**Principia Cybernetica Web (C)**

Author: F. Heylighen

Date: 20 March 1989

Parent Node(s): History of the Principia Cybernetica Project

Proposal for the Creation of a Network for Complexity Research


Definition of the domain

Motivation of the enterprise

Outline of a conceptual framework

Operationalizing the concepts

New languages and media

Some practical propositions


Last year [1988], four research groups, from the universities of Amsterdam, Brussels, Lissabon and Groningen, decided to form a consortium for proposing a large research project, on support systems for complex problem-solving, to the ESPRIT Basic Research Programme. However, the proposal was turned down. This was not unexpected, since there had only been two months between the decision to participate and the deadline for turning in the elaborated project. This period had been too short for a good coordination of the different approaches.

Yet each of the groups had quite some experience on their common research domain, which is situated somewhere near the meeting point between cybernetics and systems theory, self-organization theories, and the cognitive sciences. Each of them had already organized international meetings on the subject, of which the most recent are: "Self-Steering and Cognition in Complex Systems" (Brussels, 1987), a summer school on "Self-Organization of Cognitive Systems" (Groningen, 1988) and "Mutual Uses of Cybernetics and Science" (Amsterdam, 1989).

Being one of the participants in this development, I concluded that the research subject itself was very important, and held high promises for the future, but that there were basic obstacles to the practical organization of sufficiently large-scale projects, difficulties exemplified by the rejection of the ESPRIT-proposal.

The first obstacle is that the domain is not officially recognized, it does not fit in with any of the classifications of traditional and recent disciplines. Its proponents are scattered among the most diverse fields: physics, psychology, sociology, computer science, architecture... The result is that, on the one hand, it is difficult to be funded through the traditional channels, which do not recognize the validity of the objective - unless it is motivated in terms which are sufficiently clear and concrete to show the practical applications. On the other hand, the practitioners of the field itself have problems with cooperating, because of lacking communication channels, or even because they do not recognize each other due to a different terminology or methodology.

The second, related obstacle is that there is no generally accepted conceptual framework and methodology, no "definition" of the research domain which would give it an identity and thus would make it easily recognizable. This makes it difficult for groups using different terminologies, corresponding to different but related concepts, to coordinate their efforts by developing a collective project.

The present proposal is a first attempt to stimulate a movement which would remove these obstacles. First the proposal examines the research domain, proposing a definition, a motivation explaining why this research is very important and why it should be promoted at the present time, and a first outline of a conceptual framework together with a methodology. The term which was chosen to label the domain is complexity, denoting a concept which is essential as well in the cognitive as in the systems sciences, and which is used here in a qualitative and dynamical sense. Finally the text proposes some concrete measures for stimulating the communication and collaboration between individual researchers and teams engaged in this domain by the creation of a network.

Two major criticisms can be raised against the organization of such a network. The first one states that a network would just multiply the existing communications and thus would contribute to the existing "information overload" which is typical for the most active researchers. In other words instead of helping us to cope with complexity it would amplify the existing complexity.

I would answer to this that most of the stress accompanying the research as it is done now could be removed if more efficient systems for handling complexity and better resources (especially advanced information technology) were available, and if researchers were less uncertain about receiving the necessary contracts. The exchange of ideas and eventual collaboration with researchers working on the same domain will probably help the creation of such support systems and make it easier to get the required funding. Yet one could expect that in a first phase the creation of a network would indeed amplify the stress for certain persons. Therefore I propose that everybody involved should be free to (partially or completely) withdraw from the activities at any instant. The network should be self-organizing, i.e. be able to survive the disappearance of any of its members.

The second criticism states that there are already many associations or societies concerned with systems, cognition and/or self-organization, and that it would be more fruitful to put one's energy into the elaboration of these existing organizations, rather than in the creation of yet another one. In particular, it states that the subject of "complexity" is already fully comprehended by cybernetics and systems theory, and that there is no need to establish it as a separate domain with a separate organization.

My answer to that would be, first, that a network as envisioned here is not just another society, but a loosely coupled collection of informal communication channels, which functions not so much as an official representative of a group or domain but as a practical medium for facilitating cooperation and funding, thus bridging the gaps between the existing organizations. Second, I must of course acknowledge that the proposed research is a continuation of the systems tradition, but I think it is useful to make a distinction between what I would call "classical" cybernetics and systems theory and the present approach, which may be labelled "non-classical" or "second-order", and which studies the creation of systems rather than their static organization (see further). Moreover, the formulation of "complexity research" as a distinct domain may make it easier to insert related approaches, such as AI, non-equilibrium thermodynamics, cognitive psychology, ..., which are separate from the "systems" movement in the restricted sense.

Whether the proposed network is practically realizable or not, I hope that the present text will function as a basis for fruitful discussions, which may lead to a better formulation of an as yet ill-defined approach. I hence welcome all reactions to the proposal, either positive or negative.


The following researchers have already reacted to a previous version of the proposal and thus (directly or indirectly) contributed to the present version: Dr. R. Glanville (Portsmouth), Prof. G. de Zeeuw (Amsterdam), Dr. G.J. Dalenoort (Groningen) and Prof. J.P. Van Bendegem (Brussels).