The failed institutionalization of “complexity science”: A focus on the Santa Fe Institute’s legitimization strategy

Founders’ social networks

Why did SFI’s founders decide to begin such a project at the end of their careers? Prior to answering this question we have to distinguish between at least two types of SFI members. A first category is those whose social, symbolic, and economic capital is sufficient that their implication in the field of complexity is not done in view of self-affirmation in academia. This is precisely the category of SFI’s founders who justified their endeavor because of their multiple interests, ³⁹ the need for a new scientific paradigm in science, and the will to contribute to solving major issues affecting society. A second category of actors includes scientists who were situated in a position of weakness, due to their youth and/or to their scientific marginality. This is the category of the first SFI members who – according to their own account – lamented isolation because of their interdisciplinary approach to science. These researchers sought in SFI a home, in addition to a means of gaining scientific credibility.

Among the founders, one of them – physicist Edward Knapp – enumerates Herbert Anderson, Peter Carruthers, George Cowan, Stirling Colgate, Darragh Nagle, Nicholas Metropolis, Louis Rosen, Richard Slansky, Alwyn C. Scott, David Campbell, and himself from the Los Alamos National Laboratory; Gian-Carlo Rota from MIT, David Pines from the University of Illinois, Anthony Turkevich from the University of Chicago, Murray Gell-Mann from Caltech, and John Rubel from the Department of Defense. ⁴⁰

The group of “Senior Fellows” from LANL and their friends from other institutions gathered regularly between 1982 and 1984, before sealing the birth of their institute. During such meetings, participants debated about a series of fundamental aspects for the creation of the center, as Knapp explains: “what would be the best scientific focus for the institute, what should be the scale, what was the right mix of students and faculty, how best to broaden the initial group of founders, and the like.” ⁴¹ As Gell-Mann explained in 1994, “[i]n the beginning, we couldn’t see clearly what sorts of emerging scientific syntheses we should seek.” ⁴² Cowan related that everyone had their own idea of what they wanted. ⁴³ For instance, Metropolis initially proposed cognitive science as the central theme of the SFI. ⁴⁴ Only after several discussions were complex systems established as the general object upon which to apply interdisciplinarity. ⁴⁵ Also, initially the founder group wished to install natural sciences laboratories in the future campus, ⁴⁶ but the neighborhood ⁴⁷ and the administration of Santa Fe stopped them from doing it. ⁴⁸

In May 1984 the founding group officially established the legal entity. Since the name “Santa Fe Institute” was already used by a therapeutic service in the area, they temporarily called it the “Rio Grande Institute.” ⁴⁹ In October of the same year the center received the “nonprofit” label from the Internal Revenue Service. In 1985, as soon as the therapeutic center closed, the founders modified the institute bylaws to give it the name it has kept until today. ⁵⁰ At that time the SFI possessed neither staff nor a physical headquarters. As Knapp later related: “We joke that our early home was a desk drawer in Trustee Art Spiegel’s office in Albuquerque, and we progressed to a Santa Fe post office box before we had a place of substance.” ⁵¹ In 1985, SFI rented an office in a building in Santa Fe for its two employees, the polyvalent Ginger Richardson and the development director Ron Zee. ⁵² In 1987 the institute was installed at 1120 Canyon Road, in the Cristo Rey Convent. ⁵³ This move marked the period in which the institute became widely known, first in the United States, then in Europe and the rest of the world. In July 1991 it was to be temporarily situated at 1660 Old Pecos Trail while waiting for the fundraising for, and the acquisition and renovation of, its final home. ⁵⁴ Partially thanks to a donation by Cowan to purchase a mansion in the Santa Fe hills, the institute found its final headquarters at 1399 Hyde Park Road in July 1994.

