Tribute to Stafford Beer 9781845443900, 9780861769407

Citation preview

kyb_cover_(i).qxd

4/2/04

11:29 AM

Page 1

Volume 33 Number 3/4 2004

ISBN 0-86176-940-6

ISSN 0368-492X

Kybernetes The International Journal of Systems & Cybernetics Tribute to Stafford Beer Guest Editor: Raul Espejo

Selected as the official journal of the World Organisation of Systems and Cybernetics

www.emeraldinsight.com

Kybernetes

ISSN 0368-492X

The International Journal of Systems & Cybernetics

Volume 33 Number 3/4 2004

Tribute to Stafford Beer Guest Editor Raul Espejo

Access this journal online __________________________ 484 Editorial advisory board ___________________________ 485 Abstracts and keywords ___________________________ 486 Preface ___________________________________________ 491 Foreword Raul Espejo ____________________________________________________

492

Part I The science of the unknowable: Stafford Beer’s cybernetic informatics Andrew Pickering _______________________________________________

499

Sequences of failure in complex socio-technical systems: some implications of decision and control David Weir ____________________________________________________

Access this journal electronically The current and past volumes of this journal are available at:

www.emeraldinsight.com/ISSN 0368-492X.htm You can also search over 100 additional Emerald journals in Emerald Fulltext at:

www.emeraldinsight.com/ft See page following contents for full details of what your access includes.

522

CONTENTS

CONTENTS continued

The market economy unchecked: another version of fundamentalism? A polemical tribute to Stafford Beer Allenna Leonard ________________________________________________

538

Tigers at play: Stafford Beer’s poetry David Whittaker ________________________________________________

547

Part II City planning – ‘‘dissolving’’ urban problems insights from an application of management cybernetics Markus Schwaninger and Markus Koerner __________________________

557

The role of information and models in regulating complex commercial systems Stephen Brewis _________________________________________________

577

A self-organizing network for the systems community Jose´ Pe´rez Rı´os _________________________________________________

590

Part III Observing experiences with the VSM Werner Schuhmann _____________________________________________

609

Viability versus tribalism Trevor Hilder __________________________________________________

632

Observations on the development of cybernetic ideas in Colombia – a tribute to Stafford Beer German Bula___________________________________________________

647

Heinz and Stafford – a tribute to Heinz von Foerster and Stafford Beer Rebeca Donoso _________________________________________________

659

Part IV The footprint of complexity: the embodiment of social systems Raul Espejo ____________________________________________________

671

A Latino American Requiem for Stafford Beer Roberto Zarama, Jose´ Bermeo, Nelson Lammoglia and Jairo Villamil______

701

Fighting for science Gerard de Zeeuw________________________________________________

717

Implications for Beer’s ontological system/ metasystem dichotomy Maurice Yolles__________________________________________________

726

Part V

CONTENTS

Reflections of a Cybernetician on the Practice of Planning Stafford Beer __________________________________________________

continued 767

World in Torment: A Time Whose Idea Must Come Stafford Beer __________________________________________________

774

Knowing Norbert Stafford Beer __________________________________________________

804

Man in a garrulous silence Stafford Beer __________________________________________________

809

Ten pints of Beer: The rationale of Stafford Beer’s cybernetic books (1959-94) Discussion with Stafford Beer _____________________________________

828

A filigree friendship Stafford Beer __________________________________________________

843

What is cybernetics? Stafford Beer __________________________________________________

853

www.emeraldinsight.com/k.htm As a subscriber to this journal, you can benefit from instant, electronic access to this title via Emerald Fulltext. Your access includes a variety of features that increase the value of your journal subscription.

How to access this journal electronically To benefit from electronic access to this journal you first need to register via the Internet. Registration is simple and full instructions are available online at www.emeraldinsight.com/rpsv/librariantoolkit/emeraldadmin Once registration is completed, your institution will have instant access to all articles through the journal’s Table of Contents page at www.emeraldinsight.com/0368-492X.htm More information about the journal is also available at www.emeraldinsight.com/k.htm Our liberal institution-wide licence allows everyone within your institution to access your journal electronically, making your subscription more cost effective. Our Web site has been designed to provide you with a comprehensive, simple system that needs only minimum administration. Access is available via IP authentication or username and password.

Additional complementary services available Your access includes a variety of features that add to the functionality and value of your journal subscription: E-mail alert services These services allow you to be kept up to date with the latest additions to the journal via e-mail, as soon as new material enters the database. Further information about the services available can be found at www.emeraldinsight.com/ usertoolkit/emailalerts Research register A web-based research forum that provides insider information on research activity world-wide located at www.emeraldinsight.com/researchregister You can also register your research activity here. User services Comprehensive librarian and user toolkits have been created to help you get the most from your journal subscription. For further information about what is available visit www.emeraldinsight.com/usagetoolkit

Key features of Emerald electronic journals

Choice of access

Automatic permission to make up to 25 copies of individual articles This facility can be used for training purposes, course notes, seminars etc. This only applies to articles of which Emerald owns copyright. For further details visit www.emeraldinsight.com/copyright

Electronic access to this journal is available via a number of channels. Our Web site www.emeraldinsight.com is the recommended means of electronic access, as it provides fully searchable and value added access to the complete content of the journal. However, you can also access and search the article content of this journal through the following journal delivery services: EBSCOHost Electronic Journals Service ejournals.ebsco.com Huber E-Journals e-journals.hanshuber.com/english/index.htm Informatics J-Gate www.J-gate.informindia.co.in www.ingenta.com Minerva Electronic Online Services www.minerva.at OCLC FirstSearch www.oclc.org/firstsearch SilverLinker www.ovid.com SwetsWise www.swetswise.com TDnet www.tdnet.com

Online publishing and archiving As well as current volumes of the journal, you can also gain access to past volumes on the internet via Emerald Fulltext. Archives go back to 1994 and abstracts back to 1989. You can browse or search the database for relevant articles. Non-article content Material in our journals such as product information, industry trends, company news, conferences, etc. is available online and can be accessed by users. Key readings This feature provides abstracts of related articles chosen by the journal editor, selected to provide readers with current awareness of interesting articles from other publications in the field. Reference linking Direct links from the journal article references to abstracts of the most influential articles cited. Where possible, this link is to the full text of the article. E-mail an article Allows users to e-mail links to relevant and interesting articles to another computer for later use, reference or printing purposes.

Emerald Customer Support For customer support and technical help contact: E-mail [email protected] Web www.emeraldinsight.com/customercharter Tel +44 (0) 1274 785278 Fax +44 (0) 1274 785204

EDITORIAL ADVISORY BOARD A. Bensoussan President of INRIA, France V. Chavchanidze Institute of Cybernetics, Tbilisi University, Georgia A.B. Engel IMECC-Unicamp, Universidad Estadual de Campinas, Brazil R.L. Flood Hull University, UK F. Geyer The Netherlands Universities Institute for Co-ordination of Research in Social Sciences, Amsterdam, The Netherlands A. Ghosal Honorary Fellow, World Organisation of Systems and Cybernetics, New Delhi, India R. Glanville CybernEthics Research, UK R.W. Grubbstro¨m Linko¨ping University, Sweden Chen Hanfu Institute of Systems Science, Academia Sinica, People’s Republic of China G.J. Klir State University of New York, USA Yi Lin International Institute for General Systems Studies Inc., USA

K.E. McKee IIT Research Institute, Chicago, IL, USA M. Ma˘nescu Academician Professor, Bucharest, Romania M. Mansour Swiss Federal Institute of Technology, Switzerland K.S. Narendra Yale University, New Haven, CT, USA C.V. Negoita City University of New York, USA W. Pearlman Technion Haifa, Israel A. Raouf Pro-Rector, Ghulam Ishaq Khan (GIK) Institute of Engineering Sciences & Technology, Topi, Pakistan Y. Sawaragi Kyoto University, Japan B. Scott Cranfield University, Royal Military College of Science, Swindon, UK D.J. Stewart Human Factors Research, UK I.A. Ushakov Moscow, Russia J. van der Zouwen Free University, Amsterdam, The Netherlands

Editorial advisory board

485

Kybernetes Vol. 33 No. 3/4, 2004 p. 485 # Emerald Publishing Group Limited 0368-492X

K 33,3/4

486

Kybernetes Vol. 33 No. 3/4, 2004 Abstracts and keywords # Emerald Group Publishing Limited 0368-492X

The science of the unknowable: Stafford Beer’s cybernetic informatics Andrew Pickering Keywords Control system analysis, Network synthesis, Cybernetics, Information This paper explores the history of Stafford Beer’s work in management cybernetics, from his early conception and simulation of an adaptive automatic factory and associated experimentation in biological computing, through the development of the Viable System Model and the Team Syntegrity technique for discussion and planning. It also pursues Beer into the fields of micro- and macropolitics and spirituality. The aim is to show that all of Beer’s projects can be understood as specific instantiations and workings out of a cybernetic ontology of unknowability and becoming: a stance that recognises that the world can always surprise us and that we can never dominate it through knowledge. The thrust of Beer’s work was, thus, to construct systems that could adapt performatively to environments they could not fully control. Sequences of failure in complex sociotechnical systems: some implications of decision and control David Weir Keywords Complexity theory, Sociotechnical theory, Cybernetics This paper reviews the framework proposed by Stafford Beer in his seminal work, Decision and Control, first published in 1966, and argues that these ideas are centrally relevant to a topic on which research interest has developed subsequently, the study of crises, catastrophes and disasters in complex sociotechnical systems in high technology sectors. The problems are not discussed simply in terms of system parameters like variety, redundancy and complexity. Much empirical research supports the view that these systems typically operate in degraded mode. Such variables as hierarchical position, actors’ motivations and intentions are relevant to explain the ways in which communication systems typically operate to filter out messages from lower participants and to ignore the ‘‘soft signals’’ from small-scale and intermittent malfunctions.

