Research and Science Politics
NOT all of my time after I came to Carnegie Mellon was devoted to organization and institution building. This chapter is devoted to a discussion of our research there and my professional activities outside the university.
STUDYING ORGANIZATIONS
During the first six years of my research at GSIA, I filled out, empirically and theoretically, the decision-making framework I had laid down in Administrative Behavior. My files yield a planning document that Harold Guetzkow and I wrote on February 28, 1952, probably the basis for a grant proposal, outlining a five-year plan for research in organizations. With Administrative Behavior as its theoretical starting point, it proposed field and laboratory research as well as theoretical studies, emphasizing the need to bring together empirical findings from many sources, not just our own work, in order to build theory.
The most interesting substantive recommendation in the document was that decision making in organizations should be related to learning theory: “Our work has led us to the conclusion that there is an intimate connection between organizational structure and the learning of frames of reference and roles by members of organizations.” The source of the idea could have been a combination of Harold Guetzkow’s previous psychological research on the topic of “set,” or frame of reference, and my experience in helping organize the Economic Cooperation Administration a few years previously. The idea foreshadowed the critical importance of reference frames—we would now call it “problem representation” or the “frame problem”—in problem solving and learning. Problem representation is still high on the agenda of research in cognitive science today, thirty-nine years after the date of that memorandum.
Even before our planning memorandum was written, we had begun our first big empirical study, the “Controllership Study” (Simon et al. 1954). It took us into factories and sales offices to see what use was made by operators of blast furnaces and sales managers, as aids to their decision making, of the company accounting and cost accounting records and of the services of the accountants.
The Controllership Study was an adventure to me, both watching George Kozmetsky (whom we hired from the Harvard Business School as a young Ph.D., and who already had extraordinary facility in analyzing accounting records) extract information from our respondents, and trudging through the reddish brown dust of the National Works of U.S. Steel in McKeesport to learn how a steel mill was actually managed and how its decisions were made.
In the Controllership Study, as in the later empirical work on organizational decision making we did in GSIA, I worked with a great many colleagues, the nature of the collegial relation being a mixture of those I had experienced in Berkeley and at Illinois Tech. As at Berkeley, a number of my co-workers were young academicians who had joined our group with the understanding that they would be associated with such research projects. I was their colleague, but also their project leader. Of my co-authors during this period, eight fall in this category.
Four other co-authors were more senior faculty colleagues, my relations with them being similar to my relations with Don Smithburg and Vic Thompson, with whom I wrote the public administration book at IIT. In addition, during the first five years at GSIA, I published papers with eight students, only two of whom (Ed Feigenbaum and Allen Newell) were my own doctoral students. Following the general custom of that time in political science and economics, I did not generally co-author papers based on the doctoral dissertations of my students, and even in recent years have done so only half a dozen times, and at the student’s invitation.
At Carnegie, the project team usually met weekly to report progress, assign tasks, and, most important, to discuss the research. In the project on organizational decision making, we recorded and circulated minutes of these meetings, and in all the projects members of the team prepared frequent working papers, submitting their ideas to criticism and discussion with their colleagues.
Outside the projects, my collaborations have usually involved a single person, occasionally as many as three or four. Again, we hold regular weekly meetings, as I do with my graduate students. With individuals, the meetings are usually scheduled for an hour, with working groups, they may last several hours. Each member of the collaboration spends much more time in individual empirical or theoretical work between meetings than at the meetings.
Initial drafts of papers that are ultimately published may be written by any of the collaborators. Since I write easily and fluently, I probably do more than my share of the writing, and certainly of the final editing, of papers of which I am co-author. Names on publications are alphabetical, unless one or two members of the team are clearly principal authors.
Throughout the rest of my career, I have continued to work with many students and faculty colleagues—I have had more than eighty co-authors in all—but not usually within the framework of large empirical projects involving many people. The years at Berkeley and the first five years at Carnegie were my main experiences of large-scale empirical field work.
The Controllership Study was followed by a Ford-funded project that involved detailed case studies of specific situations in companies that gave us access to their decisions in the making. Both Dick Cyert and Jim March were attached to that project, and in that way began the collaboration that led to their pathbreaking book, The Behavioral Theory of the Firm, in 1963.