Art Spiegel was an entrepreneur who had been introduced to the group by Cowan. ⁵⁵ As the bulletins reiterated several times, “Cowan is often credited by his colleagues as the stimulus behind the Santa Fe Institute.” ⁵⁶ In a parallel way, Spiegel is described by Knapp as “very important in getting the Institute off the ground.” ⁵⁷ The archives are full of letters that he had sent to notable contacts in Santa Fe – entrepreneurs and managers in all domains, as well as scientists and local art dealers – to invite them to the first fundraisings. ⁵⁸ The SFI budget for the first year was about 100,000 dollars, mainly due to Spiegel’s activity and to senior fellows’ donations. ⁵⁹ The importance of the founders’ social and symbolic capital in launching the institute is exemplified in the friendship that united Cowan to Spiegel. ⁶⁰

The two had met in 1958 at a private dinner at the house of some common friends in Los Alamos. Their first collaborations focused on the animation of cultural activities in order to “compensa[te] for Los Alamos minimalism.” ⁶¹ The neighborhoods of what had been the headquarters of the Manhattan Project during World War II were indeed a relatively desertic area, which had been developed since then by the scientists. To get SFI started, LANL fellows activated their personal networks at several levels.

In 1986, SFI’s members increased from eight to seventeen. That same year, Ginger Richardson was hired by Cowan as polyvalent secretary in order to organize the first scientific events and write the first research projects. ⁶² She later became responsible for public relations and was made supervisor of the institute education sector. ⁶³ In the meantime, Cowan was assisted by three vice-presidents, by the president and vice-president of the Board of Trustees, as well as by the secretary. ⁶⁴ In 1987, the Board of Trustees counted twenty-seven members, among them scientists, entrepreneurs, and ex-politicians. The Scientific Advisory Board had forty members, some of which overlapped with the Board of Trustees. ⁶⁵ That year, the institute started the “Visiting Fellows” program, with six members (among whom were some of the founders). The program would last one to twelve months and provide fellows with an accommodation allowance, an office, research fund support, and a salary. The number of associates increased gradually to reach a hundred members in 1993 – this number has stayed constant until today. ⁶⁶ The visiting fellows’ only obligation has always been to spend at least one week per year at the institute – mostly coinciding with an integrative workshop. Some people stay longer, even a year for those on a sabbatical break. Among the represented nationalities, British, French, German, and Indian were the first. Rapidly new entries came in from Europe, Latin America, and Israel. In 2007 the “External Faculty” represented eighteen different countries. ⁶⁷ Despite the fact that SFI’s “community was highly influential in East Asia” and in particular in China, ⁶⁸ no member ever came from that area.

To do an assessment of the first years of the institute, Cowan wore the hat of a sociologist:

The issue of complexity has been popping up everywhere. The Santa Fe Institute has given it a little more credibility, I think, because the names attached to it are prestigious. We have a roster of National Academy types and Nobel winners, which suddenly did something very important for the whole notion, that is, to make it look more respectable. ⁶⁹

SFI’s founders were conscious of the strategic usage of their social, symbolic, and economic capitals: “We were picking the people, not the topics.” ⁷⁰ During one of the founding meetings, Metropolis subordinated the success of their scientific enterprise to the presence of some very famous people. ⁷¹ Symbolic capital often goes hand in hand with social capital: “[Gell-Mann, Pines, etc.] knew everybody. They could just pick up the phone.” ⁷² Moreover, some of these scientists were also entrepreneurs and have often directed substantial sums to the SFI to secure its functioning – Cowan for instance had founded a private bank in Los Alamos and had become a millionaire.

SFI and media reflectors were pointing to the “stars.” But what about the other members’ prosopography? As soon as SFI opened up to postdocs, the researchers coming in were difficult to categorize in some established field. For the first ten or fifteen years, most of them found in complexity science a living, strategic, temporary, and sometimes perennial shelter. SFI’s unconventional environment was constantly praised by its associates, who declared that in comparison they had always felt restricted and inhibited in traditional academic spaces. ⁷³ Such a self-representation and narrative of the institute drove a fascination in students and the general public.