The market economy unchecked: another version of fundamentalism? A polemical tribute to Stafford Beer Allenna Leonard Keywords Cybernetics, Market economy, Social benefits, Control systems A case is made that the market economy imposes constraints and makes demands that mirror the rigidity of fundamentalist religions and that this makes answering one fundamentalism with another more likely. The denial of the world’s complexity, the dismissal of the suffering of those not among the select and the obliviousness to other perspectives that characterize fundamentalist economics or religion, create a dangerous situation for all. Systems models and tools, created by Stafford Beer and many others, provide practical means to address complexity and to achieve a fair balance of stakeholder needs and interests. These tools, and the holistic philosophy underpinning them, can and should be applied to questions of public importance. To do so, we have to follow the footsteps of Stafford Beer and other pioneers in the systems and cybernetics field.

Tigers at play: Stafford Beer’s poetry David Whittaker Keywords Cybernetics, Poetry, Literature Considers Stafford Beer as a poet. Discusses Beer’s interest in a wide variety of poetic forms. Gives some technical examples of how Beer employed Welsh and Sanskrit metre and rhythm to writing in English. Also includes a translation from the Spanish; a ballade and selections from his longest poem One Person Metagame. City planning – ‘‘dissolving’’ urban problems insights from an application of management cybernetics Markus Schwaninger and Markus Koerner Keywords Cybernetics, Project management, Control systems, Urban areas Urban development in fast-growing cities is one of the huge challenges of our time, and several projects of technical cooperation are

dedicated to this issue. The aim of this paper is to help project managers to enhance their capability of dealing effectively with the formidable complexities inherent in this kind of project. For this purpose, we explore the potential of Organizational Cybernetics and Social Systems Theory in a relatively new area of application. We have developed a set of conceptual tools that are helpful in coping with dynamic complexity in change and development projects. These tools have in common an inherent logic deriving to a great extent from Stafford Beer’s Viable System Model and the St Gall framework for systemic management. The application of the tools is illustrated by a state-of-the-art case study from the realm of Technical Co-operation – the revision of the Urban Master Plan for the City of Addis Ababa, the capital of Ethiopia. However, the toolkit is in principle also applicable to any complex project of change or development.

The role of information and models in regulating complex commercial systems Stephen Brewis Keywords Cybernetics, Information, Modelling The unsophisticated bartering systems of the past have been replaced with money and highly complex business systems. This has also required the need for management systems to direct and steer these complex business systems through commercial space seeking profitable and sustainable lineages. However, not all lineages will continue to be survival worthy, imposing pressures on businesses to seek out alternative pathways. Unfortunately, the management structures of these businesses fail to manage the complexity of the evolving operational substrate. These structures are hierarchically based and use metrics that cannot be adequately disposed through the structure that created them, resulting in the inability to effectively direct and steer the business. This paper looks at the evolution of these structures and the primary role of information and regulation to sustain value propositions in modern day commercially oriented enterprises.

A self-organizing network for the systems community Jose´ Pe´rez Rı´os Keywords Cybernetics, Control systems, Information networks The number of approaches, researchers and practitioners in Systems Thinking has been growing over the last decades and particularly in recent years. This wide and deep evolution within the systems community that is a sign of vitality has also increased the communication difficulties among its components. If the systems community wants to overcome these difficulties and help cross fertilisation, adequate structures are needed to facilitate the communication processes. In this work, we present a communication and exchange of information system specifically designed for this aim. It tries to use the possibilities offered by Internet to create a systems thinkers/practitioners virtual community with the capacity to facilitate the creation of as many new partial communities as the members of the systems community wish. The design of the network system is based on Beer’s viable system model and his recursive conception of viable systems.

Observing experiences with the VSM Werner Schuhmann Keywords Cybernetics, Management, Control systems, Systems theory This paper offers a senior manager’s reflections on experiences with the viable system model (VSM) in different situations of its application. These reflections are based on second-order observation; that is on observing the forms of first-order distinctions. Recently, systems thinking developed by Niklas Luhmann serves as an epistemological background. The VSM is recognized as a valuable tool to study and to change organizations, however, its ‘‘upgrade’’ is recommended in the light of insights, such as the protologic of distinctions or the more consistent realization of systems properties, such as operational closure and self-reference.

Abstracts and keywords

487

K 33,3/4

488

Viability versus tribalism Trevor Hilder Keywords Cybernetics, Organizations, Social behaviour The viability of an organisation can be compromised by the tribal behaviour of its members. If this is not understood, attempts to reform it will be met with resistance, since the effort will be perceived as an attack against the norms of the organisation. The paper offers examples and evidence for this, suggests source materials for understanding the phenomenon, and practical techniques for overcoming it. Observations on the development of cybernetic ideas in Colombia – a tribute to Stafford Beer German Bula Keywords Cybernetics, Transformational leadership This paper is an account of management and organisational interventions in multiple institutions and enterprises in Colombia during the past 10 years, mainly influenced by Stafford Beer’s ideas and work. It offers comments about the use of these ideas in three projects; particularly it focuses on an intervention in the National Controller’s Office. These are accounts of failure and success. However, assessments of success and failure are tempered by difficulties in appreciating the complexity of social processes and our inability to see causeeffect connections. The paper offers insights about concrete aspects of these interventions and the footprints they have left in the country. Heinz and Stafford – a tribute to Heinz von Foerster and Stafford Beer Rebeca Donoso Keywords Cybernetics, Education, Poetry In June 1996, Stafford Beer and Heinz von Foerster were invited to Colombia to contribute to a number of events, among others to a symposium about second-order auditing organized by the National Controller’s Office of Colombia in Cartagena de Indias. As part of the occasion, they visited Mompox, a colonial town in the delta of the

Magdalena river. During that trip, the author met Stafford and from there on she developed a special relationship with him, similar to that she already had with Heinz. The sad death of both great men of cybernetics within a short period of time was the origin of this tribute. The footprint of complexity: the embodiment of social systems Raul Espejo Keywords Cybernetics, Complexity theory, Communications, Social interaction This paper deals with organisational complexity, seen from the perspective of its unfolding from global to local concerns. Historically, this unfolding has produced rigid social systems, where those in power positions have forced unfair constraints over the majorities at the local level, and often excluded them. There is a need to move towards flexible, fair, social systems, inclusive in character. This transformation requires an increasing appreciation of communication problems in society and the embodiment of effective social systems. This transformation is presented as a problem-solving paradigm which requires social systems with capacity to create and produce their own meanings, with capacity to manage necessary structural couplings among existing social systems, thus making this management a heuristic to produce necessary social differentiation to overcome communication failures among existing self-producing, operationally closed, social systems. A key construct used in this paper to practically produce this management is the viable system model, developed by Stafford Beer. A Latino American Requiem for Stafford Beer Roberto Zarama, Jose´ Bermeo, Nelson Lammoglia and Jairo Villamil Keywords Cybernetics, Systems theory This tribute to Stafford Beer is written from the particular situation of a Latin American country. Within this context, we believe that it could be convenient to dissolve the systems movement to assume an ethical-political position that transforms the relations of dependence in the world-system from the

point of view of economic development. A contribution to this task is the formation of amplifiers of regulatory capacity that has been developed in Colombia. In this paper, we present some preliminary advances of autoethospoiesis and some of its future perspectives of research and action. One of our main objectives is to apply this functor with recurrent, recursive, and incursive operators to the world-system model.

Fighting for science Gerard de Zeeuw Keywords Cybernetics, Systems theory What deliverables did Stafford Beer envision when he developed his ‘‘science of effective organisation’’? This paper answers this question as: the organisations that use the distinctions of Beer’s viable system model. Such organisations are part of daily life, but develop to become knowledge by continuously striving to identify experiences that falsify their existence. They will be irreducible in the sense that any acceptable model of the organisation will be the organisation itself. The notion of knowledge involved is made explicit in the paper as a tribute to Stafford Beer’s pioneering work. It allowed Stafford Beer to introduce and develop insights that began to be developed by others only much later.

Implications for Beer’s ontological system/metasystem dichotomy Maurice Yolles Keywords Cybernetics, Management Stafford Beer developed managerial cybernetics, but there were many facets of his work. Most of his work concerned epistemology, and little concerned ontology. Not all of the aspects or implications of his work has been fully recognised, and an attempt shall be made to explore one of these. In particular, this paper explores his paradigm by considering some of the epistemologically and ontological angles. Some of the implications for Beer’s work will also be shown to have led to the creation of a virtual paradigm capable of exploring his achievements ‘‘externally’’, after Go¨del.

Reflections of a Cybernetician on the Practice of Planning Stafford Beer Keywords Cybernetics, Management The author gives his reflections as a cybernetician on the practice of planning in a presentation directed at a professional audience of planners.

World in Torment: A Time Whose Idea Must Come Stafford Beer Keywords Change management, Cybernetics, Team working Presents the full text of the Presidential Address by Stafford Beer to the Triennial Congress of the World Organization of Systems and Cybernetics, New Delhi, India, January 1993. Introduces the components of contemporary change and discusses the diagnostic approach of management cybernetics. Outlines a summary theory of autonomy and considers autonomy at the global level. Offers an account of the cybernetics of chronic societary triage, developing an analysis of triage through category A, B and C partition. Produces a summary theory of team syntegrity and discusses the power and the use of the team syntegrity model. Finally, outlines an action plan for World Syntegration.

Knowing Norbert Stafford Beer Keywords Literature, Cybernetics Gives an account of Norbert Wiener’s works and how they influenced the author. The seminal Cybernetics or Control and Communications in the Animal and Machine introduces the scientific world to the notion of cybernetics and indeed, according to the author, split it down the middle. Goes on to describe Wiener’s other major works and the impact they had not only on the author but also on other writers, researchers and scientists. Concludes with a brief description of how the author helped Wiener to arrange a European trip in 1964.

