differentiation:
(At department level: "Research group" should be substituted for "department".) Hierarchical and Bureaucratic Coordination:
Integration of Multi-disciplinary Perspectives: Across specialties, the
(At departmental level: activities across research groups would be noted.) Visionary Leadership: Capacity for Understanding Direction in Which Scientific Research is Moving and Integrating Scientific Diversity:
Quality:
Sample and Data The larger study underlying this report is based on 128 research organizations in the United States: 28 research organizations where two or more major discoveries occurred and a comparison group of 100 research organizations where one or no major discovery occurred. The research focuses on four general types of organizations (universities; medical centers; free standing research institutes; and industrial research laboratories). For approximately two dozen of these organizations, we have completed in-depth case studies. Details about the sampling process to select organizations in which no major discoveries took place are available elsewhere (Hollingsworth and Hage 1996). Obviously, there are many criteria by which one might evaluate the performance of research organizations: productivity, citation indices, level of funding per scientist, and, for universities, the number of graduate students trained or the quality of the graduate program or faculty. We do not suggest that research organizations where few or no major discoveries occurred are performing poorly in science. Indeed, many excellent research organizations have hardly had a major breakthrough in bio-medical science. However, the underdeveloped state of knowledge about the conditions which facilitate or hinder the making of major discoveries justifies the emphasis of this research inquiry. This paper is based on numerous kinds of data: more than 200 in-depth interviews, archival materials, oral histories, secondary published materials, and scientific papers. Constraints of space have made it possible to list only a small fraction of these materials in the references section. Methodology Throughout, we employ comparative and narrative methodologies in the analysis (Franzosi 1998). Since the goal of the research has been to determine how the properties of research organizations are associated with the making of major discoveries, the analysis has compared organizations where major discoveries occurred with those where no major discoveries or only one discovery took place. As an additional form of comparison, historical analysis is used. Using this method, we compare some organizations with themselves, analyzing them before and after major structural and cultural changes, in order to examine how such changes related to the making of major discoveries. Despite our emphasis on the common structural and cultural properties of organizations associated with making major discoveries time and again, it is important to recognize that there has been no single route by which research organizations have acquired these common traits. Some organizations where major discoveries occurred over and over were major organizational innovations — that is they were fundamentally new kinds of organizations. This was the case with Rockefeller Institute, the Johns Hopkins Medical School, and the California Institute of Technology. Another route whereby organizations made repeated major discoveries was fundamental structural or strategic change within an existing type of organization. And the third route was by creating a new organization among the existing types of organizations. Rather than being a full report of this research project, this is only a preliminary report, concerned with the characteristics of organizations that influence their capacity to make frequent major discoveries. Irrespective of the route, those organizations where major discoveries occurred again and again had the following characteristics of the seven concepts in this study. They scored high on visionary leadership, scientific diversity, interdisciplinary and integrated activities, and quality. Conversely, they were low on differentiation and hierarchical and bureaucratic coordination, and they were moderately low on depth. Selected
case studies Highly integrated, small research institutes Two American research organizations which have had major discoveries in bio-medical science time and time again are Rockefeller Institute/University and California Institute of Technology. Rockefeller Institute/University has been the site of more major discoveries in bio-medicine in the twentieth century than any other single institution in the world, while Cal Tech has been in the company of a handful of other research organizations where a number of major discoveries occurred at multiple time points in the twentieth century. The central finding in these two case studies is that major discoveries occurred repeatedly because there was a high degree of interdisciplinary and integrated activity across diverse fields of science (thus, scientists with diverse perspectives interacted with intensity and frequency) and because of leadership which gave particular attention to the creation and maintenance of a nurturing environment, though with rigorous standards of scientific excellence. These two relatively small research organizations had moderately high levels of scientific diversity and moderately low levels of depth but had very low levels of internal differentiation (i.e., separate disciplinary departments), and had visionary leaders who provided strategies for integrating scientific diversity. They were also low on hierarchical coordination and bureaucratization. Limitations of space permit us to do little more than report in a rather preliminary fashion some of the characteristics of research organizations which are associated with major breakthroughs in bio-medical science. In a fuller forthcoming study, there will be an extensive examination of the characteristics of numerous other research organizations — even though of very high quality — where major discoveries rarely occur. The Rockefeller Institute In its early years, Rockefeller Institute acquired a culture of excellence and a structure which facilitated the making of major discoveries again and again. With the passage of time there have been modifications in the Rockefeller organization, but even so, there has been sufficient continuity in its structure and culture of excellence so that it has remained an organization where major discoveries have continued in bio-medical science. Unlike the Koch Institute in Berlin or the Pasteur Institute in Paris which were founded around great scientists and their particular research, the Rockefeller organization was a new kind of research institute which from the beginning emphasized diversity in the bio-medical sciences (Corner 1964; Dubos 1976). Not focused only on a specific area of science, the goal in the founding of the Institute was to pursue diverse areas in bio-medical sciences. Whereas earlier institutes had been constructed around bacteriology, bio-medical science was changing very rapidly by the time the Rockefeller Institute was established at the turn of the century. Bacteriology had become much more linked with pathology, and both fields were becoming more closely related to discoveries in organic and physical chemistry, as well as in physics. A broad conception of bio-medical science became the guiding philosophy of the Institute from the beginning. The consequences of these decisions about a broad, integrated approach to science are discussed below. In the early part of the twentieth century, there was only moderate diversity at Rockefeller, as reflected in the variety of research laboratories, and moderate depth because there were only a few people in each lab. However, in the recruitment process, the first director, Simon Flexner, was very receptive to recruiting scientists from different cultural and scientific areas (e.g., the Frenchman Carrell, the Austrian Landsteiner, the Japanese Noguchi, the Russian Levene, and the Germans Meltzer and Loeb). This pattern of recruitment provided diversified approaches to problems and styles of thought as well as to fields of research. Almost every one of these scientists incorporated cultural and scientific diversity in his own individual cognitive makeup, which enhanced