In the institute’s narrative, its objective was to create the conditions to explore new and risky scientific territories. ⁷⁴ In a retrospective interview published in the 2014 bulletin, Richardson considered that the mission had been achieved: “[the] success . . . [of] many . . . SFI postdocs, belies the founders’ early worries that a segue into complexity science could be a career killer.” ⁷⁵ In fact, the problem of recruitment of complex systems researchers still is a preoccupation of the SFI’s, and of complexity specialists in general. In his doctoral thesis, Williams mentioned some examples of people who had had “difficulties of reintegration” into academia after an SFI internship for the very reason that they belonged to different domains, and to no one in particular. ⁷⁶

Despite its modest size, the SFI is today widely known in the academic environments of several countries in the world – yet it is not always praised. Estimators show sympathy for the institute and claim the imitation of some of SFI’s characteristics in their practices and institutions. Selective estimators distinguish between specific people and periods of the institute’s story, and thus discard and distance themselves from some of them. Finally detractors either criticize the excessive hype around complexity and minimize SFI’s successes, ⁷⁷ blame its dominant masculinist and crypto-Catholic culture, ⁷⁸ denounce SFI’s neoliberal take, ⁷⁹ or complain about its reductionist approach. ⁸⁰

During the first years of the institute, the Senior Fellows’ symbolic capital attracted some quite well-known researchers to ally with them in their enterprise (e.g. computer scientist and psychologist John Holland from the University of Michigan, chaos physicists Doyne Farmer and Norman Packard from LANL). Furthermore, the founders’ prestige has increased the visibility and credibility of young and marginal researchers throughout SFI’s history (e.g. computer scientist Christopher Langton, the controversial founder of Artificial Life, and Welsh economist Brian Arthur, known for “complexity economics”).

Later on though, the SFI needed to find a new source of symbolic capital, especially after journalist John Horgan had criticized, on the pages of the Scientific American, complexity science for being “flaky” and the SFI for being “fact-free.” ⁸¹ In the mid-1990s it was clear to the Board of Trustees that the Senior Fellows’ prestige was no longer enough. SFI hence diversified its strategy and tried to attract rising “stars” in order to boost both the institute and the stars’ visibility. An ongoing case of this win-win association is that of computational complexity specialist Christopher Moore. Another example is represented by Greek-Brazilian statistical physicist Constantino Tsallis, who had made acknowledged contributions in the field of thermodynamics. Another example is the postdoc stay of the network theorists Duncan Watts and Mark Newman toward the end of the 1990s. ⁸² These researchers managed to collect so much scientific capital that American universities today compete to have them by offering generous salaries and important responsibilities. This is because they utilized consolidated techniques from statistical physics; they applied these techniques to new, exotic objects (social and genetic networks, epidemics, etc.); they publicized their work through general audience channels; they had efficacious institutional strategies; and they managed to perpetuate their specialty through academic teaching, conferences, and journals. But this is not the case for all complexity specialists. “[S]everal prominent SFI researchers,” as Williams wrote in his thesis, “essentially vanished from view when their time there was up” ⁸³ – whatever the country, if they want to persist and exist independently of complexity institutes and funds, complexity scientists have to pass traditional disciplinary tests.

The role of media and popular books

Most popular SFI books cannot be considered as secondary literature, but rather as what French historian Wolf Feuerhahn has called indigenous histories – that is, “histories of a discipline produced by the actors of the discipline themselves.” ⁸⁴ But what to say about enthusiastic accounts of SFI and complexity science given by certain journalists? Even if, as far as I know, they did not get any benefit from the institute, their stories contributed to entertaining a certain narrative about it. In their books the SFI appears as “the” place for interdisciplinarity and complexity science in the world.

The first book to offer a portrait of some personalities linked to the New Mexican institute (namely physicists Normand Packard, Doyne Farmer, and James Crutchfield) is Chaos, written by the New York Times journalist James Gleick. ⁸⁵ Published in 1988, this book has become one of the best-selling science books ever.