Abstracts and keywords

489

K 33,3/4

490

Man in a garrulous silence Stafford Beer Keywords Cybernetics, Culture (sociology), Systems theory Written in 1982 for a closed meeting of scholars, provides an insight into the thinking of the time and in particular to the contributions of the author to systems and cybernetics. Illustrates the challenge of technology to humankind and sees the coming of a second Tower of Babel which is as counter-productive as the first. Believed it was necessary to consider practical strategies for change. Looks at the concept of human identity, considering ‘‘Oneself: a systems viewpoint’’, ‘‘A Tripartite selfhood’’ and ‘‘The community consequence’’. Provides a schematic framework for the discussion of the social cybernetics of the human condition. Considers a basic stance, and the application of rules detected by cybernetic scholarship. Outlines strategies for inducing change and examines new dimensions for planning. Emphasizes the importance of regarding planning as a matter of obtaining recognition of the frameworks that have been introduced and then facilitating choice. Ten pints of Beer: The rationale of Stafford Beer’s cybernetic books (1959-94) Discussion with Stafford Beer Keywords Cybernetics, Organizational analysis, Management science Details a discussion with Stafford Beer about the rationale of his cybernetic books published in the twentieth century (1959-94). Includes a dialogue concerning his

contributed works: Cybernetics and Management; Decision and Control; Management Science; Designing Freedom; Platform for Change; The Heart of Enterprise; Diagnosing the System for Organisations; and Beyond Dispute. Concludes with a perspective of some of his important early work and some unpublished contributions which have been collected into one book.

A filigree friendship Stafford Beer Keywords Cybernetics, Computers This paper concerns the author’s longstanding friendship with Gordon Pask. It describes several occasions where their lives overlapped significantly, namely working with chemical computers, the founding of the Department of Cybernetics at Brunel University and working together on the design of Joan Littlewood’s Fun Palace project.

What is cybernetics? Stafford Beer Keywords Cybernetics, Systems analysis, Management science An address delivered at the University of Valladolid, Spain. Asks the Question-What is Cybernetics?. Discusses popular notions and genuine difficulties. Looks at the origins, derivations and definitions of cybernetics. Considers intrinsic control and SocioEconomic Governance in real-time. Relates cybernetics to the current world situation.

Preface Special double issue: tribute to Stafford Beer This special double issue is a tribute, in memoriam, to Stafford Beer, patron of this journal, President of the World Organisation of Systems and Cybernetics (WOSC) and one of the leading cyberneticians and systemists of his generation. He was not only a great academic and thinker but a man of wide ranging interests and activities that are well-known to his friends and colleagues, but not necessarily to a wider audience. This issue, therefore, reflects in its tributes to him, his unique contributions not only in academia but also to society at large. We would like to express our thanks to Professor Raul Espejo, the Guest Editor, and to his collaborators and contributors who have so painstakinly compiled this historic appreciation of Stafford Beer’s life and work. Many of the contributions included in this commemorative issue were based on presentations given at the event to celebrate his life held at the London School of Economics (UK) in March 2003 (Celebration of Stafford Beer’s Life and Work, 2003). The aim of this event was as the organisers said in their invitation to the celebration:

Preface

491

Working in the traditions of Warren McCulloch, Norbert Wiener and Ross Ashby Stafford Beer created and developed organisational cybernetics. He grounded his inventions, among others the Viable System Model and Team Syntegrity, in his practice as a management scientist, most notably in the political project of the Chilean Government of the early 70s. He offered a cogent, compelling and enlightened account of this practice in writings such as Decision and Control, Brain of the Firm, Designing Freedom, Platform of Change, The Heart of Enterprise and Beyond Dispute. For us the challenge is to transform his legacy into a growing body of knowledge, contributing to the production of fair societies and organisations and sustainable environments. In this meeting we want to debate and clarify the challenge of this transformation.

We believe that this special combined issue of Kybernetes has not only presented a sincere tribute to Stafford Beer’s life and work but also takes up this compelling challenge to all who knew him and his life’s endeavours. Readers may also know that we were able during Stafford Beer’s lifetime to publish, with his participation, a Special Issue of this journal (Rudall, 1993) dedicated to him. Brian H. Rudall Editor-in-Chief References Celebration of Stafford Beer’s Life and Work (2003), Kybernetes, Vol. 32 Nos 7/8, pp. 1172-4. Rudall, B.H. (Ed.) (1993), “A special issue dedicated to Stafford Beer”, Kybernetes, Vol. 22 No. 6, pp. 1-112.

Kybernetes Vol. 33 No. 3/4, 2004 p. 491 # Emerald Group Publishing Limited 0368-492X

K 33,3/4

Foreword Raul Espejo

492

Kybernetes Vol. 33 No. 3/4, 2004 pp. 492-495 q Emerald Group Publishing Limited 0368-492X

Following Stafford Beer’s death in August 2002, some of us felt the need to organise a meeting to celebrate his life in general and his contributions to knowledge and society in particular. The meeting took place at the London School of Economics on the 3rd March 2003, under the auspices of the LSE’s Complexity Research Programme and Syncho Ltd. The invitation to that meeting said, “Working in the traditions of Warren McCulloch, Norbert Wiener and Ross Ashby, Stafford Beer created and developed organisational cybernetics. He grounded his inventions, among others the viable system model (VSM) and team syntegrity, in his practice as a management scientist, most notably in the political project of the Chilean Government of the early 1970s. He offered a cogent, compelling and enlightened account of this practice in writings such as Decision and Control, Brain of the Firm, Designing Freedom, Platform of Change, The Heart of Enterprise and Beyond Dispute. For us the challenge is to transform his legacy into a growing body of knowledge, contributing to the production of fair societies and organisations and sustainable environments. In this meeting we want to debate and clarify the challenge of this transformation.” The format of the one-day meeting was invited presentations by academics and practitioners working in organisational cybernetics and applied epistemology and also by a poet. They all agreed to write formal contributions for publication in this issue of Kybernetes. After the event I made a few additional invitations for contributions to people who were close to Stafford and his work. Altogether 19 people contributed to this issue. I do not think I misinterpret them by saying that on the whole they were more interested to offer a tribute to Stafford than to break new grounds in organisational cybernetics; any progress in this direction should be received as a bonus, and not as a primary concern of this issue of Kybernetes. On the other hand, seeing all the contributions together gives a sense of the breadth and depth of some of the work in progress following Stafford’s footsteps. But they are not comprehensive. Unfortunately, there is no report of the work in progress using Stafford’s team syntegrity. This is an area of research and professional work that hopefully others will write about in the near future. How readers make sense of the 15 contributions is indeed a very personal matter and most likely there will be as many interpretations as readers. However, I had the task to give them some order. I found that they could be organised in four main strands, before a final part that included some of Stafford’s papers, published by Kybernetes from 1999 to 2002. A first group of papers offers insights about Stafford’s work and the strength of his thinking. In his paper, Andrew Pickering explores the history of

Stafford’s work in management cybernetics, from his early work in the steel industry to the development of the VSM and team syntegrity. In his view Stafford’s work makes apparent that the world can always surprise us, and that we can never dominate it through knowledge. David Weir’s paper makes use of some of Stafford’s ideas in Decision and Control, published in 1966, and argues that they are centrally relevant to the management of crises, catastrophes and disasters in complex socio-technical systems in high technology sectors. He makes apparent that topics such as redundancy, poor structures and weaknesses in information flows are critical for social systems, which typically operate in a degraded mode. Allenna Leonard highlights vividly in her paper one of Stafford’s great strengths: his ability to review holistically questions of public importance. She illustrates how the denial of the world’s complexity, characteristic of the market economy and religious fundamentalism, produces exclusion and dangerous situations for all. She argues for system models and tools, as offered by Stafford and others, to increase our appreciation of complexity. A refreshing pause, before proceeding to the second group of contributions, is offered by David Whittaker’s review of Stafford poetry. It would appear from Whittaker’s contribution that poetry was, among other aspects, a vehicle for Stafford to cross cultural boundaries. Stafford employed Welsh and Sanskrit metre and rhythm to writing in English. He also “transduced” rather than translated into English popular Spanish-language poems. The second group of contributions is focused on the application of systems thinking in general and the use of Stafford’s VSM in particular. Schwaninger and Koerner develop conceptual tools, derived from Stafford’s VSM and the St. Gall framework for systemic management, to cope with dynamic complexity in change and development projects. These tools are used in the revision of the Urban Master Plan for the City of Addis Ababa, the capital of Ethiopia. At a micro-economic level, Steve Brewis’s paper is focused on developing system tools to increase businesses’ capacity to deal with complexity. He recognises that the management structures of these businesses are, in general, inadequate to handle their operational, moment-to-moment complexity. Improving this situation requires structures that anticipate and mirror this complexity; he suggests that the VSM is particularly appropriate for this purpose. This model is also used by Jose Perez Rios in his contribution to overcome communication difficulties across dispersed communities. He explores and offers a VSM design for an Internet information exchange network. Its purpose is to enable people working in system ideas to form a global virtual community. A third group of contributions is focused on interventions in companies and government institutions influenced by Stafford’s paradigmatic thinking. Authors reflect upon their experiences working with these ideas and make apparent some of the difficulties in bringing about social transformation. Werner Schuhmann offers a senior manager’s reflections on experiences with