the potential to cross scientific disciplines. Almost all of the distinguished scientists in the history of the Rockefeller organization have internalized considerable scientific diversity, making it relatively easy for such staff to feel an affinity with others who crossed academic disciplines. From the very beginning, Rockefeller Institute did not organize the production of knowledge around academic disciplines, as was increasingly the case in major universities. The Rockefeller practices were unlike those in organizations in which academic disciplines were the dominant principle for organizing and coordinating the production of knowledge, and in which there was a tendency to recruit scientists who internalized less cultural and scientific diversity. One of the distinctive qualities of the Rockefeller organization was its tendency to recruit individuals who had been socialized in different cultures, subsystems, disciplines, and/or working environments. And because these were individuals who internalized cultural and scientific diversity at the time of recruitment, they had the potential to acquire new styles of thought and scientific competence and to internalize even greater diversity. In short, Rockefeller from the outset was a place where there was a willingness of its scientists to live and work in multiple disciplinary worlds simultaneously. As Gibbons et al. (1994) point out, achievement in science is accelerated by communication which in turn is enhanced by scientific mobility, mobility being an important precondition for the cross-fertilization of ideas. Rockefeller was quite unique in the early twentieth century in the degree to which it recruited from different parts of the world, senior scientists who had moved among different sites of knowledge production and who had worked in multiple disciplines. This also produced a scientific hybridization which over time led to new techniques, devices, and principles (e.g., instances of scientific creativity, of sudden insights, and the opening of novel pathways to solving difficult problems). A research institution such as Rockefeller had several distinct advantages over most teaching institutions. Most teaching organizations attempt to present an entire field of knowledge, and they often recruit people not because of their research excellence but to cover a particular field of knowledge. A research institute has no obligation to cover an entire field of knowledge, and can be very opportunistic in terms of the fields on which research is undertaken. If it wishes, it can neglect complete fields, recruiting scientists solely on the basis of the availability of the best people who are working on problems relevant to the institute. Moreover, a research institute, unlike a teaching organization, has the flexibility to move into new areas with considerable rapidity. And because Rockefeller was not a teaching institution, it had the luxury of being able to recruit scientists of excellence even if they were lacking in ability to speak English (Flexner 1930). Of course, the Institute was generously endowed by John D. Rockefeller, but a number of other institutes were also very well endowed (e.g., the Phipps Institute in Philadelphia, established by steel magnate Henry Phipps; the Memorial Institute for Infectious Diseases in Chicago, funded by Harold McCormick, a son-in-law of John D. Rockefeller; the Carnegie Institution in Washington, endowed by Andrew Carnegie). But none of these organizations acquired the distinction of the Rockefeller Institute, for while money was a necessary condition for an institute to produce distinguished science, it was not a sufficient condition. One of the most important conditions for an organization to have repeatedly made major discoveries across long periods of time is the quality of its leadership, a variable to which many organizational sociologists give scant attention. Over the years, Rockefeller has attempted to have directors/presidents who were capable of interacting in a meaningful way with its scientists and who also knew personally the leading bio-medical scientists of the world. Of the eight directors/presidents since the founding of the Rockefeller Institute, six made major discoveries in bio-medical science by the criteria described above, and the two who made no major discoveries (Detlev Bronk and Fred Seitz) were distinguished scientists who had been Presidents of the National Academy of Sciences. Four were Nobel laureates in physiology or medicine. The first director, Simon Flexner, left an indelible mark on the Institute. It was he who developed the expectation that Rockefeller should be headed by an individual who had
Of course, not all subsequent directors/presidents have lived up to the leadership standards of Flexner, but these are the ideals by which successive leaders have been evaluated within the organization. Flexner was aided by a Board of Scientific Directors, made up of some of the leading bio-medical scientists of America. The first president of the Board was William H. Welch of Johns Hopkins Medical School, considered the chief statesman in American medical research and education. The Board was responsible for the appointment of the scientific staff and for the establishment of the general policies concerning the scientific investigations to be undertaken. The director (Flexner) was appointed by the Scientific Directors and was to be in intimate contact with the scientific staff. There was a Board of Trustees to conduct the financial responsibilities of the Institute, but the Board of Scientific Directors played an important role in recruiting some of the most outstanding scientists ever to be assembled. The distinctive role of the Board of Scientific Directors combined with the skills of the Director to facilitate such distinguished recruiting. This process lasted until 1953, when the Institute became a small university and the Board of Scientific Directors and Board of Trustees were merged into a single Board of Trustees. Since then, the Rockefeller organization has not had a separate board of world class scientists who have made the final decisions about personnel. The quality of recruitment, while continuing to be distinguished, has not been of the same extraordinary consistency as during the time when the Board of Scientific Directors was intimately involved in making scientific appointments. At the Institute, permanent scientists were called Members and had indefinite appointments corresponding to appointments as professors in American universities. The next highest title was Associate Member which carried an appointment of three years (associate members were eligible for reappointment), but in general after a second three year term, Associate Members either were appointed as Members or resigned. Associates were appointed for two years; Assistants and Fellows for one. Eligible for reappointment, scientists in these ranks after three to five years left the Institute or rose to higher rank. Like the Kaiser-Wilhelm and Max-Planck Institutes which later had similar policies, the Rockefeller Institute was thus providing advanced training for the elite of bio-medical scientists. The process of appointment as a senior scientist (Member) was extremely rigorous. Not all senior scientists received their initial appointments at that level. For example in 1934, 46 percent of senior scientists had initially been appointed to the rank of Member, while 54 percent had been promoted from within. Members rarely left. Of Members, only Eugene Opie had resigned by the mid 1930's, and he returned. One former president of Rockefeller University confided to us that during much of its history, unless there was a strong belief at Rockefeller that a scientist was likely to win a Nobel Prize, the person was unlikely to have a permanent appointment. It was not enough that Rockefeller had the above-described leadership, outstanding scientists, a moderate to high degree of diversity, and substantial funding for major discoveries to occur time and time again. Other structural/cultural characteristics were also important. If scientific diversity was to lead to scientists having rich horizontal interaction with each other, it was important the Institute not be differentiated into academic departments or other units which would fragment the production of knowledge. The