In 1992, Waldrop published Complexity: The Emerging Science at the Edge of Order and Chaos. The year before he had signed an editorial contract with Simon & Schuster after having reported for Science about an Artificial Life conference. According to his testimony, he spent two weeks at the institute in order to have a look at some archives and especially to conduct interviews with some of its members. ⁸⁶ His book rapidly sold several thousand copies and today tops 100,000.

Also in 1992, British science writer Roger Lewin published the almost identically titled Complexity: Life at the Edge of Chaos. ⁸⁷ Like Waldrop, Lewin reported SFI’s self-promotional discourse, anchoring his tales to some key personalities.

Williams gives an interesting analysis of these books to explain their success. One of the narrative strategies is, according to him, personalization. Protagonists are often presented as misunderstood geniuses, who are as marginal as they are inventive, as eccentric as they are fascinating. They are depicted as rebels, visionaries, explorers, and dynamic people that do not give up before the obstacles of reactionary academia. Protagonists are often presented with their nicknames in order to create identification and proximity. In some cases the fascination for complexity is complemented by a bent for mysticism – sometimes Buddhist, sometimes Catholic.

If journalists have a predilection for hagiographic accounts, in their own texts SFI members focus more on theories and models. ⁸⁸ Between 1994 and 1999 the books about the institute rarefied, probably as an effect of the critiques they had received. At the end of the 1990s complexity became popular again thanks to a new wave of celebrative publications. ⁸⁹

Conscious of the importance of such publications, SFI’s Direction sent Waldrop’s book to donors in order to promote the institute. In his review of Complexity published in the 1992 winter bulletin, SFI biologist Harold Morowitz argued something that confirms Williams’ thesis: “in these days a journalist is also an actor, for support of research is geared to fame, and fame comes in part from inclusion in books of this type.” ⁹⁰ For example, despite his interlocutors always using the plural, Waldrop’s narrative of a singular “complexity science” has had performative effects by suggesting the feeling of a cohesive and coherent community – just as Gleick did with chaos theory. Several representatives of complexity science in the United States and in Europe have testified that these texts played a role in their adhesion to the field.

Besides books, SFI has always given attention to media in general. ⁹¹ Newspapers, TV, radio interviews, YouTube, and general audience conferences are all venues considered to be of critical importance. Media produce cascade effects on obtaining funding, on member enrollment, and on the credibility of the institute. The frequent references to media coverage that the institute has made in its communication material are proof of this. ⁹²

The role of federal agencies, foundations, and big enterprises

Because of their social status and their rich biographies, the Senior Fellows had, as a long-standing member of the library staff said to me, an agenda with “lots of connections,” ⁹³ from federal organizations like the National Science Foundation (NSF) and the Department of Energy (DoE) (on which the LANL depends), in the world of big money, banks, and enterprises, as well as in that of politics and think tanks (Gell-Mann, for example, had been an advisor to the RAND Corporation for years). At this level of prestige and political proximity, the distinction between professional collaborations, intellectual esteem, personal friendship, and lobbying is hard to make.

The first funding from the NSF came in 1988 as a result of a meeting organized by Cowan between the SFI’s founders and Erich Bloch, the director of the agency. For the institute, the advantage of NSF support was few bureaucratic constraints compared to other federal endowments. ⁹⁴ Despite the absence of a dedicated section for interdisciplinary projects, Bloch found a solution to fund the SFI. As Cowan explained in his memoirs, this consisted in “taxing each of the divisions in order to provide $250,000 a year in support of the Institute.” ⁹⁵ Two more people were crucial in securing NSF funding over the long term – physicists Edward Knapp and Robert Eisenstein. The first had been the president of the NSF from 1982 to 1984 and the second of the NSF Physics Division from 1992 to 1997. Knapp indeed testified that: “Perhaps my connections with the Federal government stemming from my days at the NSF and at [Universities Research Association] helped us.” ⁹⁶ In 2004, Eisenstein was appointed as the new SFI president: “[he] brings those qualities to the presidency of an institution he’s known for some 10 years. Early on during his tenure at the NSF, he was responsible for directing many of the funds that came to SFI from there.” ⁹⁷