Foreword

493

K 33,3/4

494

the VSM in different situations. These reflections are based on second order observation; that is on observing the forms of first order distinctions. He recognizes the VSM as a valuable tool to study and change organizations, however, suggests that its “upgrade” may be necessary in the light of epistemological insights such as the logic of distinctions, operational closure and self-reference. Trevor Hilder, based on interventions in different enterprises, and also on historical evidence, argues that tribal behaviours are not only exclusive but also threaten the viability of organisations. Indeed, if this is not considered attempts to reform them will be met with resistance, since the effort will be perceived as an attack against the norms of the organisation. The VSM can be seen as a heuristic to diagnose this kind of behaviours. German Bula’s paper offers an account of three processes of organisational transformation in Colombia, driven by organisational cybernetics in general and the VSM in particular. These processes, which started in 1994 and are still in progress, have influenced processes in hundreds of public and private enterprises. His account of these projects highlights not only the relevance of the work within the National Controller’s Office (or National Audit Office), in itself a major undertaken, but also the current training at the Universidad de los Andes of hundreds of people in the practices of “observing organisations”. His accounts are accounts of failure and also of success. A glimpse of Stafford extraordinary humanity is offered by Rebeca Donoso’s tribute. This is an account of her first encounter with Stafford in a remote town in the delta of the Magdalena River in Colombia, and the impact of this event in her life afterwards. This paper is one instance of the transformational processes in progress in that country, as described by Bula in the previous paper. The fourth group of contributions are mainly epistemological, and to a lesser extent methodological, reflections about the VSM. My own paper deals with organisational complexity, seen from the perspective of its unfolding from global to local concerns. Historically this unfolding has produced rigid social systems, where those in power positions have forced unfair constraints over the majorities at the local level, and often excluded them. I argue the need to move towards flexible, fair, social systems, inclusive in character. This transformation requires an increasing appreciation of communication problems in society and the embodiment of social systems. The paper by Roberto Zarama and his colleagues at the Universidad de los Andes is a tribute to Stafford written by people at the periphery of the global context. It argues, following Stafford’s cry for systems sensibility, for an ethical stance that is aware of the consequences of our social actions. A contribution to this task is their work at the University de los Andes training qualified observers of social systems (cf. paper by Bula). They give powerful insights about their current research and action. Gerard de Zeeuw’s paper offers deep epistemological insights about Stafford’s work. His basic question is, what deliverables did Stafford envision when he developed his “science of effective organisation”?

The paper answers this question as: the organisations that use the distinctions of the VSM. Such organisations, part of daily life, become knowledge by continuously striving to identify experiences that falsify their existence. This view clarifies the notion of a collective’s organisation as a knowledge generator. These are insights that began to be developed by others only much later. Last by no least is Maurice Yolles’s contribution; he argues that some aspects of Stafford’s work have not been fully recognised. He gets himself in the endeavour of exploring Stafford’s paradigmatic contributions, by considering epistemological and ontological angles of his work. Some of the implications of this exploration for Stafford’s work are shown to have led to the creation of a virtual paradigm capable of exploring his achievements “externally”, after Go¨del. In part V a collection of Stafford’s contributions to Kybernetes, published from 1993 to 2002, have been included. Of course each of the contributions stands on its own, as an independent, self-contained contribution, however their ordering shows a path, useful to me, to make sense of all of them as a totality. I have included four of Stafford’s illustrations as entries to each of the first four parts of this issue of Kybernetes. They were copied from “Reproof of the Soul, Haiku by Vanilla Beer, Illustrated by Stafford Beer” (www.vanillabeer.com), with the permission of the author, herself a painter. In her own words: “we (Stafford and Vanilla) both thought it was funny that he should illustrate a book I had written – swapping roles”. Finally, I want to thank Allenna Leonard for her collaboration, in many different ways, throughout the editing of this issue.

Foreword

495

File name{KYB}Issues/33_3-4/Part 1.3d 09:50 – 1/4/04 – Total page(s) 3

PART I Illustration by Stafford Beer (I)

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

The science of the unknowable: Stafford Beer’s cybernetic informatics

The science of the unknowable

499

Andrew Pickering Department of Sociology, University of Illinois, Illinois, USA Keywords Control system analysis, Network synthesis, Cybernetics, Information Abstract This paper explores the history of Stafford Beer’s work in management cybernetics, from his early conception and simulation of an adaptive automatic factory and associated experimentation in biological computing, through the development of the Viable System Model and the Team Syntegrity technique for discussion and planning. It also pursues Beer into the fields of micro- and macropolitics and spirituality. The aim is to show that all of Beer’s projects can be understood as specific instantiations and workings out of a cybernetic ontology of unknowability and becoming: a stance that recognises that the world can always surprise us and that we can never dominate it through knowledge. The thrust of Beer’s work was, thus, to construct systems that could adapt performatively to environments they could not fully control.

This paper derives from a larger project exploring the history of cybernetics in Britain during and after World War II [1]. The project focuses on the work of four British cyberneticians – Grey Walter, Ross Ashby, Stafford Beer and Gordon Pask; here I focus on Stafford Beer, the founder of the field called management cybernetics, and his work in informatics [2]. Anthony Stafford Beer was born in London in 1926. He joined the British Army in 1944 after just 1 year as an undergraduate in London, and served in India and Britain. He left the army in 1949, and between 1949 and 1970 he worked in the steel and publishing industries and ran his own consulting company. From 1970 until his death in August 2002, he worked as an independent management consultant (Times, 2002) (Figure 1). I begin with an overview of Beer’s general perspective on information science and information systems, and also intends bring out the singularity of the cybernetic approach. From 1950s onwards, Beer was a remorseless critic of the ways in which computers were being deployed in the industry, essentially to replace the existing paper systems. He felt that this did nothing to change the existing organisational forms, and that something more imaginative was required. His argument was that the postwar world was a new kind of world. Specifically, the pace of change had increased markedly since the war, and the important thing for organisations was that they should be adaptive – light on their feet and ready to accommodate themselves to the new situations which would arise faster and faster as time passed [3]. For a special issue of Kybernetes in memory of Stafford Beer, edited by Raul Espejo.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 499-521 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523535

K 33,3/4

500

Figure 1. Stafford Beer (a) in the early 1960s, and (b) in 1975

To render organisations adaptable, according to Beer, required reorganising them to make possible specific patterns of information flow and transformation. We can turn to some examples shortly, but first I want to emphasise the gap between the mainstream vision of informatics and the cybernetic. What we need to think about here is ontology – the question of what the world is like. Mainstream informatics presumes a very familiar ontology. The world is a regular, law-like place that can be known more or less exhaustively. It is a place that can therefore be controlled and dominated through knowledge. That is the logic behind the creation of bigger databases and faster information systems. Of course, this ontology does recognise the existence of the unknown, but only as something to be conquered, to be drawn into the realm of the known. Cybernetics turned this picture inside out and exemplified a different and much less familiar ontology. Beer (1959, p. 17) argued that there exists in the world a class of “exceedingly complex systems”, including the brain, firm and economy, which are in principle unknowable. However, we gather much data on them, and we can never know them completely, they can always surprise us. Such systems can never be dominated by knowledge, and instead we have to learn somehow to cope with them, and cybernetics was, then, the science of dealing with the unknown, the science of adaptation – an extremely odd sort of science [4]. This ontology of the unknowable is the key thing to grasp in thinking about cybernetics. Two corollaries of it are worth mentioning. First, it thematises time. By definition, one has to deal with the unknown in time, as it happens. No amount of information about the past, as stored in conventional information systems, can ever prepare us for genuine unpredictable novelty. As Beer (1972, p. 199) ironically put it: “Look straight ahead down the motorway while

you are driving flat out. Most enterprises are directed with the driver’s eyes fixed on the rear-view mirror”. This real-time/retrospective contrast is an important angle on the specificity of cybernetics. The other corollary is this. Conventional informatics is, as I would say, representational – meaning, again, that it is all about the accumulation of data and knowledge. One might eventually want to draw upon that knowledge for action, but that is not the defining feature of an information system. The information system is, as it were, detachable from the action. Cybernetics viewed information systems differently. If we have to continually deal with the unexpected as a practical matter, then the accumulation of representational knowledge seems less relevant. Instead what one wants is a performative information system, geared straight into the action, not detachable at all. One would not care exactly what information was flowing through the system and how, as long as its output was an adaptive transformation of the organisation to its environment. This contrast between the representationalism of conventional information systems and the performativity of cybernetic ones is very important. In 1962, in one of his more visionary moments, Beer described electronic computers as dinosaurs, looking forward to the day when they would be supplanted by another class of information processing devices that simply would not have representational intermediate states at all (Beer, 1962a, p. 220) [5]. When I came across that idea I was amazed. Something beyond the computer? What is this man talking about? Is he mad? Beer was not mad. Now, we turn to history, starting with a brief detour through the work of Beer’s friend, W. Ross Ashby (Figure 2), born in 1903, died in 1972, the doyen of the English cyberneticians [6]. We need to think especially about a device that Ashby built in his spare time in 1948, his famous homeostat (Ashby, 1948, 1952). This was an electromechanical device intended to mimic

The science of the unknowable

501

Figure 2. Ross Ashby

K 33,3/4

502

the biological process of homeostasis – the ability of organisms to maintain “essential variables” such as blood temperature constant in the face of fluctuations in their environment. Without going into details, in Ashby’s homeostat the essential variable was the electric current flowing through a moveable needle dipping into a trough of water on top of the device, and the machine’s environment was constituted by electrical interconnections to other homeostats. The trick in maintaining homeostasis was that when the current within a given homeostat went beyond some preassigned limit, a relay would trip, operating a stepping switch which would change the electrical resistance of the homeostat’s inner circuitry, with the sequence of different values for the resistance being determined from a table of random numbers. The homeostat would thus, as it were, randomly reconfigure itself. If the current were to continue to go beyond its limit, the machine would reconfigure itself again and again until homeostasis was achieved. The homeostat was thus, as Ashby called it, an ultrastable device; whatever one did to it, it would eventually find its way back to homeostatic equilibrium with its environment. It was a device for staying the same. Another British cybernetician, Walter (1953), sarcastically referred to it as Machina Sopora. I need to make three remarks on the homeostat (Figure 3). First, I hope it is clear how it fits in with my earlier remarks on ontology. The homeostat was a device that dealt with unknown. It did this in real time – it reacted to fluctuations in its environment as they happened, and it did so in a performative rather than a representational fashion: it did not seek to know the world representationally; it simply materially reconfigured itself as the occasion arose. If you have the hang of that, then you have the hang of what was most distinctive about British cybernetics. If orreries – those beautiful early-modern models of the solar system – were the mechanical emblems of the ontology of the knowable, then the homeostat was the emblem of the cybernetic ontology of unknowability. Second, the homeostat was the centrepiece of Ashby’s first book, Design for a Brain (1952), and Ashby intended it as a model of the brain inasmuch as it learnt to cope with its environment. But, again, it was a performative brain – as distinguished, for example, from the rational representationalist brain that was later exemplified symbolically as AI. My third remark is this. We can think of the homeostat as a controller, and much of Ashby’s cybernetics focused precisely on questions of control (Ashby, 1956) [7]. His key result here was the Law of Requisite Variety – Ashby’s law, as Beer called it. Variety is a measure of the number of states a system can take up – 25, as it happens, in the case of Ashby’s first homeostats, with their different possible internal electrical resistances. The Law of Requisite Variety stated that a system could succeed as a homeostatic controller only if it disposed of as much variety as the environment in which it existed. A homeostat could maintain its ultrastable condition when connected to