In the minutes of the working sessions in which Harold Guetzkow, Jim March, I, and others participated, our language for describing decision-making processes developed in an interesting way. We had long talked about “decision premises” and “islands” of such premises that could be communicated from one manager in an organization to another as inputs to decisions. When Harold brought to these sessions his earlier research on problem solving, we began more and more to see decision-making processes as essentially the same as problem-solving processes. My own economic theorizing was leading me in a similar direction. Hence, as early as 1951, the language of problem solving began to creep into the minutes of our working sessions.
Our growing interest in problem solving led us to restudy the writings of psychologists who had done research on that topic, especially Gestaltists such as Norman Maier, Max Wertheimer, Karl Duncker, and George Katona. This was a first tentative step on the road that led soon afterward to my collaboration with Al Newell and Cliff Shaw in building a computer simulation of heuristic problem solving.
Jim, Harold, and I also undertook to produce a “propositional inventory” of organization theory that Bernard Berelson of the Ford Foundation commissioned. Jim March and I co-authored, with Harold’s help, a book called Organizations (1958), which represents our interpretation of that assignment. It was successful in systematizing organization theory, less successful in marshaling the empirical evidence to support its propositions.
In particular, though we were acutely sensitive to the need for using field research to test and extend our theories of organizational behavior, we did not know how to compare information from diverse case studies that used relatively informal and wholly unstandardized methods for gathering and analyzing data. Much research effort was subsequently aimed at that problem at Carnegie, but we cannot say that we learned to handle all, or most, of the methodological difficulties. Perhaps our foremost contribution has been to show how human thinking-aloud protocols can be used as objective data, especially in conjunction with computer simulations.
Berelson did not think our book contained much of a propositional inventory, but the theory that it does contain has aged well. Jim and I have talked a few times about revising it, but have agreed that it would be difficult to come together again from our diverging professional paths. (Perhaps each of us should issue an independent second edition and see which sells better!)
Harold left in 1957 for other, presumably greener pastures, and Jim followed in 1964. After 1956 I was considerably diverted by the new computer simulation work. In 1961, Dick Cyert became dean of GSIA, greatly curtailing his research time. Since our research on organizational decision making was nearly unique, at least in the United States, we were generally unsuccessful in recruiting replacements for the original quadrumvirate who had provided the leadership. They simply were not being produced by other Ph.D. programs.
By the early 1960s, the Golden Age of organization theory and the behavioral theory of the firm had ended at Carnegie Institute of Technology. As we shall see, GSIA came to be dominated by research on sophisticated mathematical techniques in operations research and economics and by neoclassical economic theory. The economists’ aborted revolution of 1951 achieved a large measure of success in the 1960s.
THE MATHEMATICAL SIDE
In my management science and econometric research, I retained my consulting ties and close contacts with the Cowles Commission and, through it, with the RAND Corporation, then enjoying tremendous success and visibility as a new way of enlisting research talent to help advance applied goals. Beginning in 1952, as I discussed earlier, I made frequent trips to RAND in Santa Monica and spent a number of summers and one full year (1960–61) there during the following decade.
At Carnegie Tech, my management science research, as distinguished from my research on organizational behavior, was part of Bill Cooper’s project, as explained in chapter 9. The plans for that project were laid out in a basic planning memorandum dated February 21, 1950, a year and a half prior to the crisis I have told about. Under the heading of “production technology,” both the Cooper-Charnes linear programming work and the Holt-Modigliani-Muth-Simon*1 factory scheduling research were foreshadowed.
There was also a good deal of emphasis on the effect of financial data on decisions, an area of special interest to Bill Cooper. This led ultimately to such products as Charles Bonini’s computer simulation of a business firm, which demonstrated how changes in the accounting system could trigger changes in operating decisions.
My own mathematical and econometric work at this time had a number of themes, both substantive and methodological. A good overview of it can be found in Models of Man (1957a), a collection of some of the papers I wrote during my first five years at Carnegie. Part I contains papers on causality, deriving from my work with the Cowles Commission on the identification problem. Part II undertakes to show the utility of making mathematical translations of several current theories of social interaction (the theories of George Homans and Leon Festinger). Part III proposes several models that explicate how Barnard’s and my organization theory can be related to the economic theory of the firm. Part IV formalizes and explores the concept of bounded rationality. Four chapters deserve special mention for their influence on subsequent events in my life and in the economics profession.