Later on, some of the federal agencies’ funds came in to support computer scientist Christopher Langton’s agent-based simulation called Swarm (an open-source simulation package used to model the interaction of a high number of social or biological agents and their emergent collective behavior). In April 1994, the SFI received a scholarship of US$323,000 from the Defense Advanced Research Projects Agency (DARPA) – the military and the intelligence were interested in an application of this platform in order to solve problems “related to the way troops adapt to changing circumstances on the battlefield. The army wants to know how much local decision-making power troops on the ground should be given and how much of that decision-making power should remain in the hands of commanding officers.” ⁹⁸ According to the first librarian of the institute, Margaret Alexander, Ellen Goldberg “decided not to take DARPA money anymore, because it was kinda dark and scary.” ⁹⁹ In fact, as a mathematician from the institute told me, some of the resident scientists keep responding to DARPA calls on a personal basis, but not without internal resistance, namely from postdocs. ¹⁰⁰

Since the beginning, the funding coming from foundations was also very important. In 1987, substantial contributions came from Russell Sage, Alfred Sloan, Richard Lounsbery, and the Heinz Company. ¹⁰¹ In 1988, John and Catherine MacArthur were added to the list. ¹⁰² If some of them became regular donors to the SFI, some others have come in and out. Overall, the place of foundation funding gradually decreased to 30 percent in 1996. ¹⁰³ By 2006 it had diminished to only 6 percent of the total budget. ¹⁰⁴ Yet during the 2010s, individual, family, and philanthropic foundations again became very important and constituted up to half of the SFI budget. ¹⁰⁵ This was mainly due to the decrease in individuals’ and trustees’ gifts, and to investment loss after the 2008 economic crisis; the director of communications of the SFI explained that “[a]s the economy gets tight, people . . . want more and more tangible results. And we are a theoretical institute and we are never gonna produce tangible results.” ¹⁰⁶ As the federal agencies fundraiser explained to me, government funds stayed the same before and after the Lehman Brothers’ crash thanks to the “American Recovery and Reinvestment Act” from 2009, ¹⁰⁷ but if private donors were the main contributors before the crash, foundations took their place afterwards. Some years following the crisis, the contribution of private and foundation donations equalized, while the investment income increased in size. Figure 1 illustrates the evolution of funds over three periods, including the first phase under Cowan’s direction (1991), the second one led by Goldberg (1997), and the post-2008 crisis (2013).

OPEN IN VIEWER

Figure 1.

The evolution of SFI funding in three different periods, from the 1991, 1997, and 2013 Annual Reports.

Less data are available about Senior Fellows’ and other SFI members’ relationships with private foundations. It is nevertheless arguable that the founders had close contacts with that world too. For example, Gell-Mann had strong connections with the MacArthur Foundation – one of the first and most assiduous of SFI’s financers. ¹⁰⁸ From 1979 to 2002 he had directed the organization, with the freedom to give funds to whomever he thought worthy. ¹⁰⁹

What Figure 1 shows but does not explain are companies’ gifts. Since the first important workshop organized by the institute to put physicists and economists together to work on financial economics, the SFI entertained excellent relations with big corporations like Citigroup. Its CEO John Reed was indeed one of the first and most important funders of SFI’s research projects with his “$250,000 to study the global economy.” ¹¹⁰ Cowan explained that, in 1987, Robert Adams – a member of the SFI board and head of the Smithsonian Museum in Washington – spoke with Reed about the institute, suggesting that he “come to Santa Fe with a few of his staff to talk about some of the problems in international finance that his company was attempting to solve.” ¹¹¹ Reed was keen for new mathematical models to avoid loss from economic crises and stock exchange crashes. ¹¹²