The science of the unknowable

503

Figure 3. The homeostat: partial photograph of four interconnected homeostat units

another homeostat with the same number of internal states, but might fail against one having twice that number. That is enough to get us back to Stafford Beer. If Ashby did cybernetics as a “pure science”, Beer was an applied cybernetician, which is what interests me – I am especially interested in what cybernetics looked like when it was put to work in the world. Ashby’s homeostat was at the heart of all of Beer’s attempts to conceptualise and design adaptive organisations, and now we can go through some of these as they emerged in the period from 1950s to 1970s. In the 1950s, Beer’s cybernetics revolved around the contemporary fantasy of the “automatic factory”, in which all operations were to be controlled by computers rather than people. Beer (1962a, p. 164) likened current visions of the automatic factory to a “spinal dog” – a dog whose nervous system had been severed below the brain. Such an animal can, apparently, continue to live and display bodily reflexes, but it cannot learn and adapt to the changing circumstances. Adding a brain is thus required, to move from an automatic factory to the cybernetic factory, and this, Beer argued, should be an Ashbean homeostat.

K 33,3/4

504

Figure 4. The cybernetic factory

I cannot go into detail here, so let me discuss some figures from a major paper written by Beer in 1960 by way of an overview (Beer, 1962a). Figure 4 is a logic diagram of the cybernetic factory. The T- and V-machines are what we would call neural nets: the T-machine collects data on the state of the factory and its environment and translates them into meaningful form; the V-machine reverses the operation, issuing commands for action in the spaces of buying, production and selling. Between them lies the U-machine, which is the homeostat, the artificial brain, which seeks to find and maintain a balance between the inner and outer conditions of the firm – trying to keep the firm operating in a liveable segment of phase-space. Figure 5 is a more suggestive figure, a painting by Beer, labelled “general picture of the whole theory” (the T-, U- and V-machines are indistinctly labelled in the smaller painting at the lower left). The cybernetic factory was not a pure theory. By 1960, Beer had at least simulated a cybernetic factory at Templeborough Rolling Mills, a subsidiary of his employer, United Steel, and the next figure might help us to understand things better. In Figure 6, the lines of circles and squares marked “sensation”

The science of the unknowable

505

Figure 5. The factory as brain

to “judgements” correspond to the numerical inputs to the T-machine: tons bought, cost of raw material, cash at bank, value of credits, etc. At Templeborough, all these data were statistically processed, analysed and transformed into 12 variables, first six referring to the inner state of the mill, and next six to its economic environment. Figures were generated at the mill every day – as close to real time as one could get – and each day’s figures were stored as the “generalised gestalt memories” indicated at the lower left and right of the figure. Beer claimed to see how all these data collection and processing, including changes in the classification system, could be accomplished automatically, although in fact it was still done clerically in the mill according to protocols devised by operational research (OR) scientists – this was one sense in which the mill had become a simulation of a fully cybernetic factory.

K 33,3/4

506

Figure 6. Simulation of a cybernetic factory

The other sense of simulation concerned the U-machine. As indicated in the lower centre of the figure, the two gestalt memories of the factory defined two phase-spaces in terms of the relevant parameters, and the job of the U-machine was to strike a homeostatic balance between them. But nothing like a functioning U-machine had yet been devised. The U-machine at Templeborough was still constituted by the decisions of human managers, though currently they were precisely positioned in an information space defined by the simulated T- and V-machines. Hence, by 1960, Beer had constructed a simulation of a cybernetic factory that promised to dispense entirely with human personnel, though humans in fact still filled the gaps for machines which were not yet in place. Beer could see how to complete the automatic T- and V-machines, though the U-machine remained unspecified. Nevertheless, Beer (1962a, p. 212) wrote, “Before long a decision will be taken as to which fabric to use in the first attempt to build a U-machine in actual hardware (or colloid, or protein)”. The vision of the adaptive factory not just running smoothly, but also evolving and changing all on its own without any human intervention is itself amazing, but Beer’s attempts to construct the U-machine homeostat are where the story gets really interesting. The requirements for the U-machine were that, first, it should be able to internally reconfigure itself, like Ashby’s original homeostat, and that, second, in accordance with Ashby’s law, it must have high variety, to have a chance of coping with the complexity of its environment. These days, we might think of somehow programming a computer to fulfil this function, but Beer argued that this was not necessarily the way to go. Computers were extremely expensive in 1950s and 1960s, and besides, Beer (1962a, p. 209, 215) had come up with a different idea: As a constructor of machines man has become accustomed to regard his materials as inert lumps of matter which have to be fashioned and assembled to make a useful system. He does not normally think first of materials as having an intrinsically high variety which has to be constrained. . . [But] We do not want a lot of bits and pieces which we have got to put together. Because once we settle for [that], we have got to have a blueprint. We have got to design the damn thing; and that is just what we do not want to do.

What is all this about? Ashby had built an electromechanical equivalent of a homeostatic biological system and called it a brain. Beer’s idea was to turn Ashby’s idea by another 1808: he wanted somehow to enrol a naturally occurring homeostatic system as the brain of the cybernetic factory. He had conceived the idea, I would say, of a non-representational, adaptive, biological computer. This was the machine which he hoped would supercede the electronic computer; the referent of his remark about dinosaurs. During the second half of the 1950s, Berar (1959, p. 162) embarked on “an almost unbounded survey of naturally occurring systems in search of materials for the construction of cybernetic machines”.

The science of the unknowable

507

K 33,3/4

508

Beer (1962b) wrote a brief report on the state-of-the-art, which fairly makes mind-boggling reading. Let me just mention some of the systems he discussed there to convey the flavour of it. The list includes a successful attempt to use positive and negative feedbacks to train young children to solve simultaneous equations without teaching them the relevant mathematics – to turn children into a performative (rather than cognitive) mathematical machine – and it goes on to discuss an extension of the same tactics to mice! This is, I would guess, the origin of the mouse-computer that turns up in both Douglas Adams’ HitchHikers Guide to the Universe and Terry Pratchett’s Discworld series of fantasy novels [8]. Beer also reported attempts to induce small organisms, Daphnia collected from a local pond, to ingest iron filings so that input and output couplings to them could be achieved via magnetic fields, and another attempt to use a population of the protozoon Euglena via optical couplings. (The problem was how to contrive inputs and outputs to these systems.) Beer’s last attempt in this series was not to use specific organisms, but an entire pond ecosystem as a homeostatic controller, on which he reported that, “Currently there are a few of the usual creatures visible to the naked eye (Hydra, Cyclops, Daphnia, and a leech); microscopically there is the expected multitude of micro-organisms. . .. The state of this research at the moment,” he said, “is that I tinker with this tank from time to time in the middle of the night” Beer (1962b, p. 31). In the end, this wonderful line of research was founded, not on any point of principle, but on Beer’s practical failure to achieve a useful coupling to any biological system of sufficiently high variety. I do want to note, however, that I admire Beer’s imagination enormously in this phase of his work [9]. I also want to mention that it is clear from subsequent developments that the homeostatic system, Beer really had in mind, was something like the human spinal cord and brain. He never mentioned this in his work on biological computers, but the image that sticks in my mind is that the brain of the cybernetic factory should really have been an unconscious human body, floating in a vat of nutrients and with electronic readouts tapping its higher and lower reflexes – something vaguely reminiscent of the movie The Matrix. This horrible image helps me to appreciate the magnitude of the gap between the cybernetic information systems and more conventional approaches. Now, we can return to something more like normality. Beer’s dreams of biological controllers came to an end in the early 1960s, but this provoked a transformation rather than an abandonment of his vision of the cybernetic factory. Beer’s (1972) book, Brain of the Firm, laid out a new vision of what he called the viable system model (VSM) [10]. The VSM took up Beer’s earlier plan for a cybernetic factory and transformed it along two axes. First, the simulation of the cybernetic factory just discussed became, in effect, the thing itself. Beer dropped the ambition to dispense entirely with human beings and instead argued that human managers should be positioned within purposefully designed information flows at just

those points that would ideally have been occupied by homeostatic ponds or trained mice. Second, Beer extended and elaborated his conception of information flows considerably. The aim of the firm had to be to survive in an environment that was fluctuating and changing. How was this now to be accomplished? The place to look for inspiration, according to Beer, was, once more, nature. Biological organisms have already mastered the trick of survival and adaptation, and Beer’s idea was therefore to read biological organisms as exemplary viable systems in general – we should transplant their key features to the structure of the firm. In particular, Beer chose the human nervous system as his model. If his original idea was that the firm needed to contain an artificial brain (made of magnetic Daphnia or leeches), the idea of the VSM was that the firm should become a brain,a cyborg brain with human brains lodged within it. (Another weird image, if you think about it too hard.) The spirit of the VSM is nicely expressed in the juxtaposition of two figures from Brain of the Firm: one, a schematic of the human body; and other of the firm (Figure 7). Very briefly, Beer argued that one needs to distinguish, at minimum, five levels or systems of control in any viable system. In this figure, System one consists of four subsidiaries of a larger organisation, labelled A, B, C and D, analogous to arms and legs, the heart, kidneys, etc. System two, the equivalent of the sympathetic nervous system, connects them to one another and to System three, and seeks to damp out destructive interactions between the subsidiaries. System three – the pons and medulla of the VSM – consists of a set of OR models of production that enables management to react to fluctuations in Systems one and two – by reallocating resources, for example. System