Chapter 14 of Models of Man, “A Behavioral Model of Rational Choice” (first published as 1955a), mostly written in 1952 during my first RAND summer, represents my first major step toward formalizing the psychological theory of bounded rationality. Although the term satisficing, a key concept in my subsequent work, was not used in it, the satisficing concept—searching for “good enough” actions rather than optimal ones—is already present. Of all my writings on this topic, the “Behavioral Model” paper comes closest to the mathematical format with which economists are comfortable. Hence, economists who wish to refer to bounded rationality and satisficing most commonly choose this paper for citation.
What made the paper distinct from most contemporary economic writing was its explicit concern for the process of making decisions, for procedural and not just substantive rationality. Because of its concern with process, the paper also represents a first step toward computer simulation of human behavior. The manuscript contained an appendix, never published except as part of a RAND technical report, outlining how a chess-playing program for a computer could compensate for its bounded rationality by using selective search guided by heuristics (see note 4 of the published paper). This proposal for a heuristic chess program is an extension of ideas that Claude Shannon had published in 1950, but my emphasis was on human simulation rather than chess prowess.
What motivated this part of the work was a talk that John von Neumann gave at the 1952 RAND summer seminar, emphasizing the difficult problems that had to be solved in order to program a computer to play good chess. I thought von Neumann overestimated the difficulties substantially, and moreover I believed that I had some solutions for them which I proposed in the appendix to the RAND technical report. This chess discussion disappeared from the final manuscript of my paper. Perhaps referees thought it irrelevant or perhaps I excised it when Al Newell began to carry the chess ideas much farther as he developed his 1954 paper on the subject.
Chapter 15 of Models of Man, “Rational Choice and the Structure of the Environment” (first published as Simon 1956), was a companion piece to chapter 14. Adopting again a satisficing point of view, it provides a sort of Darwinian model of rationality. Bracketing satisficing with Darwinian may appear contradictory, for evolutionists sometimes talk about survival of the fittest. But in fact, natural selection only predicts that survivors will be fit enough, that is, fitter than their losing competitors; it postulates satisficing, not optimizing. The paper showed how relatively simple choice mechanisms could enable an organism, searching through its life’s maze, to survive in an uncertain environment in which several incommensurable needs had to be met. It depicted a procedural rationality for organisms that was squarely based on satisficing rather than optimizing.
Chapter 15 had a curious by-product—the only short story I have ever written. Since no one has ever told me that it has any literary merit, it should probably be read as philosophy, not literature. It uses the metaphor of the maze to explore the relation between satisficing and basic human values, a free translation of chapter 15 from model to metaphor; I will recount it in the next chapter.
Chapter 11 of Models of Man, “A Formal Theory of the Employment Relation,” was a harbinger of the New Institutional Economics that has been nurtured in recent years by Oliver Williamson and others. In it I attempted to cast some aspects of organization theory in the mold of neoclassical reasoning. In this sense it is reactionary, a throwback. In my 1977 Ely Lecture before the American Economic Association, I apologized for it thus:
In my 1951 paper, I defined the characteristics of an employment contract. . . . My argument requires a theorem and fifteen numbered equations, and assumes that both employer and employee maximize their utilities. Actually, the underlying functional argument is very simple. . . . The rigorous economic argument, involving the idea of maximizing behavior by employer and employee, is readily translatable into a simple qualitative argument that an employment contract may be a functional (“reasonable”) way of dealing with certain kinds of uncertainty.
Thus, most of my theory of the employment contract can be expressed without either equations or maximization. But in that qualitative form it would not have captured the attention of economists, who, in the “new institutional economics,” continue to pour the new wine into the old bottle of neoclassical reasoning. At least we have some new wine.
Chapter 13, “Application of Servomechanism Theory to Production Control,” started Charles Holt and me off on our operations research venture in scheduling a paint factory. More accurately, it led us to a new operations research tool for dealing with a broad class of dynamic planning problems under uncertainty, for which the paint factory provided an application.
By making strong approximating assumptions about costs, we were able to solve an exact maximization problem with little computation. That is to say, we satisficed by finding the optimal policy for a gross approximation to the real world. The dynamic programming project later was joined by Franco Modigliani, and by several graduate students, including John Muth. Its main direct product was the Holt, Modigliani, Muth, and Simon book, Planning Production, Inventories, and Work Force (1960). It also had an indirect product—rational expectations—which belongs to the next segment of the chronicle.
THE SYSTEMS RESEARCH LABORATORY
Apart from the consulting relation that took me to the RAND Corporation in 1952, another aspect of my association with RAND from 1952 to 1955 has a great bearing on what followed.
