Developing second-order cybernetics: a collection of papers from the UK Cybernetics Society


ISSN: 0368-492X

Article publication date: 1 October 2004


(2004), "Developing second-order cybernetics: a collection of papers from the UK Cybernetics Society", Kybernetes, Vol. 33 No. 9/10.



Emerald Group Publishing Limited

Copyright © 2004, Emerald Group Publishing Limited

Developing second-order cybernetics: a collection of papers from the UK Cybernetics Society

Developing second-order cybernetics: a collection of papers from the UK Cybernetics Society

Early in 2002 Ranulph Granville came up with a suggestion that the (UK) Cybernetics Society should hold a one day workshop on the theme of “Second Order Cybernetics (SOC): What is it? What is it good for?”

As time progressed it became evident that not all members of the society were au fait with the tenets and history of SOC which fact itself argued for the value and relevance of such a workshop. It also meant that conservative voices preferred a more traditional format of paper presentations rather than a workshop. In the end a compromise of sorts was reached where for its annual one-day conference (“Developing Second Order Cybernetics”, 14 September 2002) the Society invited papers with at least an indirect connection to SOC, while reserving the final one hour or so of the day for a debate around the theme of “What is the value of SOC?”

It is perhaps worth reminding the reader that the distinction between a first- and second-order cybernetics was first proposed by Heinz von Foerster, one of the founding fathers of cybernetics, in the early 1970s. According to von Foerster, first-order cybernetics is “the study of observed systems” and SOC is the “study of observing systems”. I will not elaborate further here. My paper (Scott, this issue) and Glanville's paper (this issue) present accounts of the origins and subsequent development of the first- and second-order distinction.

The conference brought forth a fascinating set of papers. It also provoked a very interesting debate. Roughly this latter could be characterised as having two sets of voices – one that, although prepared to discuss models of how observing systems work or to discuss the operational role of the observer in carrying out experiments and reporting that data to fellow observers, remained sceptical or not particularly interested in the grander claims that SOC represents a radical paradigmatic departure for all sciences in terms of its epistemological and ethical consequences. The other set of voices more closely followed von Foerster in emphasising the epistemological and ethical issues and showed more interest in the relevance of SOC concepts and tenets for the social sciences and applied disciplines such as management, psychotherapy, education and design. I will leave the reader the chance of examining the collection presented here in order to match particular contributions to the voice sets described above.

As a matter of detail I should also point out that one of the authors was not able to present at the conference in the event. However, his paper was submitted as relevant to the conference theme and, following the usual reviewing process, has been accepted for publication. When I say “the usual reviewing process” I should perhaps also point out that each contributor reviewed at least two of the other papers in the collection.

While there are clear differences in positioning with respect to fundamental assumptions about SOC and the cybernetics enterprise as a whole, it was still possible for reviewers to help ensure that the papers do indeed present their ideas and concerns clearly and with appropriate scholarly rigour. Again, I will leave it to the reader to ascertain how well these aims were achieved.

In the event, most authors took the opportunity to revisit their papers in detail or, in some cases, to rewrite de novo. Thus, the collection here is strictly not the proceedings of the 1-day conference. Rather, it is part of the fruit that that day has brought forth. Other fruit includes an ongoing extensive mailing list discussion and additional material on the society's web site (available at:

About the papers themselves I will dare to make one or two comments by way of orientation. Of course my interpretation is just that, my interpretation. As noted earlier, my own paper is a historical account. In it I try to capture the essence of the recursive turn whereby in von Foerster's phrase “it behoves the observer to enter the domain of his own descriptions”. Critically this move reveals to the scientist studying the evolution and workings of self-organising observing systems but he himself is just such a system. I go on to discuss the ethical and epistemological consequences of this view.

Glanville's paper addresses the theme “What is the purpose of SOC?” showing how SOC itself elucidates the relativistic and reflexive character of the concepts “goal” and “purpose”. This seems to be a cautionary tale, warning us that even our most cherished fundamental concepts carry with them the marks of our own distinction making.