According to SFI’s philosophy, “[c]ompetitive peer-reviewed grant funding provides credibility for the Institute’s science, while the availability of unrestricted funds from contributions allows investment in cutting-edge, high-risk ideas.” ¹¹³ Another opening to the world of enterprises occurred in 1988 with the entry of some new members in the Board of Trustees from the oil and weapons sectors. ¹¹⁴ Here again, contacts preceded SFI’s birth. Trustees were co-opted among the founders’ friends, as was the case for Robert Anderson from the Hondo Oil Company who was brought in by Cowan. ¹¹⁵ In 1989, other funders were added to the list – among them Barclay’s and the Trinity Capital Corporation. ¹¹⁶ In 1990 the number of foundations was nine and that of companies twelve, with some new entries such as IBM, Morgan Stanley, and Volvo. ¹¹⁷ That same year the Board of Trustees got new members from the aeronautics, finance, and informatics sectors. ¹¹⁸ For several years most of the SFI trustees came from Silicon Valley and Wall Street. ¹¹⁹ Yet it was under Goldberg’s presidency (1996–2002) that private and business donations increased to two-thirds of the institute’s funding.

Today the institute is provided with a “strategic partnership” office, which operates in three sections. One is in charge of collecting public funds from federal agencies through the search for new scholarships and the writing of projects to respond to the calls. Another one is in charge of searching for and keeping in contact with individual contributors. The third fundraising tool is the “Business Network for Complex Systems Research” (or BusNet): even if the SFI Mission Statements of 1986 stated that one of the objectives of the institute should be to “serve the industry,” ¹²⁰ the BusNet was not created until 1992. In 2015 it was renamed the “Applied Complexity Network” (or ACtioN) by the current president, David Krakauer. ¹²¹ Its task is to provide consulting services to enterprises, non-governmental organizations, and administrations.

The role of scientific publications

In March 1987, the SFI started a collaboration with the Addison-Wesley Publishing Company, with the aim of releasing a series of books under the name “Santa Fe Institute Studies in the Sciences of Complexity.” ¹²² The series included conference and summer school proceedings, as well as individual and collective monographs.

In a 1989 bulletin, the person responsible for scientific editions, Ronda Butler-Villa, explained the importance of such publications in an article entitled “Getting the Word Out”: “the publications program provides the only tangible evidence of what is being accomplished [at the SFI].” ¹²³ Butler-Villa explained that 1,500 copies of volumes were generally printed, half of which were commonly sold in the first six months, while the other half were stocked in bookshops. No one of these volumes has ever overcome the symbolic bestseller threshold of 10,000 copies.

The editorial strategy was changed in 1992, with the perspective of reducing the printing costs and increasing the efficacy of such publications. ¹²⁴ In 1994, the number of titles was fifty, but the high cost of production gradually dissuaded both SFI and the editor from adding new ones. ¹²⁵ Even if, in 1998, the contract with Addison-Wesley was replaced by a new one with Oxford University Press, ¹²⁶ Richardson explained that “these volumes lost with time their value. . . . As time went on, there were many new ways to spread the word.” ¹²⁷

A section of the SFI bulletin has often been dedicated to scientific article announcements. In the 1991 fall–winter issue the institute mentioned the increasing number of specialized publications authored by its affiliated members: twenty-eight in 1990, thirty-one in 1991. ¹²⁸ In 2007, Geoffrey West highlighted the scientific excellence of SFI members by pointing to the prestige of the journals in which they publish: “Although SFI is dedicated to risk-taking and supporting more speculative investigation, our researchers continue to be published and recognized in leading scientific journals”; ¹²⁹ “[o]ur researchers routinely publish in top journals including Science, Nature and the Proceedings of the National Academy of Sciences. In fact, in 2007, we had more than 100 publications in such journals.” ¹³⁰ In their turn, Nature and Science roundly shared their enthusiastic support for SFI and contributed to its scientific legitimacy by regularly dedicating their journalistic and promotional sections to its activities: “The journal Nature touted SFI as a place where ‘through the interaction of talented people there may well arise new visions of how the world is put together,’ and Science described it as an intellectual playground where scientists knocked around new ideas ‘like volleyballs’.” ¹³¹

It is important to note that not all of these articles covered the field of complexity sciences – the reference is thus to any scientific article published by SFI members, so as to realize a form of capital transfer from more established fields to the emerging one.