organization needed to be quite integrated, which was the case. First, there was no institutionalization of disciplinary departments. It is true that early on the Institute was organized into the Department of the Laboratories, with a number of labs, and the Department of the Hospital. The first of its kind, essentially a laboratory for the study of human biology and pathology, the hospital was always small, with a number of its staff without MD degrees. Over time a number of its permanent staff have been members of the National Academy of Sciences. Here was the realization of the ideal of the dialectic between clinical and basic science. A case in point, Avery’s discovery of the chemical substance responsible for bacterial transformations (a major turning point in the transition from medical microbiology to molecular biology) was made by a group trained in medicine and working in the Rockefeller research hospital. Their research, focused on understanding of biological phenomenon, influenced the practice of medicine only indirectly, but led to one of the most important biological discoveries of the twentieth century. Staff of the Department of Laboratories and the Department of the Hospital intermingled on a daily basis. Indeed, the key to much of the Rockefeller organization’s success in making numerous major discoveries was the scientific integration of the institute. Diversity and depth of knowledge within a well integrated research organization have the potential to change the way people view problems and to minimize their tendency to make mistakes and to work on trivial problems. In the final analysis, in order to make major discoveries, scientists must work on significant problems which are "do-able." And the greater the research group's diversity and depth within an integrated structure, the greater the likelihood that scientists will not stray into unproductive areas. Frequent and intense interaction among people of like minds (with low levels of diversity) tends not to lead to major breakthroughs. Oswald Avery’s career at the Rockefeller Institute is an interesting case in point. Appointed at age 37 to Rockefeller, Avery at the time was a highly competent researcher who had worked across several fields, but had demonstrated little creativity and very little originality as a scientist. Once in the Rockefeller context (with considerable diversity and depth), Avery’s potential emerged, and by the time (1944) he published his classic paper with MacLeod and McCarty on DNA and transformation, he had internalized vast areas of the fields of bacteriology, immunology, chemistry, and bio-chemistry. Our research on American research organizations suggests that the context in which scientists are embedded influences their performance. If scientists work in environments where there is considerable diversity and depth, and have frequent and intense interaction with those having complementary interests, they increase the probability that the quality of their work will improve. It is the diversity of disciplines and paradigms to which individuals are exposed in frequent and intense interactions that increases the tendency for breakthroughs. Intellectual and social integration was maintained at Rockefeller Institute by a variety of devices. There was high quality food at lunch, served at tables for eight. The idea was that a single conversation could take place at such a table, but not at a larger one. Eating meals together while conversing about serious scientific matters was an important part of the Rockefeller culture and an important means of integrating the scientific diversity and depth of the Institute. Even though there was considerable diversity at the Rockefeller, the type of diversity was very different from that which existed at the colleges of Oxford and Cambridge, where eating was also an important part of the culture. At the British colleges, diversity ranged all across the board (e.g., from archaeology to mathematics). With so much diversity, it was considered ill-mannered to talk about one's work at the high table at the colleges of Oxford and Cambridge, as many of those present would be unable to comprehend the line of discussion. But at Rockefeller, diversity was only within the biological sciences and the norm was to carry on lively lunch time discussions about biomedical and related sciences. The lunch table was a great learning experience where people had intense discussions about new approaches to bio-medical problems. Integration was also facilitated by the weekly conferences everyone was expected to attend, at which scientists reported about their work. Within the Hospital Division, there were afternoon tea times which most faculty attended. One of the most important integrating devices was the journal club, especially the Hospital Journal Club. Throughout the academic year, the Hospital Journal Club generally met twice a month. Everyone was expected to attend, and to be prepared to report on a paper outside one’s research field but of general interest to everyone. No one knew in advance who would be called upon to present materials to the journal club. Why would world class scientists agree to participate in such an activity? They did this because they knew they were at a great organization, and believed that one reason it was so distinguished was that they were continuously learning from each other. This kind of regular reading outside one’s own specialization and in areas of interest to others in the organization was one way of continuing to expand intellectual horizons. When one reads the archives of Rockefeller Institute/University, one is struck by the prominent role of Simon Flexner in establishing in the early twentieth century a culture of nurturing young scientists. Once institutionalized, this trait persisted among leaders of Rockefeller (notably at the lab level), especially in the habits of Avery, Bronk, Bloebel, and Wiesel. Finally, the environment of New York was a great asset for the Rockefeller Institute. Before jet aircraft, most distinguished foreign scientists traveling to America arrived in New York and invariably visited the Rockefeller Institute. Certainly no other bio-medical research organization in America was so favored with foreign scientists. It was located in a magnificent part of New York City. After the Second World War, its neighbors were the New York Hospital, Cornell University Medical College, and the Sloan Kettering Institute for Cancer Research. Its environment provided an opportunity for access to many of the latest ways of thinking about bio-medical science. If a small institute could not have all the diverse ways of approaching bio-medical science on its permanent staff, it nevertheless had the opportunity to have many of the world’s leading scientists passing through the institute with reports about their latest work. Rockefeller Foundation grants to young British and European scholars brought to New York the cream of the crop of young scientists, for stays of varying lengths. As a result of the structure and culture of the Rockefeller organization, it is still one of the world's great centers for bio-medical research. Yet after the mid 1950's, it no longer excelled quite so much relative to all other bio-medical research organizations, as during the previous half century. First, jet aircraft had an extraordinary effect in shifting the location of excellence in bio-medical science, as the elite of the scientific community no longer believed it was necessary to be located on the east coast. Distinguished scientists were increasingly willing to live elsewhere. Moreover, as the National Institutes of Health began to fund bio-medical research on a grand scale, the Rockefeller organization no longer had the same kind of distinctive financial advantages. Indeed, no major private institution could any longer attain and continue scientific excellence without federal funding. As a result of new modes of transportation and new sources of funding, Rockefeller was no longer able to dominate the world of bio-medical science. The internal structure of Rockefeller also changed, a modification which had some adverse effect on its long term future, though not enough effect for it to lose its prominence as a center of excellence in research. In 1953, the Institute appointed Detlev Bronk, President of Johns Hopkins