Alex Andrew makes two contributions. One is a set of reminiscences of the early days of cybernetics and his own involvement with Warren McCulloch's group at MIT and research on the cat spinal cord and frog vision. The latter is of particular interest as later developments gave rise to the notion of the organizational closure of the nervous system, a key contributing concept in von Foerster's first- and second-order distinction.

Alex's other contribution is a critique of constructivism, the epistemological paradigm which identifies its roots in the ideas of von Foerster on SOC. Alex questions whether in fact constructivism has anything new to say to the already sceptical scientific practitioner.

Petros Gelepithis provides a critique of cybernetics and extant cognitive science with its AI based models of symbol processing, arguing that what has been lacking from both is a formal theory of communication. He goes on to outline the form of such a theory and goes on to discuss its relevance as a basis for improving the quality of communication and understanding in human society. The relevance to SOC is a shared concern with observers in communication. The more radical epistemological and ethical concerns of von Foerster are not addressed.

Gerard de Zeeuw, as part of a lifelong research programme addressing the evolution and form of scientific models and methodologies and using terminology distinct from that of von Foerster, develops a very sophisticated account of how projects aimed at dealing with social ills and the like may emerge as knowledge domains that are constituted by collectives. The design of collectives that serve as knowledge constituters is linked to SOC.

Tony Booth's paper addresses the vexing problem of exactly how, given our current level of knowledge of physical phenomena at the quantum level, does an observing system, human or artificial, heed some of those events and capture them as data. Referees agree that Booth's approach with his interdisciplinary synthesis of concepts is innovative and, although at an early stage of development, well deserving of publication as part of this collection. My own reading of Booth's paper is that the observing system is ipso facto a meaning generating system which, with respect to epistemology and ethics, brings him to a similar position to that of von Foerster.

Finally, Nick Green's paper is a tour de force in which he presents an overview of Gordon Pask's Conversation Theory (CT) and (as developed in his later years) Interaction of Actors Theory (IA). Green appears to me to do two things which are particularly innovative. He collects together what he sees as Pask's key concepts as a set of “axioms”. Essentially these “axioms” are the collection of conditions that together account for the workings of self-organising systems that have evolved to converse and interact as distinct, embodied actors. Although not formally stated as axioms as understood mathematics, Green's approach has validity as the axioms in question are to be understood as a concurrently applied set. In other words, they gain their a priori axiomatic status by being considered as a coherent collection of propositions that together provide access to the full richness of Pask's theorising. Green insists that Pask was indeed developing a full blown cosmology and that his ideas and models have relevance for all extant natural sciences. It is these aspects of Pask that Green wishes to bring to the fore. In the extended mailing list discussions that have taken place since the conference it has become clear that some of Green's interpretations of Pask's writings are controversial. In particular, in emphasising the cosmology, he has allided some of the careful distinctions that Pask makes (as some of us read him) that distinguishes his conversation theory and interaction of actors theory from the physical sciences, where the former as theories of conceptualisation and culture have the explicit second-order attributes of being relativistic (to an observer's perspective) and reflexive (explaining the observer to himself). Where Green appears to be exactly right is that in his later years Pask emphasised the kinetic status of CT and IA. Green sees this as not only a commentary about the form of Pask's later models of the dynamics of conceptualisation and interaction, to be distinguished from earlier kinematic or snapshot models of conceptual structures and domains of interaction, but also as a cris de couer which insists on our accepting, as embodied actors, that we are indeed living and breathing dynamical systems, conceptualising, conversing and interacting. Somewhere in his writings von Foerster says, “Life is studied in vivo, not in vitrio.” As an aphorism for catching the import of Green's reading of Pask, I suggest the following: “One cannot not do cosmology”. The, in my view, missing second-order turn might be captured similarly by the aphorism, “One cannot not do culture”.

Bernard Scott Cranfield University, Royal Military College of Science, Shrivenham, UK