The role of summer schools

As sociologists of sciences Kate Bulpin and Susan Molyneux-Hodgson have shown, educative devices are important in the creation of new technoscientific domains. ¹³² In the case of complexity sciences, a specific bachelor’s degree does not exist while master’s degrees and PhDs are very rare. ¹³³ The device that importantly contributed to getting complexity science into existence was the Complex Systems Summer School (CSSS).

This was conceived by Peter Carruthers and first launched in 1988. ¹³⁴ Mike Simmons, one of their kingpins, described CSSS as an occasion to reinforce the academic existence of the institute and to “maintain SFI’s position in the forefront of this exciting new field of complexity.” ¹³⁵ At the beginning the target was the United States. With time, the students diversified in terms of the nationalities represented. ¹³⁶

The first CSSS was organized with the support of several research centers, universities, and federal agencies. For the second summer school, some Silicon Valley companies provided computer materials for free. ¹³⁷ For several years the event took place at the local Saint John High School. ¹³⁸ There were five initial students who had been mainly brought in by physicist Philip Anderson. ¹³⁹ A workshop session in order to ensure that participants could all handle modeling software was also organized. The school remained without charge for unconfirmed scientists until 2008, after which people had to pay. Today students, post-docs, and faculty pay US$4,500, while other professionals pay US$6,000.

Since the beginning the institute attributed an important place to this educational device. First, to produce new complexity adepts in America and around the world, and second, to fix the international standards of complexity science: “the [SFI Lectures in Complex Systems] series’ cumulative effect is not only to record what has already been done in the field but also to contribute to defining complexity’s future evolution.” ¹⁴⁰

CSSS lectures were published as proceedings by Addison-Wesley. Overall, SFI published half a dozen volumes as a result, before it was considered not to be viable anymore. ¹⁴¹ CSSS and their publications were important for the establishment of some common ground for complexity science inquiry tools. ¹⁴² These varied through time and scientists were invited to innovate by creating or bringing in other means of inquiry. CSSS was also a place for socialization: “Well over a thousand students have participated since its debut in 1988, three points. ‘Each brings the SFI mode of thinking back to his or her home institution’.” ¹⁴³

The success of the summer school led the institute to organize a winter version in January 1992 in Tucson (Arizona), the first of a long series that, through the “International Outreach Program,” touched China, Eastern Europe, and South America. ¹⁴⁴ The program was reinforced in 2013 with the implementation of a series of Massive Open Online Courses. ¹⁴⁵

The role of pedagogical projects

Another pedagogical device that SFI put in place at the beginning of the 1990s had an even longer-term perspective – it concerned students from secondary level to postdoc.

The first program of this kind was organized in spring 1990 and involved undergraduates. ¹⁴⁶ Students could participate in one of the research programs held by the institute, or spend a study period under the guidance of an SFI member. ¹⁴⁷

In 1991 the institute opened up the possibility for PhD candidates to write their thesis under the supervision of one of its members. ¹⁴⁸ In 2002 the program was financed by the NSF and gave freedom to candidates to choose their topic. They had to spend two and a half months in residence. ¹⁴⁹

The postdoctoral was launched in 1989 and used to last one to three years. ¹⁵⁰ eBay’s founder, Pierre Omidyar, became one of the strongest supporters of SFI postdocs. Besides his gifts to the institute and his participation on the Board of Trustees, in 2008 Omidyar implemented a funding program for some postdoctoral scholarships for which candidates can freely choose their topic ¹⁵¹ and are supposed to become complexity science ambassadors. ¹⁵²