University as its Director. As noted, under Bronk's stewardship the Board of Scientific Directors was abolished and the Board of Trustees became the sole governing authority of the organization. Also, under Bronk’s leadership, Rockefeller Institute became the only purely graduate university in the United States. It was an exceptional university, as there was no formal schedule of classes, only eight courses were offered, and no courses were required. It has long had many more faculty than students, and over time the students have been unusually gifted with a very high degree of self-discipline. Because of their small numbers, students have been able to receive very rigorous training in a highly nurturing environment. But as a result of the new role of the organization, the nature of the faculty changed. While under Bronk and his successor Seitz, a number of distinguished new faculty were appointed, the organization (without its Board of Scientific Directors) also began to appoint some less distinguished scientists to tenured positions. One reason this occurred was the increase in diversity in the organization. With increased size, it became more difficult for the President of the University to have the same level of knowledge of the scientific qualities of each scientist. Moreover, Bronk, while a brilliant administrator with a great deal of charisma, was not in the mainstream of bio-medical science when he became Director of Rockefeller Institute. Though he had been a distinguished bio-physicist, he had long been an administrator. Moreover, his successor, the distinguished physicist Fred Seitz, had been President of the National Academy of Sciences and had never been a biological scientist. He simply did not know the bio-medical science community with the same degree of intimacy as Rockefeller's first two directors. More importantly for the ability of the Rockefeller University to make major discoveries, its internal structure was moderately modified in response to changes in the larger funding environment. Once funding from the National Institutes of Health became accessible, various labs grew in size, turned somewhat inward, and became somewhat more autonomous. No longer did everyone have lunch together. By 1970, there were too many senior scientists, and too many postdocs and students for the early twentieth century type of communication and integration to exist. Most labs began to have their own journal clubs and attendance at the weekly scientific presentations dropped off dramatically. No longer was there the same degree of horizontal communication as during the first half of the century. And yet, even if Rockefeller University today is less scientifically integrated, it is still much less differentiated internally than every other American university. The fact that there are still no departments and that a lab closes down when the head departs means that the organization has an extraordinary amount of flexibility to adapt to new knowledge. This flexibility and adaptiveness explain why Rockefeller University, despite it small size, still towers over all other research organizations in America. Today it has a higher proportion of its faculty as members of the National Academy of Sciences and as Howard Hughes investigators than any other research organization in America. Moreover, its scientists receive more funding from the National Institutes of Health per scientist for bio-medical research than those in any other research organization in America. And it is still an organization where a number of major discoveries occur. California Institute of Technology Like Rockefeller Institute/University, Cal Tech when it was founded in the early part of the twentieth century was a new kind of organization. Though a university, it had divisions (e.g., chemistry, engineering) but no departments, and it had a plural executive, and was also to be very small. It has historically had a moderate degree of diversity and a high degree of integration. Neither organization attempted to do everything in the biological sciences. Because Cal Tech had no medical or agricultural school, its scope in the biological sciences was quite limited. With somewhat narrow scope, its degree of scientific diversity was somewhat limited, and because of its small size, it did not have great depth in many areas. In most research organizations, it is considerable diversity combined with moderate to high depth which leads to fragmentation and internal differentiation. However, Cal Tech, like Rockefeller, has never been differentiated into departments like large American universities. In its early years, a triumvirate of the physicist Robert A. Millikan, the chemist Arthur A. Noyes, and astrophysicist George Ellery Hale guided Cal Tech to national prominence in the physical sciences and engineering. They institutionalized an organizational culture which has lasted to the present. The Institute would do only a few things but would do them with excellence. There was to be heavy emphasis on research and advanced training, and there would be no mass education. Moderately high salaries would attract a high quality faculty and generous fellowships would attract outstanding graduate students. There would be relatively few students, and classes would be very small. Excellence was to be the hallmark of the organization. In a relatively short period Cal Tech attained a distinguished institutional reputation. By 1929 it was among the leading three physics and chemistry research centers in America. Whereas Johns Hopkins and the University of Chicago had also attained prominence in a rather short period of time as the creations of single patrons, Cal Tech was supported by a diverse group of local elites. And while Johns Hopkins and the University of Chicago were also relatively small organizations, each was somewhat broader in scope (Servos 1990: 264-275). Today, Cal Tech is still relatively small and continues to have moderate scientific diversity and depth, and a very low degree of differentiation. In the early 1990's, the Institute had 1, 875 students (more than 1,000 of them graduate students), a faculty of approximately 270, and about 700 fellows and visiting professors. Cal Tech faculty and students over the years have made a substantial number of major discoveries in the bio-medical sciences. Its faculty and administration have received 21 Nobel Prizes and 31 National Medals of Science. The key to its having so many major discoveries has been its interdisciplinary culture. As recent president Thomas E. Everhart observed, people can cut across fields of inquiry more readily than in most places because of its small size. It is "at the intersection of traditional fields of study such as physics, chemistry, and biology where major discoveries have occurred and Cal Tech has always been in pursuit of major discoveries" (Los Angeles Times, January 12, 1992: 6). Cal Tech had no biology until the late 1920's, but when the Institution decided to create a biology division, it did so by recruiting geneticist T. H. Morgan of Columbia University, the most distinguished biologist in America. This appointment sustained Cal Tech's emphasis on excellence. Morgan was not only a member of the National Academy of Sciences, but had served as President of the Academy. He had excellent connections with the National Research Council and the Rockefeller Foundation, and within four years of his arriving at Cal Tech, his discoveries resulted in his being awarded a Nobel Prize. The award turned world attention to the Institute's new biology program. Entering the field of biology relatively late, Cal Tech was free of the institutional inertia facing many research organizations in the biological sciences. In a number of universities, biology was fragmented into many different disciplines (e.g., natural history, botany, zoology, eugenics). With Morgan's arrival, Cal Tech developed biology with a firm foundation in genetics. In many ways this was fortuitous. No one at the time knew that by the 1950's the biological sciences would be firmly grounded in the science of molecular genetics. Building a division with strength in this area, Cal Tech was in the forefront of the way in which biological research would move. While in some respects he was past his prime, he was still a major leader in the biological sciences and was an excellent