During the GSIA period up to 1952, while we were busy with our studies of decision making in organizations, four people—John Kennedy, William Biel, Robert Chapman, and Allen Newell (all in the Santa Monica part of the forest)—conceived the grand, or grandiose, design of studying in the laboratory the behavior of an air defense organization. The laboratory (christened SRL, the Systems Research Laboratory) would simulate an entire air-defense early warning station, staffed with perhaps fifty men. The U.S. Air Force was to supply the budget and the airmen-subjects for the simulation.
When, in the planning stages of this imaginative project, the group came to me for advice, because of my previous experience in conducting the California SRA experiment and other organizational studies, I became a consultant; and, on my first visit to RAND early in 1952, I met Allen Newell. I had not known him previously, but I was familiar with a couple of mathematical documents he had written, in which he tried to formalize organization theory. I was only mildly impressed by the mathematics, which seemed to me to contain more definitions than theorems (always a bad sign for formal theories), but I was well disposed to anyone who had the disposition and skill for applying mathematics to these kinds of questions.
In our first five minutes of conversation, Al and I discovered our ideological affinity. We launched at once into an animated discussion, recognizing that though our vocabularies were different, we both viewed the human mind as a symbol-manipulating (my term) or information-processing (his term) system.
The SRL experiments provided the most microscopic data one could want on how radar operators and air controllers made their decisions. A vast body of such data was accumulated from a series of experiments conducted over three years, during which all of the communications between subjects were recorded. Yet Al and I suffered from continuing frustration in trying to write formal descriptions of the process. Somehow, we lacked the necessary language and technology to describe thinking people as information processors. As a result, we were never able to produce a good analysis of the phenomena we observed in the SRL experiments, and only one relatively innocuous paper was ever published about them.*2
The frustration that Al and I experienced with the SRL experimental data had major consequences, the first of which is reported in chapter 12 of this volume. In simplest terms, it determined the rest of my life. It put me in a maze I have never escaped from—or wanted to.
THE POLITICS OF SOCIAL SCIENCE
My gravitation toward the sun soon led me into professional activities in political science, and in social science generally, outside the university. Since my involvement in science politics began while I was at Illinois Tech, I will return briefly to that scene, beginning the tale just at the end of World War II.
In view of my education as a political scientist, it is perhaps not surprising that I devote considerable space here to the politics of science. My activities in this domain divide into two parts, the first concerned with the social sciences, the second with the relations between the social and natural sciences. I tell the social sciences part of the story here, the social and natural sciences part in chapter 19, as it relates more closely to events that took place after my election to the National Academy of Sciences in 1967.
Political Science
I entered into the world of science politics as a political scientist—that is to say, into a world quite segregated from natural science (“real” science, some would call it). During my tenure at Illinois Tech, I was active in the American Political Science Association, as one of the Young Turks fomenting the behavioral revolution in that discipline whose origins were recounted in chapter 4. Administrative Behavior and the published articles that had preceded it gave me some standing among the insurgents, but we had little success in finding a credible senior to lead us (V. O. Key, who was our candidate, was a gradualist, more or less committed to behavioralism, but not to revolution).
All I now recall of this activity was meeting with other conspirators in my hotel room at the (then new) downtown Washington Hilton during an APSA convention; and defeating the re-election of the incumbent APSA secretary at a later APSA meeting in Chicago. Our goal was to make sure that research within the behavioral framework would be received sympathetically by the American Political Science Review and would have an appropriate place in the annual meetings of the association. The behavioralists gradually captured the association, but it was the passage of time and the retirement of the old guard rather than revolutionary activity that accomplished the change.
About 1951 Bernard Berelson came to consult me about the prospective program of the new Behavioral Science Division of the Ford Foundation, including a very tentative plan for a training center. Arnold Thackray (1984), in a historical account of the Ford program, notes that I was early designated as a potential adviser, possibly because I was acquainted with Don Marquis of the University of Michigan, one of the framers of the program.
My first advice to Bernie, given at a meeting he had called in Chicago, was that the Behavioral Science money of the Foundation should be used for research grants, and not for such boondoggles as a training center. The most useful role for the program, I thought, was to fill the gap that had been left by the omission of the social sciences from the charter (and hence the funding) of the National Science Foundation.