The science of the unknowable

509

Figure 7. Control systems in (a) the firm, and (b) the human body

K 33,3/4

510

four – the base of the brain itself – was envisaged as a decision-making environment for higher management, modelled on the World War II operations room. It would collect and display information from the lower systems and from the outside world and, very importantly, it would run a set of computer programs that higher management could consult on the possible future effects of major decisions. At the same time, this operations room was intended to function as a club-room for senior management – a place to hang out, even when major decisions were not at stake. Finally, System five was the location of the most senior management whom Beer regarded as the cortex of the firm. Their vision of the firm and its future, whatever it was, was to be negotiated into reality in reciprocally vetoing homeostatic interactions with System four. Here, I should return to the question of ontology. Despite my earlier emphasis on performance vs representation, it is clear that the VSM did incorporate significant representational elements, especially the computer models running in Systems three and four. But one should note that one function of the programs running at the System three level was statistical filtration – that is, to junk almost all the information that arrived there rather than to store it. Second, in the VSM the models at levels three and four were to be continually updated in comparisons between their predictions and the actual performance of the firm and its environment. This updating recognised even in the realm of representation that the world remained an unknowable place; the utility of the models had to be found out continually in real-time experience. It is important to emphasise that the VSM was not a theoretical conceit. All of Beer’s consulting work was based on it, and by the early 1980s he could already list amongst his clients small businesses and large industries, publishers, banks and insurance companies, transportation, education and health organisations, and governments and international agencies (Beer, 1989a, pp. 34-5). I cannot go into examples, but I can note that in much of this work, the VSM functioned as a diagnostic tool – comparison with the VSM diagram was a way of singling out organisational problems that needed to be addressed. Beer claimed that one could go very quickly to the heart of an organisation’s problems in this fashion – though addressing the problems took much longer. Only on one major occasion Beer had the chance to implement the VSM from the ground up – when he was invited to help design and implement a control system for the entire economy of Chile, under the newly elected Marxist regime led by Salvador Allende. From 1971 to 1973, Beer threw himself into Project Cybersyn as it was called (for “cybernetic synergy”); a lot was done in a very short period of time, and I can just summarise what was accomplished [11]. By installing telex facilities, a real-time communication network called Cybernet was established, linking much of Chile’s industrial base to one computer in Santiago. A set of programs called Cyberstride was written to process and filter the incoming data at the System three level, and another program, CHECO, was written to simulate the overall behaviour of the Chilean

economy at the System four level. The System four operations room was also getting into shape by 1973, as shown in Figure 8. This cybernetisation of the Chilean economy was an extremely ambitious project which, alas, never had chance to go into full operation. On 11 September 1973, General Pinochet launched a successful coup against the Allende government, and Allende himself died that day. Some members of Beer’s group fled the country; others were jailed. Chile and Cybersyn are as far as I want to go in tracing out Beer’s development of the VSM in management cybernetics, and I would pause now to take stock of where we have got to, before branching out in different directions. I can say this: what we have been exploring so far is Beer’s implementation of the cybernetic ontology of unknowability in the construction of adaptive information systems, and what I have been trying to show is, first, the intriguing and imaginative singularity of this work – how different it is from conventional informatics. The key contrast here is Beer’s emphasis on real-time performance rather than data-processing as a self-contained activity, thematised for me by his idea of biological rather than digital computation. Second, I have been trying to show that, despite the seeming paradoxicality of it, one can indeed construct adaptive systems, systems that adapt to and transform themselves in the face of the unknown, as in Beer’s implementations of the VSM.

The science of the unknowable

511

Figure 8. Operations room of Project Cybersyn

K 33,3/4

512

I could finish the paper at this point – there is much to think about in those aspects of Beer’s work already discussed. But there are further aspects of his work that also deserve attention here. Beer’s daughter, Vanilla, recalls that “Stafford and I generally ran Jesus and Marx together in an attempt to produce metanoyic possibilities”, so the last sections of this talk make some observations on Beer’s politics and spirituality [12]. First, politics. Beer sometimes described himself as “an old-fashioned Marxist” (Miller, 2002) or even as “somewhat to the left of Marx” [13]. Hence, no doubt, his enthusiasm for the Chile project. But his writings cannot be construed as contributions to orthodox Marxist debates, and we need to look at them from a different angle [14]. Running through Beer’s cybernetics is a profound concern for democracy, at the level of both theory and practice. The theoretical concern followed directly from the Law of Requisite Variety. In our dealings with others, it is imperative that we dispose sufficient variety to accommodate their variety, and vice versa, so decision-making should take the form of a homeostatic reciprocal reconfiguration of all parties concerned – rather than, say, the imposition of the will of some on others. But what interests me most is not Beer’s democratic theory, but his attempts to, as it were, engineer democracy, to make it real. What is at stake here? An ideal democracy might be one in which everyone discusses any given issue with everyone else until a conclusion eventually emerges that all are willing to abide by. The trouble with this ideal is that it quickly becomes practically impossible as the number of people involved increases. There are many ad hoc solutions to this problem to be found in the present world, ranging from committee meetings of small numbers of representatives up to simple voting in national elections. Beer was very critical of all of these in their handling of variety. The agenda of committee meetings is easily rigged to define the possible outcomes, for example, leaving little space for open-ended adaptation. Beer’s problem can thus be understood as one trying to design better arrangements for democratic and adaptive decision-making. If conventional politics is about advocating specific plans of action, then Beer’s work has to be seen as a species of sub-politics or infra-politics – the attempt to establish a suitably democratic and adaptive ground on which conventional politics can be conducted. So much for generalities. What did this mean in practice? We can start by thinking again about the VSM. Beer repeatedly stressed various aspects of this. First, just like the comparable levels of the human body, the various components of System one of the viable system were supposed to be quasi-autonomous. Most of the time, the heart and the lungs do their own thing, quite independently of any conscious control, and Beer thus envisaged the individual divisions of the firm in System one going their own way, in charge of their own destinies, for most of the time. This was one sense in which the VSM could be said to be democratic by design. Of course, the various levels were also

interconnected and responsive to one another – they were not entirely autonomous – but again Beer thought of the coupling as essentially homeostatic. Experiments at each level could be thought of as analogous to the random reconfigurations of Ashby’s mechanical homeostats, and the evolution of the entire system would be determined by a process of reciprocal vetoing between such experiments at all levels, just like an array of homeostats coming into equilibrium with one another. Again, the VSM envisaged democratic relations, here between levels, rather than unidirectional control. Beyond this coupling of levels lay the question of what the viable system was for: what did the organisation as a whole seek to hold constant in the face of its unpredictable interactions with its environment – what were the system’s goals? Again, Beer stressed the importance of not thinking of such goals as hierarchically determined, as simply given by the organisation’s “brain”. In Chile, in parallel with the technicalities of Cybersyn, Beer worked to develop what was called the “People Project”, intended to provide real-time feedback from the people on the conduct of government (Beer, 1981). One endearing aspect of this was Beer’s idea that people should be equipped with “algedometers” – dials that they would set to indicate the magnitude of their pleasure or displeasure. Beer imagined, for example, a situation in which the integrated result of everyone’s individual algedometer settings would be displayed on a giant algedometer next to a politician giving a speech on TV. Not only could the people see their own collective response to the speech, but they could also see that the politician could see it, and they could expect him or her to act accordingly – and then they could respond to that. This was one way in which everyone could contribute to “design the nation”, as Beer put it. In such ways and more, then, the VSM is a kind of techno-social diagram of an adaptive democracy – a map of how people might be arranged and connected to involve them all in their collective adaptation to a fluctuating and ultimately unknowable world. I find this very interesting. I am especially struck by the concreteness and specificity of Beer’s approach. I have read many books of political theory which go through fascinating and complicated arguments only to reach the obvious conclusion – that democracy is a good thing and we need more of it; Beer is one of the few thinkers I have come across who had anything new to say at this subpolitical level about how democracy might be made in practice. Again we see the characteristic cybernetic concern with performance rather than theory and representation. Not everyone saw the Chile project the way I have just described it, however. In Chile itself, as well as the US and Britain, Cybersyn was criticised, from left and right, as technocratic (Miller, 2002). Beer rejected the charge along the lines I have just discussed, but it is easy to see what was on the critics’ minds. The organic liveliness that Beer wanted to foster within the VSM structure could readily be denatured into a command structure – homeostatic couplings could be replaced by a one-way flow of orders – and indeed some factions in Chile

The science of the unknowable

513

K 33,3/4

514

Figure 9. The syntegration icosahedron

hoped to operate the Cybersyn apparatus just that way after the coup. It is therefore interesting to look at another of Beer’s projects that evolved over the years alongside the VSM. This was what Beer called the syntegration approach to decision-making, which implemented a quite different organisational diagram (Beer, 1994b). The question here was how to bring together a group of people to discuss their collective future without specifying any detailed agenda, and the solution this time was an appeal to geometry. Team syntegrity was based on the figure of the icosahedron – the regular three-dimensional structure that has 20 triangular faces, 12 vertices and 30 edges (Figure 9). Participants would be assigned to one edge or another, and loosely defined discussion topics initially assigned to each vertex. Each participant would then take part with the appropriate others in discussions of the two topics at the ends of their edge. The discussions might come to some conclusion on given topics, redefine them, suggest new ones, or whatever, and then the whole cycle of discussion would start up again. In this way, novel thoughts and proposals would eventually “reverberate” around the whole structure from vertex to vertex and return to the original proposers in modified form, and the upshot of the procedure would ideally be a consensual understanding of the group and its purposes that could not have been foreseen by anyone in advance.