person to assist in building a biology program. Morgan thought in strategic terms and wanted to develop a biology program with strong interaction with physics and chemistry. Though he did not have a strong grasp of either physics or chemistry, he played an important role in developing a culture for the Biological Division which would emphasize genetics, embryology, physiology, biophysics, and biochemistry, all in cooperation with physicists and chemists. While he lacked strength in some of these areas, he brought to Cal Tech a remarkable group of younger colleagues from Columbia: Alfred H. Sturtevart, Calvin B. Bridges, Jack Schultz, and Theodosius Dobzhansky. Kohler (1994: 124) argues that Morgan's group had never been as productive and as innovative as it was in the 1930's after arriving at Cal Tech. The emphasis was on diversity and cooperation. "Doors between laboratory rooms were always kept open, there were no individual offices or telephones, and work places were communal". Just as Flexner had been instrumental in institutionalizing at Rockefeller a culture of providing rigorous criticism within a nurturing environment, Morgan played an important role in institutionalizing this kind of culture in the Biology Division at Cal Tech. At Columbia University and at Cal Tech, the Morgan group engaged in a great deal of social interaction within a rigorous scientific environment. This kind of culture persisted in Cal Tech’s Biological Division for many years. During the 1930's the Cal Tech biology group attempted to connect Drosophila genetics with developmental and evolutionary biology (Kohler, 1994: 176), a strategy which reached its apex with the work of Ed Lewis, awarded the 1995 Nobel Prize in Physiology or Medicine. The innovations and excitement came from work of the generation after Morgan and from the culture which Morgan had been instrumental in establishing. Sturtevart and Schultz in particular helped to foster a spirit of crossing fields. The group made serious efforts to integrate experimental embryology and biochemistry with genetics. The next generation working at Cal Tech (George Beadle, Boris Ephrussi, and Theodosius Dobzhansky) advanced new systems of development and evolutionary genetics (Kohler 1994: 177). The culture of working across fields at Cal Tech facilitated the moving of maize geneticist Beadle into Drosophila genetics to work with Ephrussi, the Russian-born, Paris-trained experimental embryologist in genetics and his eventual link with the biochemist Edmund Tatum. Shortly after Morgan arrived at Cal Tech, Linus Pauling began to express an interest in biology and to participate in biology seminars. At a small organization like Cal Tech, scientists all across the campus were interested in what others were doing. And because of its culture of excellence, faculty assumed that all other faculty were outstanding people in their fields, and thus most people were anxious to learn a bout the work of their colleagues. Pauling was one of the leading theoretical chemists in the world. Because of his interest in molecules, he probably would have drifted into biology in any event. But by the mid-1930's, after encouragement from the Rockefeller Foundation, he began to work on the chemical bonding between the hemoglobin and oxygen materials, and later on the general problem of the structure of proteins. By the early 1950's, Pauling was producing reliable descriptions of the physical properties of giant protein molecules. Since Pauling was always interested in understanding chemical substances in terms of their protons and electrons, it is not surprising that he was the boundary crosser par excellence. One of the most important individuals in establishing the field of molecular biology, Pauling did work on the helical hypothesis which laid essential ground work for Crick and Watson's discovery of the double helix structure (Goodstein 1984; Watson, 1968; Kay 1993). It was the rich horizontal communication across fields at Cal Tech which was important in facilitating Pauling's discoveries in so many different areas. It is extremely doubtful had he been at a larger American research university, fragmented into many academic departments, that he would have had as much interaction with scientists in other areas and that his productivity and creativity would have been so considerable in so many different areas. To put it another way, the interdisciplinary and integrated culture of Cal Tech was critical to his thinking and to his achievements. Just as scientific integration was facilitated at the Rockefeller Institute by its dining room, similarly horizontal communication and scientific integration were promoted by Cal Tech's famous Athenaeum dining room. For many years, it was the place where faculty met, learned what others were doing, and often agreed to collaborate on research projects. The Athenaeum had an ambiance of comfort and prosperity. As at Rockefeller, many of the tables seated eight people, enough for a single conversation. The incentives for the faculty to go there were multiple: The food was outstanding, it was a very comfortable place to dine, it was a place to learn what the local faculty were doing, and to meet leading scientists from all over the world who were visiting Cal Tech. Despite all of Cal Tech's resources, divisions have from time to time fluctuated in quality. By the late 1940's, Morgan was no longer on the scene, Calvin Bridges was dead, and a number of Morgan's young associates had moved on. While Sturtevart and others were still there, the Biology Division did not have the same intellectual excitement and productivity as in the late 1930's. However, because of the interdisciplinary and cooperative culture ingrained at Cal Tech, the dip in the Biology Division's quality was a temporary phenomenon. Recruited to return to Cal Tech in 1946, George Beadle as leader of the Biology Division worked closely with Pauling to make Cal Tech again one of the premier centers of the world in the biological sciences. They recruited Max Delbrück, a physicist and a leader in phage genetics, and other distinguished scientists, so that Cal Tech soon became one of the world's premier centers in interdisciplinary biology with strength grounded in genetics but extending to physics and to various other fields of chemistry. Delbrück did much to perpetuate the nurturing environment which Morgan and others had developed. While Delbrück was hardly surpassed by anyone in setting rigorous scientific standards, his work groups, whether at Cal Tech or at Cold Spring Harbor, joined hard work with play. At Cal Tech, his group of students, postdocs, and colleagues frequently made trips to the desert for recreation, while at Cold Spring Harbor, there was swimming and tennis, as well as other play. Everyone was on a first name basis, but recognition was earned on the basis of excellence in science. Over time Cal Tech has continued to be distinguished in the biological sciences, and one of the most important reasons for this persistent ability to make major breakthroughs has been the integration of scientific diversity. Another factor has been the foundation on which the Biology Division was originally based, strength not just in genetics but in the genetics of Drosophila. And finally, the Institute has exercised unusual care in the making of tenured appointments. We have thus far identified only one person prior to 1950, who was granted a tenured appointment in the Biology Division without first having spent a prolonged period of time at the Institute. They did not recruit faculty without knowing a great deal about them. It is quite remarkable how the original strength in the genetics of Drosophila has continued to evolve, in substantial part because of the characteristics of the organization. The continued excellence in this research tradition has also been due to the strategic vision of Morgan of integrating genetics with other fields of knowledge. One impressive embodiment of these traditions has been Seymour Benzer. Originally trained in physics, he was inspired by Max Delbrück and moved into the fields of molecular biology and virus genetics, where he became famous for mapping the fine structure of genes. After discovering that exposure of a population of Drosophila to a mutagen generated an extraordinary variety of mutants affecting various aspects