The advice I offered at the Chicago meeting was followed up by a long letter of December 10 to Berelson. I argued that “the rest of the Ford Foundation program is so heavily oriented toward application that Program Five resources should be jealously guarded for fundamental work.” I again expressed my opposition to the training center idea, opining that it would be much better to distribute the money among three or four institutions, to free a few senior faculty in each for research and working with talented advanced students. These schools should be required to cooperate so that promising graduate students could divide their doctoral work among several institutions, and the Foundation should “season the brew with a competitive fellowship scheme at the undergraduate and graduate levels.”
Expressing a concern I had long had that students are seldom exposed to the social sciences before fixing on a career in natural science, I proposed making grants to permit a few institutions “to experiment with procedures for attracting to the social sciences at an early stage in their training a few men with outstanding research potential.” I also commented on the Foundation’s proposal to support interdisciplinary research:
The problem is less one of bringing unlike social scientists together than one of bringing unlike social sciences together in one man. There has been failure after failure of interdisciplinary “teams” to integrate anything . . . except to the extent that individual team members became interdisciplinary. I would not give a dollar to assist a typical political scientist to collaborate with a typical economist unless each one of them gave me a sworn statement that he would study seriously and not in a dilettante’s way the discipline of the other for at least a year.
After more than three pages of such criticism and comments, I expressed the
hope that, in spite of these points of disagreement, I conveyed to you in Chicago my enthusiasm for the general objectives and emphases you have pointed up in the program. . . . [I feel we have] the same basic convictions as to . . . the job . . . to be done: to provide the trained human talent that is needed for the progress of the behavioral sciences, to facilitate the cumulative and interactive development of theory and empirical research, and to build research progress around the core disciplines . . . represented by social psychology and sociology.
But the Ford Foundation was already more than half committed to creating a Center for Advanced Study in the Behavioral Sciences (CASBS), so I was soon a member of the committee that was helping Berelson to plan it. The idea of such a center had emerged from the fertile mind of Paul Lazarsfeld of Columbia University (a long-time associate of Berelson), who conceived of it as a place where young postdocs would come to sit at the feet of the masters and learn good methodology.
As this seemed to me (and others) a terrible idea, we gradually turned the plans around to eliminate the proposed bimodal distribution of “juniors” and “seniors,” and substitute a cadre of scholars distributed over the whole range of age and seniority. The Center would be a place for research and writing, rather than for postdoctoral training, and it would recruit the best social scientists it could find, who would spend a year there, preferably at a time when they had completed a major piece of research and were in the reflective and writing-up stage.
Part of my reason for opposing the original plan were my doubts that the elders were the persons best equipped to teach good social science method, and that the postdoctoral “juniors” would want to continue to sit at their feet (I was about thirty-five at the time). Perhaps more important, I thought that there was no lack of social scientists who knew how to do good empirical and theoretical research. What was lacking was money to support it. If there was to be a Center at all, its resources should be devoted, so far as possible, to subsidizing research. This change in the plan for CASBS destroyed Paul Lazarsfeld’s dream, and although we remained friends to the end of his life, he never quite forgave me my part in its destruction.
On another issue I was less persuasive. After several possible sites for the Center had been considered, the hill above the Stanford campus was selected—a beautiful spot in Lotus Land that I would not have chosen to enhance productivity. This had the unforeseen (or was it unforeseen?) consequence that Stanford University was able in the 1950s and 1960s to assemble a formidable social science faculty from California-smitten veterans of the Center.
The main consequences for my personal career of this involvement with the Ford Foundation and CASBS was that I became visible throughout social science circles and acquired an enormous amount of information about the current research picture, especially the names and numbers of all the players. Another consequence was that, by virtue of my strong opposition to the Center, I always felt embarrassed at the idea of spending a year there, and never did. (There were other reasons for my reluctance, too—notably its lack of suitable computing facilities.)
The Social Science Research Council
During the 1950s I was invited to membership on a couple of committees of the Social Science Research Council (SSRC), an organization that had been created in the 1920s by Charles Merriam and others to encourage interdisciplinary research in social science. A Committee on Business Enterprise Research brought together economists interested in the business firm with sociologists and others interested in organizations. Some years later, a committee of three (Carl Hovland, George Miller, and myself) was set up to administer a small grant from the Ford Foundation for what we would now call cognitive science.
In 1958, I was invited to join the Council’s Board of Directors, continuing to serve on it until 1971 (incuding five years as chairman). During most of that time, Pendleton Herring was president of the council. At the top of my agenda for SSRC I placed the support of mathematical training for social scientists and the erosion of boundaries between disciplines.
