One sees more references to complex systems now than general system(s) theory. Institutes of Complex Systems provide a home base for scientists and scholars who continue cross-disciplinary studies in the spirit of Bertalanffy and Rapoport.
The shift in popularity from general systems to complex systems is visible in Google's ngram viewer. The ngram viewer is a free service that displays frequency of word usage in English language literature scanned by Google.
References to "complex systems" have far surpassed references to "general systems."
References to general systems peaked in 1974 then declined. The same is true of references to "Bertalanffy" and "Anatol Rapoport" and "general system theory." All peaked around the same time, from 1969 (Rapoport) to 1974 (general system theory).
References to "complex systems" tracked upward together with "general systems" until 1974 when the two began to diverge. General Systems fell in usage but references to complex systems continued to rise into the 21st Century.
Organizations and conferences devoted to General Systems still exist, but they do not receive much attention from scientists. By contrast, over 100 institutes devoted to Complex Systems have sprung up around the world, many within the past ten years.
Here is a sampling, along with introductions from their websites:
That is only a sampling of more than 100 institutes now devoted to the study of complex systems. Non-English-language sites include the Institute for Complex Systems (CNR-ISC) in Rome, Italy, which was the source of the research about flocks of birds turning in a way that resembled liquid helium.
One answer to the question, "What is the future of general systems?" appears to be, "Its spirit continues in institutes for the study of complex systems." Almost all are affiliated with major universities.
None of these institutes explicitly cites Bertalanffy or mentions General Systems. All of them are devoted to cross-disciplinary studies in search of principles, findings, and mathematical models useful for analyzing a range of systems. That is what Bertalanffy, Rapoport, and others set out to do.
Ironically, complexity (as such) was never mentioned by the founders of General Systems. Locating interdisciplinary and cross-situational patterns, principles, laws, and formulas was at the heart of the matter for them, whether systems were simple or complex.
Complexity was also not a key focus of this toolkit. Most of the principles are simple, focusing on patterns students might understand immediately with a few examples.
By contrast, the typical Institute for Complexity Studies is intended for PhD students and postdocs and career scientists. The publications and research projects emanating from these institutes justify "complexity" as part of the name.
I tend to think, however, this complexity is a side-effect of being on the cutting edge of science. The simplest systems yielded to scientific understanding long ago.
For intermediate students interested in complex systems, one of the best places to go is the complexity explorer web site sponsored by the Santa Fe Institute. It is devoted to listing "online courses and other educational materials related to complex systems science" including the excellent, free MOOCs (online courses) offered by Santa Fe Institute staff.
When I found out the Santa Fe Institute was offering a MOOC on Complex Systems in Spring, 2013, I signed up for it. This was my first experience as a student after teaching for 30 years, and it was fun.
The course was taught by Melanie Mitchell, whose 2010 book Complexity: A Guided Tour was much like the course. It is a good book to read, for students interested in the next step up in complexity from this toolkit.
Melanie discussed the issue of how to define complexity in her first unit. She included video interviews with several other luminaries from the Santa Fe Institute, regarding various definitions of complexity.
This actually revealed that the issue is not clearly settled. That was Dr. Mitchell's opinion as well.
Several operational definitions of complexity are available. Most relate complexity to the length or difficulty of descriptions of a system, which might include the number of interacting agents in a system, the non-linearity of its behavior, or the non-uniformity of its structure, its unpredictability, or the amount of computing power require to characterize the system.
Meanwhile, subjects covered under the label of Complex Systems are often about simplification. They show how apparently complex systems can be generated by iterative processes. For example:
See a pattern there? It almost invites a general system principle: "Wherever you see a Complex Systems demonstration, you will find some complicated thing arising from iterations and combinations of much simpler processes."
The MOOC on Complex Systems used the free simulation tool NetLogo. It is truly a great instructional tool, because students can run simulations in their own simplified environment, on their own computers, and see for themselves what happens as various systems, grow, interact, or evolve.
In NetLogo, each advancing moment of time is represented by one cycle of the program. Things onscreen move (spread, grow, evolve) as the program steps through time, and of course, the process can be controlled, tweaked, slowed or sped up, run in reverse, as parameters are altered, just to see what will happen.
NetLogo is a small taste of what simulations in general do for systems analysis. They are indispensable for high level scientists. For beginners using a simple program like NetLogo, they make abstract principles of system dynamics tangible: little worlds unfolding on a computer screen.
Most instructive of all, you can play with the micro-world in front of you and see what happens. Again and again, small changes in the initial conditions produce large changes in the growth or change within a system.
The label complex systems is now locked into the titles of institutions all over the world and will not change. But the whole enterprise could just as well be labeled, "Finding simplicity in apparent complexity."
Studies of complex systems are not about celebrating or multiplying complexities. They are about detecting simplifying principles that underlie apparent complexity.
In short, complexity is not the focus, goal, or desirable outcome of complexity studies. The goal is understanding and analyzing complex-looking systems, and that means striving for clear and simple explanations wherever possible.
The simpler and more efficient a rule or algorithm that solves a complex problem, the more elegant it seems. A great example comes from Anatol Rapoport.
In 1979 Rapoport entered a game theory competition sponsored by Robert Axelrod, based on the Prisoner's Dilemma game. This is a game invented at Princeton's Institute for Advanced Studies in the 1950s in which trust and betrayal are the two main options.
The tournament played computer against computer, and Rapoport's entry won first prize, yet it only required 4 lines of code (5 in Fortran). The strategy of Rapoport's program was tit-for-tat, and there were two rules:
Cooperate on the first round.
On round t>1, do what your opponent did on round t-1.
The importance of this finding is hard to overstate. Prisoner's Dilemma captures the essence of conflicts between hostile parties. If tit-for-tat is the best overall solution, that has implications for how to do effective diplomacy.
The winning solution, the best way to negotiate in a stalemate between trust and betrayal, is 4 or 5 lines of code. That is a fine example of how systems analysis, at its best, looks for simple solutions to complex problems.
If you were at an Institute for Complex Systems, would a solution like tit-for-tat put you out of business? Not at all. Once you have a good tool, you move on to the next level of complexity, studying when and how best to use it.
If you were a PhD student with a special interest in conflict resolution, you might study how the tit-for-tat strategy has been used or misused in the past. When has worked or not worked, and why?
How might it be applied in the Middle East, or on the Korean peninsula, or in other political hot spots? A simple tool can be employed to ask and answer complex questions.
Much of the work going under the name of Complex Systems has this characteristic, it seems to me. Principles such as synchronization, scaling, and the variation-with-selective-retention pattern in creativity can greatly simplify the analysis of complex systems.
A tool proves itself by its utility. To judge this, you must become handy with the tool. You must use it, put it to work, and see what benefits it provides. That was a lesson I learned again from the "law of requisite variety" discussed on the previous page.
As always with tools, different people will get different results. Some will feel inept and put the tools aside, others will move on to more powerful tools.
If you find a tool at the right level of complexity, one that explains previously unexplained phenomena of interest to you, then you become comfortable with it. Like any tool used repeatedly, it soon becomes familiar, easy to use, and applied almost automatically to new situations.
Therefore, for those who already find the General Systems principles in this toolkit easy and obvious, good! That indicates readiness for greater challenges.
Meanwhile, the simple principles of system analysis do not become obsolete, just because one moves beyond them or starts to find them obvious. They become implicit, called upon easily and automatically as needed to apply analytic abilities to new tasks.
Write to Dr. Dewey at firstname.lastname@example.org.
Copyright © 2017 Russ Dewey