of behavior, Benzer and his associates identified mutations affecting visual reception, circadian rhythm, memory, excitable membrane channels, synaptic function, as well as hereditary pathological defects analogous to such human disorders as muscle and brain degeneration. As a result of Benzer's pioneering discoveries, Drosophila has become a model system for molecular neurogenetics, since his work revealed that genes so involved have identifiable close counterparts in the human cell, and allow the localization of genes responsible for various human hereditary defects. This has opened the way to using knowledge obtained from fruit flies to investigate human behavioral mechanisms and their disorders (Citations by Prize Committees for the Lasker Prize 1971; Horwitz Prize 1976; Wolf Prize 1991; Crafoord Prize 1993). Most small institutions, as implied here, have only a limited range (e.g., scope or diversity of activity) in the biological sciences, and this is one of the key reasons why their diversity does not become so highly differentiated. It is rare that the small institution has a great deal of depth in a particular area. But in the case of Cal Tech, it was the considerable depth in Drosophila genetics since the time of Morgan, embedded in a small institution where the drosophilists could frequently and intensely interact with scientists in other areas of science, that has facilitated this field in producing so many major discoveries. One of the most significant breakthroughs was by Ed Lewis who worked for more than thirty years in linking Drosophila genetics to embryology and evolution, work which eventually led to his receiving the Nobel Prize in 1995 (earlier he had won the Lasker and Wolf Prizes for this work). Lewis' work is very seminal in understanding how genes control development. Some observers herald Lewis' work as one of the most important discoveries of the twentieth century. It is significant that this work emerged at Cal Tech, the institution with more than half a century of experience in Drosophila genetics, where from the outset of biology there had been considerable diversity and interest in using Drosophila genetics to pursue problems of development (Kohler 1994). The small size of Cal Tech facilitated knowledge of development by a geneticist like Lewis, who learned about the chemistry of protein research of Pauling and others. In short, the structure and culture of Cal Tech facilitated the ability of its faculty to capitalize on its scientific diversity. (Interviews with Ed Lewis, March 25, December 21, 1994.). The interdisciplinary/integrated culture at Cal Tech's is well illustrated by the interaction among basic scientists, engineers, and computer scientists on campus. Two of its very distinguished biological scientists, William J. Dreyer and his now-famous student Leroy Hood, long expounded the thesis that what drives the pace of scientific progress is not so much the quality of scientific talent but the availability of new technologies, particularly new instrumentation which permits scientists to pursue new research strategies. And because of the ease with which biologists, engineers, and computer scientists could communicate at Cal Tech, the development of new instrumentation for scientific advance has been particularly notable there, especially in the biological sciences. In the early 1980's, some of the proteins of most interest to molecular biologists existed in millionths-of-a-gram quantities on the surface of cells, so that obtaining enough of a particular protein to conduct any research often required ten years and was very costly. Leroy Hood addressed the problem by constructing a machine that could work with extremely complex chemistries automatically and required only a tiny amount of protein. Hood and his co-workers developed a protein sequencer which had the ability to work with 1/10,000th of the amount of protein previously required. In one of the machine's early applications, Hood was able to determine the molecular structure of interferon. After developing the protein sequencer, Hood's transdisciplinary group went on to develop a DNA synthesizer which would create strands of DNA; a protein synthesizer for making protein in the lab; and a laser-equipped sequencer for determining the structure of DNA. The first gene synthesis required 20 assistants almost six years to conduct, whereas Hood remarked that in 1985 that his lab synthesized the same gene in less than a day. Some have compared the effect of these two sequencers and two synthesizers to the effect that cyclotrons and particle accelerators had on the field of high energy physics, permitting physicists to penetrate the basic structure of matter. These machines, in conjunction with other technological advances of the last decade such as recombinant DNA and monoclonal antibody techniques have allowed researchers "to manipulate and analyze genes and proteins in a way that was utterly impossible before" (Leroy Hood, quoted in 'Los Angeles Times,' Oct. 20, 1985: 56; interviews with Leroy Hood, July 29, 1995;. August 28, 1996). Significantly, Hood's work was complementary to that of fellow Cal Tech biologists Benzer and Lewis. Like Benzer, Hood also worked in the area of neurobiology and his research has also been quite fundamental to developmental biology, in particular in understanding how genes are regulated, how the nervous system functions, and how a fertilized egg becomes a complete organism. What was different about Hood's lab at Cal Tech was its size. By worldwide standards, and certainly by those at Cal Tech, Hood's lab during the 1980's and the early 1990's was extraordinarily large for a biology lab. The size of his operation was out of kilter with the norm of having small labs at Cal Tech. The friction between the norms of his lab and the norms of small labs at the Institute eventually became a contributing factor to his leaving Cal Tech and moving to a large public university, the University of Washington in Seattle. There he has established a new Bio-Molecular Technology Department, with people from genetics, medicine, engineering, and computer science. But it was Cal Tech which made it possible to develop the model of laboratory diversity which Hood values highly and which led to the new Seattle facility. Some observers believe that Cal Tech is too small to continue making major discoveries in the bio-medical sciences. Obviously, only time can tell. The suggestion of this project is that when scientific diversity is highly integrated and interdisciplinary, when there are leaders with a strategic vision, when there is a nurturing environment but with rigorous standards, and when there is capacity to identify and recruit individuals of considerable excellence, the organization will continue to have high potential to make major discoveries. Significantly, Cal Tech in the 1990's is continuing to demonstrate flexibility and adaptiveness to the fast-paced work of scientific change. In part, it is doing this by developing new structural arrangements, one interesting example being the Beckman Institute. The vision driving Cal Tech's Beckman Institute is that within the next several decades, chemistry, biology, and medicine will increasingly converge as fields of knowledge. The Director of the Beckman Institute, Harry Gray, Professor of Chemistry, views all of this in the tradition of Linus Pauling, and the inclusive view of knowledge. His view is that a science should embrace as its own the diverse disciplines to which it contributes. Gray and his colleagues argue that there is a total restructuring of the field of chemistry taking place as a result of new tools available to chemists. Others at Cal Tech argue, also in the tradition of Pauling, that in terms of theory, there is a narrowing of differences among the fields of biology, materials science, polymer science, and chemistry. And at the Institute, there is increasing emphasis on the necessity of making fundamental changes in the way that the next generation of scientists will be trained. According to Peter Dervan, Professor of Chemistry at Cal Tech, it is necessary for scientists to break out of their disciplinary boundaries and put people with different specialties next to each other so they can get inside one another's minds and learn from each other. Large Research Organizations and Problems of