Syntegration figured ever larger in the consulting work of Beer and others in the 1990s, and Beer (1989b, p. 122; 1994b, p. 12) described it, with some justification, as a process of “perfect democracy” – clearly it was a process with no privileged centre whatsoever. Again, what I find striking about it is its specificity: syntegration was not a theoretical argument; it was a practical sub-political set-up in which democratic deliberation and planning could be conducted in a completely open-ended fashion. So far I have been talking about what one could call the inner sub-politics of organisations. How could a firm or a nation arrange its own internal affairs to be democratically adaptive? But Beer was also concerned throughout his life with inter-organisational and international conduct: how should systems conduct their interactions with one another? Here, from 1950s onwards, Beer’s rhetoric was always one of “crisis”. Since World War II, the world has been changing faster and faster in unpredictable ways, but our institutions cannot recognise this, they are not adaptable. Instead of coping with the inexhaustible variety of the world by deploying their own variety, they seek to fix their environments, and since the world is ontologically incapable of being pinned down and fixed, this necessarily has disastrous results – typically manifest in the Third World, the environment and so on. Hence, we need to redesign our institutions cybernetically, precisely as adaptive systems. When I first encountered this rhetoric, I wanted to ignore it. It was both self-serving and dated. We all used to talk like that in the 1960s but, in fact, the world has not come to an end since then. Oddly enough, though, just while I have been writing about Beer, his stories have started to seem very relevant indeed. Everything that has happened since those planes flew into the World Trade Centre and the Pentagon speaks of an Anglo-American attempt to freeze the world, to stop it displaying any variety at all – running from endless “security” checks and imprisonment without trial to the invasions of Afghanistan and Iraq. Global politics has collapsed into one-bit discriminations (Beer, 1993, p. 33) between “us” and “them”, the goodies and the baddies – and you would have to be mad to believe that things will get better because of this instead of worse. As Beer (2001) wrote, “Last month the tragic events in New York cybernetically interpreted look quite different from the interpretation supplied by world leaders – and therefore the strategies now pursued are quite mistaken in cybernetic ways”. Of course, many interpretations of recent events are possible, involving, for instance, the concealed interests of Texas oil money. The interesting thing about Beer’s macropolitical analysis – worked out at great length in his reflections on the Chilean experience – is that it again revolved, literally and metaphorically, around a diagram of information flows, in which media and government models systematically reduce the variety recognisable in crisis situations and thus exacerbate the very crises they

The science of the unknowable

515

K 33,3/4

516

Figure 10. The cybernetics of crisis

seek to represent and manage. Here, Beer’s subpolitics extended itself into the field of intersystemic and international relations (Figure 10) [15]. Where have we got to now? One might think of informatics and politics as two distinct projects, taking place in quite different and disjoint social arenas. What fascinates me about Beer’s work is that in it, informatics and politics were continuous with one another, with his distinctive diagrams of information flows and transformations fusing the two together, with one another and with the cybernetic ontology of unknowability and with a wild, if unrealised, vision of biological computing. Now, we can move to my last topic, the spiritual aspect of Beer’s work. Beer grew up in the Anglican church, converted to Catholicism for 24 years and ended his life as a self-described Tantric Yogi. The affinity between cybernetics and Eastern mystical religion, especially Buddhism, is widely recognised. It is enough to note that one of the best popular introductions to recent scientific work on self-organising systems is The Web of Life, written by New Age guru Fritjof Capra, also the author of the famous Tao of Physics (Capra, 1975, 1996). But these connections are usually made in a rather generic fashion. Capra’s argument, for example, is that humanity is entangled in a complex system of relations with all of the plants, animals and minerals that comprise the planet Earth, and because of those entanglements it is in our interest to respect and care for non-humans and humans alike, much as the Buddha encouraged us to do. In Beer’s case, however, the connections between the spiritual and technical aspects of his work are much more specific and concrete than that. I do not feel that I fully understand any of these yet, but let me just mention three of them. (1) The icosahedron. At a mundane level, it would appear that Beer could have chosen any of the regular polyhedra as the basis for his

syntegration approach to collective decision-making. It is true that the icosahedron accommodated a relatively large number of participants relative to other polyhedra, but what made this figure particularly attractive to Beer and others was a species of number-mysticism. Especially, under a certain geometrical projection, the icosahedron gave rise to a figure known as an enneagram, whose mystical significance Beer traced back to Sufism and to the Vedas. During the Chile project, a Buddhist monk gave Beer a mandala which turned out to include an enneagram, and which Beer used in his meditational practices thereafter (Beer, 1994b, Chapter 12) (Figure 11) [16]. (2) The VSM. An important aspect of the VSM which I have not mentioned earlier is that it supposed viable systems to be recursive. Each component of System one of any viable system was supposed to be itself a viable system. Thus, under higher magnification, each System one in Figure 7 was supposed to consist of its own five element system, and so on, both up and down the scale. Since the body has mind and consciousness, this implied, for Beer, that different levels of consciousness could be traced down to the individual cells of the body, and upwards beyond the body, to a kind of group consciousness that arose in syntegration and eventually to the cosmos itself (Figure 12). (3) The VSM again. While this originated as a map of the physiology of the human nervous system, Beer also regarded it as a map of the mystical,

The science of the unknowable

517

Figure 11. The enneagram

K 33,3/4

518

Figure 12. Recursive layers of consciousness

spiritual body. The yogic chakras could themselves thus, be mapped onto the elements of the VSM and ascribed their own consciousness along the lines just mentioned, a consciousness to which, as Beer (1994b, p. 247) put it, “I attest from yogic experience myself”. The previous mapping would then connect us to the divine at the cosmic scale – to divinity as the ultimate unknowable with which its constituent elements, including the human, endeavour to cope and live in the presence of. Again, what interests me most here is that Beer did not attempt to separate his spirituality from his technical work, along the lines of the Modern settlement as Latour (1993) calls it, where the spiritual and the scientific, say, are thought to exist in separate realms. Nor were the two realms connected by a sort of parallelism. Instead, like the technical and political, they were fused together. The icosahedron was at once a map for organising mundane social relations and a meditational device. The VSM was a map of finite human organisations and, at the same time, of the spiritual order of the cosmos. It is worth noting that what emerges here is a very “earthy” view of the spiritual as continuous with the secular, a view in which, presumably, all the elements of the spiritual world themselves become open-ended in time in relation to one another. This is a very different theology from the one I was taught at school, with the Christian God as eternal, unchanging and quite apart from His earthly creation. We have travelled along way from the Templeborough steel mill, via biological computers, the organisation as a performative brain and subpolitical diagrams of democracy, to arrive at the yogic chakras and cosmic consciousness. What interests me so much about Beer’s work – and British cybernetics in general – is the distinctive character of its interventions in so

many fields that we usually think of as disjoint: informatics, management, computing, politics and spirituality (and this list goes on). I am also struck by the unity of these interventions, which can all be seen as working through of the ontology of unknowability or becoming in a way that breaks down modern disciplinary distinctions. We can perhaps find some inspiration here for our own work. Notes 1. An earlier and shorter version of this paper was presented at the Second Conference on the History and Heritage of Scientific and Technical Information Systems, Chemical Heritage Foundation, Philadelphia, 16-17 November 2002, and is to appear in Mary Ellen Bowden and Warden Boyd Rayward (Eds), Proceedings of the 2002 Conference on the History and Heritage of Scientific Information Systems (American Society for Scientific Information/Chemical Heritage Foundation). The present paper is a revision of a talk presented at the Department of Information and Media Studies, University of Aarhus, Denmark, 10 April 2003. I am grateful for comments received at both meetings, and also for suggestions from Raul Espejo on the final draft. Other studies deriving from this project include Pickering (2002, 2003a, b, c, forthcoming a, b). During 2002-2003, the project was supported by the National Science Foundation under Grant No. SES-0094504, and I was based at the Science Studies Unit, Edinburgh University. I thank the Unit Director, David Bloor, and all its members for their hospitality. 2. Beer provided me with a considerable amount of information before his death. I thank Beer’s partner, Allenna Leonard, and his daughter, Vanilla Beer, for their assistance and encouragement in my research. I also thank Eden Miller for enlightening discussions of Beer’s work in Chile and for allowing me to read and cite her unpublished work. 3. See, for example, Beer (1959) – his first book. 4. On the affinity between the cybernetic ontology and that at which I arrived in my analyses of scientific practice (Pickering, 1995, 2002). My particular interest in the history of cybernetics is to see how this ontology engaged in a wide variety of real-world projects. I emphasise these ontological considerations because it seems to me that an ontology of representation and knowability exercises a certain hegemony over our imaginations. We (academics especially) often find it hard to imagine the world as a place of unknowability and becoming, in which, as it were, actions speak louder than words. My ambition in examining the history of cybernetics is to help revive our ontological imaginations (my own as much as anyone else’s) – to help us see the world differently and to exemplify the sorts of practical projects that might suggest. 5. The page citation here, as below, refers to the more accessible 1994 reprint of this paper. 6. For more on Ashby, see Asaro (1998) and Pickering (2003a). 7. For further discussion of the cybernetic conception of “control”, see Pickering (2003b). 8. In the Hitch-hiker’s Guide, the earth is a giant analogue computer built by mice-like beings to answer the Ultimate Question. On the earth as an analogue computer, see Blohm et al. (1986). 9. One could develop this point further in a discussion of hylozoism, the idea that matter is active and thus that we should enrol it in our projects rather than bending it to our will. A key text here would be Pebbles to Computers (Blohm et al., 1986). 10. This book was significantly extended in its second edition (Beer, 1981) and eventually formed part of a trilogy with Beer (1979, 1985). For more on the VSM, see Espejo and Harnden (1989).

