Complex Adaptive Systems
To understand the body of knowledge that is today known as Complex Adaptive Systems (CAS), it is important in my opinion to understand the streams of thought that have woven themselves together to create its philosophical and practical foundations. In the early decades of the twentieth century, nascient systems thinkers began observing the characteristics of natural systems and classifying the properties of these systems. At the root of systems thinking has long been the idea that understanding the workings of a system in one context (say evolutionary biology) can lead to the transfer of that understanding to building new artifical systems in new domains.
Two schools of systems thought developed in the 1940's. Nortbert Wiener was instrumental in developing the field of Cybernetics. Wiener's key insight was identifying the crucial role played by feedback in natural systems. This enabled Wiener to make strides in the design of control systems for Anti-Aircraft guns to help the war effort. The concept of feedback control systems has since become pivotal in the development of industrial manufacturing systems and process automation. The ability for components of a system to exert influence over one another through feedback mechansms is a key underpinning of current CAS thinking.
Click here for more on Norbert Wiener
The other seminal school of systems thought was the attempt at a General Systems Theory by Austrian biologist Ludwig Von Bertalanffy. Bertalanffy's vision was to develop a set of interdisciplinary ideas that were applicable to complex systems wherever they might appear, from biology to financial markets to complex government bodies, to any thing that contained certain organizational concepts and relationships between the component parts. It is hard to overestimate the streams of thought set in motion by the early work of Bertalanffy and his collaborators such as Anatol Rapoport, Kenneth Boulding and Ross Ashby.
Click here for more on Ludwig Von Bertalanffy
Two contemporaries of both Wiener and Bertalanffy who had considerable input into the thought stream that eventually led to complex adaptive systems and artificial life are Alan Turing and John von Neumann. Neither are thought of as systems scientists, though each had important things to say on the possibility of artificial life and modeling complex systems by computer.
Click here for more on Alan Turing and John von
Neumann
The 1960's and 70's saw an attempt by systems thinkers such as Heinz Von Foerster, Gregory Bateson, Margret Mead and Humberto Maturana to take the best of the work from Wiener and Bertalanffy and produce a more articulated and comprehensive systems theory. This in my opinion set in motion the development of ideas that led to catastrophe theory, chaos theory and the working out of ideas on complexity, self organization and connectionism.
Click here for more on Gregory Bateson and Margret
Mead
Click here for more on Humberto Maturana
While the development of systems thinking and theory was taking place in the 60's and 70's, computer scientists such as John Holland and Lawrence Fogel began incorporating ideas from evolution, genetics and natural selection into computer programs. These computer programs were initially aimed at problem solving and optimization, however, it soon became clear that the concepts of evolutionary computation had much wider application. Chris Langton, a maverick computer scientist, along with collegues such as Norm Packard and Doyne Farmer, took the ideas of evolutionary computation to their next logical step. Their insight was that perhaps the combination of the ideas fermenting in evolutionary computation, coupled with the frameworks being developed under the banners of complexity science and chaos, could lead to the development of artifical life. After an intial conference organized by Langton in 1988 under the auspices of the Sante Fe Institute sparked huge interest in the academic community about the feasibility of encoding the fundamentals of life into computer systems, many new approaches were developed. The need for an all embracing theory on which the artifical life vision could be constructed led to the eventual development by Holland and others of the notion of a Complex Adaptive System.
Click here for more on Langton, Packard and Farmer
Holland articulated his vision of the nature and characteristics of a Complex Adaptive System in his seminal book Hidden Order. His thoughts are by no means the only views on what constitutes a CAS, however, in many respects they are the clearest exposition to date and the ideas are straight forward enough that they can actually be developed into computer software and explored insilico. I know this as I took this step as part of my Masters dissertation. I will describle the key components of the Holland model next and then proceed to look at the criticisms of his ideas and other frameworks for CAS that have been proposed.