Making Major Discoveries As research organizations increase in diversity and depth, there is a tendency for them to experience more differentiation and less integration. These changes often lead to increases in hierarchical and bureaucratic coordination, with negative consequences for making major discoveries. Changes in biological and medical knowledge have marked consequences for the diversity and depth of research organizations, requiring specialties that reflect these new areas of research if the organization is not to become anachronistic. Thus, universities, research institutes and medical schools added bio-chemistry once the field developed. Genetics, bio-physics and various medical and surgical sub-specialties became attached to medical schools and other research organizations across the years, usually requiring the hiring of several, or even many people so that there was enough depth in an area. New kinds of instrumentation and other technologies also require the hiring of personnel. But to add personnel in new specialties, to build diversity with the required spread of talents, and to create depth in each area implies a growth in the size of the research organization. The problem is how the organization incorporates the expansion of knowledge and growth in size. When research organizations respond to growth by differentiating into new departments and by imposing hierarchical and bureaucratic controls, these processes lead to a decline in integration and diminish the possibility of making major discoveries, even if organizations become highly productive in the number of scientific papers produced. Examples of research organizations that have responded to changes in biological and medical knowledge by differentiating into new departments and growing in size are found in many large American research universities (e.g., Illinois [Champaign-Urbana], California [Berkeley], Minnesota, Michigan) and many medical schools (e.g., the schools attached to UCLA, Yale, and University of Pennsylvania). Many of these were places that appeared at certain moments of time poised to be the contexts of major discoveries, but such discoveries did not occur. It is worth stressing that after World War II, these universities and medical schools had large amounts of research funding, were highly productive, and had a number of scientists in the National Academy of Sciences. But they did not have the structural and cultural contexts suited for major discoveries or breakthroughs in the bio-medical sciences. Significantly, by the late 1980's and early 1990's, the University of California, Berkeley, recognized that the fragmented and differentiated structure of the biological sciences substantially hampered the quality of biological research, and it restructured the biological sciences, abolishing more than a dozen departments in the process. After years of attempting to restructure the biological sciences, today, Berkeley finally has a much more integrated biology program and has a structure which provides the potential for major discoveries to occur. Differentiation, size, and bureaucratization are also variables which place constraints on organizational flexibility. Science is very dynamic, and if research organizations are to adapt to the changes in the world of science, they must have a structure which is highly flexible. It is flexibility in adapting to change in science which provides the potential for making major discoveries across time. The lack of flexibility has been a problem with most of the nation's medical schools. At the time many medical schools came into existence, they were dominated by clinicians, especially the departments of medicine and surgery. Most medical schools have been very sharply differentiated between the clinical sciences and basic sciences. And therefore it has been difficult for the basic science departments in many medical schools to attain the autonomy and organizational environment necessary to become distinguished. Thus, although the medical schools of the universities of Michigan, Minnesota, Pennsylvania, and Wisconsin have been either very strong or quite good over time in the clinical sciences, they have produced either no or very few major discoveries in the twentieth century. Strategies for Adding Diversity and Depth and for Enhancing Scientific Integration in Large Research Organizations Another set of findings in this research relate to different strategies for adding diversity and depth without necessarily differentiating into more departments, even in large research organizations. Since growth in size leads to the differentiation of departments and hierarchical and bureaucratic coordination, and this diminishes social integration and the capacity to make major discoveries, the problem is how organizations can respond to changes in knowledge by increasing diversity and depth without growing in size. We have found a number of interesting strategies organizations pursued, several of which are briefly discussed:
The first strategy, of limiting growth while emphasizing diversity, has been chosen by a few private research universities. Departments, seeking to increase depth and scope in their discipline, attempt to add staff and consequently get larger, often recruiting staff more like themselves in the process. An interesting contrary example is Harvard, where a major function of the President since the presidency of James Conant is to convene an ad hoc committee of distinguished scholars to assess every tenured appointment in the College of Arts and Sciences. These ad hoc committees have frequently contested the judgment of academic departments, repeatedly vetoing a tenured appointment previously approved by a department. This process, by contesting tendencies among departments to reproduce themselves, has increased diversity and has facilitated Harvard's flexibility in adapting to the larger world of science, giving it a distinct margin of advantage over other universities. A second strategy is either to create an integrated program that stresses diversity, depth, and integration or to prevent the processes of differentiation by pursuing the same goals within a single department. Two illustrations of this strategy are the biology department at MIT and the basic sciences at the University of California at San Francisco, perhaps the two leading contemporary institutions in the biological sciences in the United States, if not the world. Another strategy is to create a small interdisciplinary research institute or center attached to a university or medical school, an institute which is largely autonomous and is high in diversity, depth, and integration. The University of Wisconsin followed this strategy for several decades with success, with the Enzyme Institute and the McArdle Cancer Institute. Both were small institutes, which allowed for a highly focused and reflective attack on research, and permitted intense and frequent interaction among faculty from diverse backgrounds who had full time appointments major discoveries occurred in these institutes but not in the departments of the university. Conclusion Organizations with major discoveries recurring again and again have tended to be those in which there is a high degree of interaction among scientists across diverse fields of science. Over time, biomedical knowledge has become increasingly complex (e.g., involving both more fields of knowledge and greater depth), and if research organizations are to make major discoveries, it is necessary for them to incorporate new knowledge and depth in new fields of knowledge in such a way that scientists can interact with intensity and frequency across diverse fields of knowledge. However, as research organizations add new fields of knowledge and depth, it is important that the parts of the organization be well-integrated and not highly differentiated from one another; otherwise there will not be the horizontal communication with frequent and intense interaction across diverse fields that is prerequisite if major discoveries are to occur. Research organizations where a number of major discoveries have occurred have had a distinctive style of leadership, that is leaders who have had
