I
often find myself examining details of the early history of the Santa Fe Institute to help explain the choice of the word "complexity" to describe our research agenda. Occasional critical comments have, in effect, stated that the word has many definitions and remains ambiguous. I have no problem with these comments, but they do amuse me, and perhaps some personal experiences will explain why.
Early in the discussions that led to the organization of the Santa Fe Institute, my principal interest was in drafting a broad charter that would encourage physical scientists to interact with life scientists, neuroscientists, cognitive scientists, social scientists, and even legislators and government executives. Through my research experience at Los Alamos, I had become deeply impressed with the interactions between biologists, chemists, and physicists which led to molecular biology. The first talk I heard that hinted at the nature of the genetic code was given by George Gamow on a visit to Los Alamos in the early 1950s. In later years, I had many opportunities to encourage similar interdisciplinary research. I was usually pleased with the result.
My interest in broadening the concerns of physical scientists was further stimulated when I joined the White House Science Council (WHSC) and, like so many scientists before me, observed the dichotomy between the expertise of physical scientists and the daily preoccupations of government with social issues, human behavior, and constituencies. Although scientists were constantly being told to avoid these worldly concerns, our advice was somehow expected to take them into account or risk being politely dismissed as irrelevant. I began to voice my puzzlement: Must we always limit ourselves to the world of elegant analysis, or could we also consider problems for which there was no single answer?
The suggestion that there might be a place where such questions could be better addressed was encouraged by fellow members of the WHSC. In discussions at Los Alamos with other Senior Fellows who met regularly, enthusiasm for a broadened approach ran high but each person's approach differed considerably in detail. There was an initial discussion of expanding our interests in the life sciences. Then the conversations turned to cognitive science. However, a consensus never emerged. I was frustrated. Finally, when I asked Herb Anderson how he would resolve the situation, he suggested a founding symposium in which a number of eminent scientists might describe their views and interests.
The title chosen for the first two founding symposia in the fall of 1984, "Emerging Syntheses in Science," was suggested by their chairman, Murray Gell-Mann. Initially, complexity was not a major theme. A few of the papers presented mentioned complexity in connection with living systems. Gell-Mann included complexity in biology as a possible topic of interest. Steve Wolfram discussed his interest in complex systems theory, emphasizing the very complex behavior that could be calculated for simple systems governed by simple local rules. The topic of complexity began to resonate.
I sensed the possibility of a consensus and was elated. But a consensus on what? In conversations with various individuals, it became increasingly clear to me that each had his own interpretation of the meaning of "complexity." What these interpretations seemed to have in common was the sense that complexity advocated crossing the line that divided elegant analysis from statistics. Of course, the increasing acceptance of the work of Gibbs within the past two generations had established a precedent. But his results were for systems with identical particles, or agents. What if they were different? The question pervaded all aspects of the life and social sciences without losing its connection to the physical sciences.
Scientists in disparate fields sensed a broadening of their charters. To those who had been invited to the symposia, this seemed welcome and overdue. The ambiguity of "complexity" was a virtue, because it could be defined within various fields to fit different research interests. By the time I gave the concluding talk, I was a convert. I rather arbitrarily indicated that our research focus would be on complex and interactive systems.
In early 1986, when we were hosted by David Robinson at the Carnegie Corporation in New York to present our plans to possible corporate funders, our research theme was complex adaptive systems (CAS). It was a term already familiar to Jack Complexity? Who Ordered That?