During 1992–3, the SFI launched its first education program in high schools. ¹⁵³ With the support of a grant from the Pinewood Foundation, around sixty students from all Santa Fe high schools participated in a cycle of conferences held by a few members of the research center. The one and a half hour meetings were organized once a month and included a question time to answer students’ queries. Another device started in 1998 and lasted several years. It implied the usage of agent-based models in order to teach students the basis of coding and a decentralized way of thinking. ¹⁵⁴

Some years later the SFI extended its pedagogical activities into secondary-level schools and deployed resources to train hundreds of teachers in using computational tools so that they could transmit SFI’s approach to 6–12-year-old children. The program was called the “New Mexico Adventures in Modeling.” ¹⁵⁵ In the light of the political interest in STEM, the SFI anticipated that complexity sciences would become pivotal in the twenty-first century. ¹⁵⁶

The role of integrative workshops

The integrative workshops represented the sole public activity of the institute until 1986 and are still organized today. Halfway between a conference and a summer school, this kind of event can last one or two weeks, and attract two to three dozen participants. Each attendant is a speaker and contributes by teaching – during a brainstorming phase – the other members how the chosen study object is approached in his/her discipline. In the phase of synthesis, attendees propose possible bridges between different disciplines. ¹⁵⁷

Since its first research projects in 1986, the SFI has claimed to work on “complex adaptive systems” or CAS, in reference to evolving aspects of living and social systems. The founding workshop of CAS was held in 1986 with the support of the Alfred Sloan Foundation. ¹⁵⁸ If until the mid-1990s integration was mainly sought for building a general theory of complexity, after Horgan’s disruptive article in Scientific American and under Goldberg’s presidency interdisciplinary efforts were redirected toward local general theories. ¹⁵⁹ Some workshops in particular contributed to providing visibility for the institute. This is the case regarding the 1987 meeting between physicists and economists entitled “Founding Workshop on Evolutionary Paths of the Global Economy.” ¹⁶⁰

More or less multidisciplinary, such workshops have covered a vast array of subjects and, as soon as the institute got up to full speed, more than twenty were organized each year. ¹⁶¹ They have been the place to weave collaborations, establish new fields of research and specialties, and contribute to creating the myth of the SFI as an open-minded and stimulating intellectual environment. ¹⁶²

The role of scientific journals

To found a new domain, it is necessary to launch a scientific journal. In 1990, Simmons announced the foundation of Complexity: An International Journal of Complex and Adaptive Systems. ¹⁶³ The first editorial director was biologist Harold Morowitz. Mathematician John Casti assisted him as executive editor. ¹⁶⁴ Simmons explained the reason for the journal as follows: “Even as complexity is emerging as a new field of research, it is in danger of fragmenting into a number of disjoint subfields that will not communicate adequately with each other because each will be served by its own specialized journals.” ¹⁶⁵ Complexity was conceived as a space of integration between different models in the perspective of founding a common theory of complex systems. First published in 1995 by John Wiley, it was conceived as a quarterly, international, and electronic journal. But it never benefitted from great visibility and reputation, especially in light of the abandonment of the project of an integrated theory of complex systems by the SFI itself after Horgan’s article. Several of my interviewees explained to me that they prefer to publish their best articles in traditional disciplinary journals, which also have higher impact factors.

The role of prizes and rewards

Yet another strategy for establishing a field is to create specific awards that compensate the people working in it, but complexity prizes are very rare in academia. ¹⁶⁶ Once more, the SFI has thus leaned on the symbolic capital of other disciplines’ prizes. That is why, in the course of the Bulletin issues, the institute celebrated all its members’ awards (career prizes or those for best articles, medals, nominations to the National Academy of Science, scholarships, simple marks of esteem from newspapers, etc.). ¹⁶⁷ The conditio sine qua non for scientists to appear in these listings was their belonging to the institute, not the fact of being awarded for research in complexity science. ¹⁶⁸

In Bourdieu’s terms, this operation can be seen as a form of capital conversion: scientists’ symbolic capital gained in more traditional frontiers of science has been used to legitimize SFI and its science through a sort of mirror effect.