Nurturing here is defined as a combination of activities by leaders: rigorous and thorough review (carried out with sensitivity) of work of those in the work group, stimulation of new ideas and work arrangements, sometimes shielding researchers from outside criticism, willingness to be patient in waiting for results, and interacting socially with those in the work group. The diversity of disciplines and depth of knowledge within a well-integrated research group or lab have the potential to change the way people view problems and to minimize their tendency to make mistakes and to work on trivial problems. In the final analysis, scientists, in order to make major discoveries, must work on significant problems which are "do-able." And the greater the research group’s diversity and depth, the greater the likelihood that scientists will not stray into unproductive areas. If scientists work in environments where there is diversity across disciplines and depth, and have frequent and intense interaction with those having complementary interests, the probability is increased that the quality of their work will improve. It is the diversity of disciplines and paradigms to which individuals are exposed with frequent and intense interaction that increases the tendency for creativity, and for breakthroughs to occur. Working in an interdisciplinary environment without intense and frequent interaction among members of the work group does not tend to lead to new ways of thinking (e.g., major discoveries). Changes in biological and medical knowledge have marked consequences for the diversity and depth of research organizations, requiring new specialities if the organization is not to become anachronistic. As knowledge expands, new disciplines and sub-specialities come into existence, and there are pressures on the organization to appoint several or even many people in the new fields of knowledge so that there is sufficient depth in an area. New kinds of instrumentation and other technologies also require the hiring of personnel. But the addition of personnel in new specialities with the required spread of talents, and the addition of depth in each area, imply a growth in the size of the research organization. Increases in scientific diversity and depth, if not properly managed, can ultimately limit the capacity of a research organization to make major discoveries. As research organizations increase in the number of disciplines and expand their depth in each one, there is a tendency for them to become more differentiated and less integrated. These changes are often accompanied by increases in hierarchical coordination and bureaucratization, with negative consequences for making major discoveries. Over time, as research organizations have differentiated into more departments and other sub-units, recruitment and the search for additional funding have been delegated to lower levels. Academic departments are inherently socially conservative and tend to select people who reproduce their thinking. Thus differentiation has tended to limit the process of crossing academic disciplines, and to hamper scientific integration, both very important for major discoveries in biomedical science. More specifically, increases in size and the decentralization of research and personnel decisions to lower units such as departments lead to more bureaucratic rules and the use of budgetary controls. And the formalization of rules and increases in structural differentiation tend to decrease the frequency and intensity of interaction across specialities and thus social integration. In sum, this research adds to the theoretical literature the finding that increases in size lead to differentiation and to fewer major breakthroughs in bio-medical science. Acknowledgments We are deeply indebted to Ragnar Björk, Jerald Hage, Gerald Edelman, Peter Weingart, Nico Stehr, and Julie Klein for their contributions to this paper. Julie K. Sweeney of the University of Wisconsin patiently and thoughtfully assisted with data for the project. Archivists at the Karolinska Institute, California Institute of Technology, the Royal Swedish Academy of Sciences, and the Rockefeller Archive Center provided very valuable assistance. For their financial support, we thank the Rockefeller Foundation, the Sloan Foundation, the Andrew W. Mellon Foundation, the Swedish Council for Higher Education, and the University of Wisconsin. References* Allen, G. E. 1978a. 'Opposition to the Mendelian-chromosome Theory: The Physiological and Developmental Genetics of Richard Goldschmidt.' Journal of the History of Biology 7: 55-87. Allen, G. E. 1978b. Thomas Hunt Morgan: The Man and His Science. Princeton: Princeton University Press. Allison, P.D., and Scott Long, J. 1990. 'Departmental Effects on Scientific Productivity.' American Sociological Review 55: 469-478. Ben-David, J. 1960. 'Scientific Productivity and Academic Organization in Nineteenth Century Medicine.' American Sociological Review 25: 828-843. Ben-David, J. 1971. The Scientist's Role in Society: A Comparative Study. Englewood Cliffs, NJ: Prentice-Hall. Ben-David, J. 1977. Centers of Learning: Britain, France, Germany, United States. New York: McGraw Hill. Corner, G.W. 1964. A History of the Rockefeller Institute. New York: Rockefeller University Press. Dasgupta, P.,and David, P.A. 1993. 'Toward a New Economics of Science.' MERIT Research Paper 94-003. Maastricht: Maastricht Economic Research Institute on Innovation and Technology. Dubos, R.J. 1976. The Professor, the Institute, and DNA. New York: Rockefeller University Press. Flexner, A. 1930. Universities: American, German, English. Oxford, UK: Oxford University Press. Franzosi, R. 1998. ‘Narrative Analysis - Or Why (And How) Sociologists Should Be Interested in Narrative.’ Annual Review of Sociology 24: 517-554. Fujimura, J. 1987. 'The Molecular Biology Bandwagon in Cancer Research: Where Social Worlds Meet.' Social Problems. 35: 261-283. Gibbons, M., C. Limoges, H. Notwotny, S. Schwartzman, P. Scott. 1994. The New Production of Knowledge. London: Sage Publications. Hollingsworth, J.R., and Hage, J. 1996. 'Organizational Characteristics Which Facilitate Major Discoveries in the Bio-Medical Sciences.' Document presented to Science and Technology Studies Program, National Science Foundation. Kay, L.E. 1993. The Molecular Vision of Life: Caltech, the Rockefeller Foundation and the Rise of the New Biology. New York: Oxford University Press. Kohler, R.E. 1994. The Lords of the Fly. Chicago: University of Chicago Press. Latour, B. 1987. Science in Action. How To Follow Scientists and Engineers Through Society. Cambridge: Harvard University Press. Latour, B., and Woolgar, S. 1979. Laboratory Life: The Construction of Scientific Facts. Princeton, NJ: Princeton University Press. Los Angeles Times, October 20, 1985, January 12, 1992. Lynch, M. 1985. Art and Artifact in Laboratory Science: A Study of Shop Work and Shop Talk in a Research Laboratory. London: Routledge and Kegan Paul. Lynch, M. 1993. Scientific Practice and Ordinary Action: Ethnomethodological and Social Studies of Science. Cambridge: Cambridge University Press. Merton, R. 1961. 'Singletons and Multiples in Scientific Discovery.' Proceedings of the American Philosophical Society 55: 470-86. Merton, R. 1973. The Sociology of Science: Theoretical and Empirical Investigations. Chicago: University of Chicago Press. Olby, R. 1979. The Path to the Double Helix. Seattle: University of Washington Press. Pelz, D.C., and Andrews, F.M. 1966. Scientists in Organizations: Productive Climates for Research and Development. New York: John Wiley. Polanyi, M. 1966. The Tacit Dimension. London: Routledge and Kegan Paul. Rheinberger, H-J. 1997. Toward a History of Epistemic Things: Synthesizing Proteins in the Test Tube. Stanford: Stanford University Press. Rosenberg, N. 1994. Exploring the Black Box. Technology, Economies, and History. New York: Cambridge University Press. Servos, J.W. 1990. Physical Chemistry from Oswald to Pauling: The Making of Science in America. Princeton: Princeton University Press. Shapin, S. 1995. 'Here and Everywhere: Sociology of Scientific Knowledge.' Annual Review of Sociology 21: 289-321. Watson, J.D. 1968. The Double Helix. New York: Athenaeum. Zuckerman, H. 1977. Scientific Elite: Nobel Laureates in the United States. New York and London: Free Press. * Extensive use has been made of archival materials, interviews conducted by the authors, and oral histories of scientists. Space constraints prevent the listing thereof. |