As the council had been pioneering since 1952 in activities aimed at raising the level of mathematical sophistication in social science, on this topic I simply supported what was already going on and encouraged the expansion of mathematical to include computer simulation. The council’s committee on mathematical training for social scientists was later spun off as an independent entity, mostly funded by the National Science Foundation. This committee, over the years, was highly effective in diffusing mathematical skills, particularly in sociology and political science (mathematization was already proceeding apace in economics).
The encouragement of interdisciplinary activity had been a principal motivation for founding the Social Science Research Council. But at the meetings of the council, I was appalled at how often I heard such phrases as, “as a historian, I . . . ,” “as an economist, I . . . ,” “as a sociologist, I . . . ,” and so on. I challenged these phrases each time I heard them, but it was like trying to purge ainnuh (“Isn’t it so?”) from the lexicon of a native of Milwaukee.
My free-floating experiences at the University of Chicago, and my many interdisciplinary contacts at the Cowles Commission and RAND, had not prepared me for the fierce disciplinary loyalties I encountered in the heart of the social science establishment. I came to see that disciplines play the same role in academe as nations in the international system. Academicians typically live out their whole careers within the culture of a discipline, rarely shaking off the parochialism this isolated existence engenders. (Still later I learned from my encounters with economics that disciplines undertake imperialistic adventures with the same zest that nations do.)
As best I could, I encouraged those activities of the council that cut across boundaries, and questioned activities that seemed more appropriate for the professional associations of the disciplines. My efforts had, at most, modest success. In 1966, the year before I took over the chair of the council, there were ten committees that were genuinely interdisciplinary, but at least six whose activities lay within single disciplines. That ratio didn’t change much over the years of my activity in the council.
But there was a bigger problem. The Ford Foundation had become deeply enamored of area studies, which produce experts in Russian or East Asian or Latin American or African affairs. Almost all of its grant money, and especially its fellowship money, was being diverted in this direction. Funds for the council’s traditional fellowships for interdisciplinary training had almost dried up. By 1966, the council already had nine committees dedicated to area studies, not counting the fellowship selection committees for the training of area specialists.
But weren’t these area studies just the sort of interdisciplinarianism I was looking for? Perhaps I judged them too harshly and from insufficient information, but I thought not. Too often, they seemed to aim at training disciplinary specialization within area specialization: experts on the Russian economy, the Chinese government, the Indonesian family. And because of the necessary emphasis on language skills, combined with a frequent attraction to current events, they seemed often to degenerate into high-grade journalism.
Arguments addressed to the Ford Foundation that this was not the way to go had no effect. Foot dragging had little more. (“Whose food I eat, his songs I sing.”) At least, I thought, we should be making strong efforts to convert area studies into genuine social science. To do that, we should establish committees for comparative research on important topics, drawing together the appropriate area specialists. That would impel them to conceptualize what they were doing at a higher theoretical level.
We never found the funding to do much of this, and I was never certain how much understanding and support I had from the council staff. When I made my periodic speech on this topic, heads nodded in agreement, but no visible action followed. I am sure they were more realistic than I about how movable our funders were.
As I look at the scene today, however, I see more comparative analysis among cultures, much of it sponsored by SSRC, and I am correspondingly more positive about the long-run effects of the area studies programs. I don’t think I can claim any personal credit for these newer developments.
My service on the Board of the SSRC had little apparent effect on the direction of development of the social sciences, and probably was not highly productive. But it was educational to me, and extremely pleasant. In addition to its bimonthly meetings in New York, each autumn the council held a larger gathering at Skytop, a resort in the Poconos, where I had the opportunity to become good friends and engage in stimulating conversation with a great many leading social scientists.
In my activities with the Ford Foundation and with SSRC, I had my first experiences of seeking to influence large affairs, where it is never clear whether one’s efforts have any result but where a result even of size epsilon could be important. (Epsilon times infinity can be a large number.)
Notes
*1. The economist Charles Holt came to GSIA from the University of Chicago and Franco Modigliani from the University of Illinois. John Muth was one of our own graduate students.
*2. SRL was not all futility, however. If it made little substantive contribution to basic science, its technology was subsequently more fully automated with the help of computers, and its training program was spun off to the Systems Development Corporation, which for many years was responsible for training military personnel to operate the Dew Line air defense system and later systems.