The science of the unknowable

519

K 33,3/4

520

11. The second edition of Brain of the Firm (Beer, 1981) includes a long history and discussion of the Chile project. Miller (2002) is an important historical study of Cybernsyn. 12. The quotation is from an e-mail to the author, 3 April 2003. The Oxford English Dictionary defines “metanoia” as “the setting up [of] an immense new inward movement for obtaining the rule of life; a change of the inner man . . .” 13. E-mail from Vanilla Beer to the author, 3 April 2003. 14. “Stafford was fond of telling the story about Marx that had him saying Thank God I’m not a Marxist. He didn’t usually describe himself in this context, but Stafford had a great deal of admiration for Marx, especially his early writings on alienation. He wasn’t much of a fan of Das Capital mostly on the grounds of dull and repetitive” (e-mail from Allenna Leonard, 5 April 2003). 15. See also Beer (1993, pp. 37-42) on the World Syntegration project. 16. In this connection, it would be interesting to explore further the connections betwen Beer’s work and that of Buckminster Fuller (discussed in Beer, 1994b, passim). The same Buddhist monk reappears in Lilly (1972).

References Asaro, P. (1998), “Design for a mind: the mechanistic philosophy of W. Ross Ashby”, Draft, University of Illinois (unpublished). Ashby, W.R. (1948), “Design for a brain”, Electronic Engineering, Vol. 20, pp. 379-83. Ashby, W.R. (1952), Design for a Brain: The Origin of Adaptive Behaviour, 2nd ed. 1960, Wiley, New York, NY. Ashby, W.R. (1956), An Introduction to Cybernetics, Wiley, New York, NY. Beer, S. (1959), Cybernetics and Management, English Universities Press, London. Beer, S. (1962a), “Towards the automatic factory”, in von Foerster, H. and Zopf, G. (Eds), Principles of Self-Organization: Transactions of the University of Illinois Symposium on Self-Organization, 8-9 June 1961, Robert Allerton Park [sic: actually 1960], Pergamon, New York, NY, pp. 25-89. Reprinted in Beer, How Many Grapes Went into the Wine? Stafford Beer on the Art and Science of Holistic Management, Wiley, New York, NY, 1994, pp. 163-225. Beer, S. (1962b), “A progress note on research into a cybernetic analogue of fabric”, Artorga, Communication 40, April 1962. Reprinted in Beer, How Many Grapes Went into the Wine? Stafford Beer on the Art and Science of Holistic Management, Wiley, New York, NY, pp. 24-32. Beer, S. (1972), Brain of the Firm, Penguin, London. Beer, S. (1979), The Heart of the Enterprise, Wiley, New York, NY. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, New York, NY. Beer, S. (1985), Diagnosing the System for Organizations, Wiley, New York, NY. Beer, S. (1989a), “The viable system model: its provenance, development, methodology and pathology”, in Espejo, R. and Harnden, R. (Eds), The Viable System Model: Interpretations and Applications of Stafford Beer’s VSM, Wiley, New York, NY, pp. 11-37, Reprinted from Journal of the Operational Research Society, (1984) Vol. 35, pp. 7-26. Beer, S. (1989b), “On suicidal rabbits: a relativity of systems”, Systems Practice, Vol. 3, pp. 115-24. Beer, S. (1993), “World in torment: a time whose idea must come”, Kybernetes, Vol. 22, pp. 15-43. Beer, S. (1994b), Beyond Dispute: The Invention of Team Syntegrity, Wiley, New York, NY.

Beer, S. (2001), “What is cybernetics?”, Acceptance Speech for an Honorary Degree at the University of Valladolid, Mexico, October 2001 (See Kybernetes, Vol. 31 No. 2, 2002, pp. 209-19. Blohm, H., Beer, S. and Suzuki, D. (1986), Pebbles to Computers: The Thread, Oxford University Press, Toronto. Capra, F. (1975), The Tao of Physics, Wildwood House. Capra, F. (1996), The Web of Life: A New Scientific Understanding of Living Systems, Anchor Books, New York, NY. Espejo, R. and Harnden, R. (Eds) (1989), The Viable System Model: Interpretations, and Applications of Stafford Beer’s VSM, Wiley, New York, NY. Latour, B. (1993), We Have Never Been Modern, Harvard University Press, Cambridge, MA. Lilly, J. (1972), The Center of the Cyclone: An Autobiography of Inner Space, Julian Press, New York, NY. Miller, E. (2002), “Designing freedom, regulating a nation: socialist cybernetics in Allende’s chile”, Draft, MIT (unpublished). Pickering, A. (1995), The Mangle of Practice: Time, Agency, and Science, University of Chicago Press, Chicago. Pickering, A. (2002), “Cybernetics and the mangle: Ashby, Beer and Pask”, in Pestre, D. and Dahan, A. (Eds), Social Studies of Science, La Science des Anne´es 1950 (French translation) Vol. 32, Presses de l’EHESS, Paris, pp. 413-37. Pickering, A. (2003a), “The cybernetic Brain in Britain: Ross Ashby’s Design for a Brain”, Paper presented at the History and Philosophy of Science Colloquium, 12 March 2003, University of Leeds. Pickering, A. (2003b), “On Gordon Pask: cybernetics as art”, Paper presented at the Institute for Studies of Science, Technology and Innovation, 16 June 2003, University of Edinburgh. Pickering (2003c), “Cybernetics and madness”, Paper presented at the Wellcome Centre for the History of Medicine Colloquium, 27 May 2003, Glasgow University. Pickering, A. (forthcoming a), “The tortoise against modernity: Grey Walter, the brain, engineering and entertainment”, Experimental Cultures: Configurations between Science, Art, and Technology, 1830-1950, Max Planck Institute for the History of Science, Berlin, Preprint 213, pp. 109-22. To appear in a volume edited by Schmidgen, H. and Rheinberger, H-J. Pickering, A. (forthcoming b), “A gallery of monsters: cybernetics and self-organisation, 1940-1970”, in Stefano, Franchi and Gu¨ven, Guzeldere (Eds), Constructions of the Mind, MIT Press, Cambridge, MA. Times (2002), “Obituary: Stafford Beer”, 9 September 2002. Walter, W.G. (1953), The Living Brain, Duckworth, London. Further reading Beer, S. (1974), “Cybernetics of National Development”, The Zaheer Foundation Lecture, New Delhi, India, in Beer, How Many Grapes Went into the Wine? Stafford Beer on the Art and Science of Holistic Management, Wiley, New York, NY, pp. 316-43. Beer, S. (1994a) How Many Grapes Went into the Wine? Stafford Beer on the Art and Science of Holistic Management, in Harnden, R. and Leonard, A. (Eds), Wiley, New York, NY.

The science of the unknowable

521

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

522

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Sequences of failure in complex socio-technical systems Some implications of decision and control David Weir Ceram Sophia Antipolis, France Keywords Complexity theory, Sociotechnical theory, Cybernetics Abstract This paper reviews the framework proposed by Stafford Beer in his seminal work, Decision and Control, first published in 1966, and argues that these ideas are centrally relevant to a topic on which research interest has developed subsequently, the study of crises, catastrophes and disasters in complex socio-technical systems in high technology sectors. The problems are not discussed simply in terms of system parameters like variety, redundancy and complexity. Much empirical research supports the view that these systems typically operate in degraded mode. Such variables as hierarchical position, actors’ motivations and intentions are relevant to explain the ways in which communication systems typically operate to filter out messages from lower participants and to ignore the “soft signals” from small-scale and intermittent malfunctions.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 522-537 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523544

There is no question that Decision and Control is one of the most important books of its time and that Stafford Beer was one of those great, illuminating geniuses whose contributions to the development of management science has been enormously significant. But was his work, so much of its time, of lasting importance? Has the discipline moved on, leaving Stafford’s corpus of merely historical significance? On the other hand, I believe that the ideas which engaged Stafford should continue to inform our discourse, and that his insights still have powerful implications for our professional practice as scientists and managers. In a sense my interests have always been rather the inverse of those of Stafford Beer. He came from Operational Research with a determination to demonstrate, as he puts it in the opening words of Decision and Control how “real science penetrated the zone of managerial intuition” (Beer, 1966, p. 1). I first met him in 1970 when I came as a junior colleague to join a team that already included Stafford in his prime at the Manchester Business School. I arrived with a background in empirical research in criminology and sociology obsessed with the understanding that on the whole, things did not work, at least not in the way that they were supposed. Managers understood this truth perfectly. Later, this crude appreciation was parlayed for better digestion by the theoretically-minded into the more technically sounding formulation of “all systems operate in degraded mode” (Weir, 1996, p. 1). The original version of this paper was given at the Conference to honour the work of Stafford Beer held at the London School of Economics on 3 March 2003.

Finally, after 20 years, in Dumas’ terms, we ended up again as colleagues when as Dean of the Newcastle Business School, it was my good fortune to re-encounter Stafford and to persuade him to work as a Visiting Professor on a new generation of researchers and scholars trying their best in an unsupportive and unpromising intellectual environment. I was delighted to rediscover that in practice we agreed about just about everything that really mattered. This paper is not the place for long encomia, but it is impossible to pass on without recording my own tribute, insignificant amongst so many, but sincere and heartfelt to one whose personal and professional example was and will remain as a “shining city on a great hill”. As a person and a scholar, he was a great, shaggy inspiration; the sort of person you hope to encounter when you enter academic life and, of course, rarely do. The heart of Stafford’s work, as far as my interests are concerned, lies in some of the theses expounded in Decision and Control and in particular in chapter 15 which deals with “Controlling Enterprises”. For this audience it is unnecessary to recapitulate the whole argument, but I select certain passages to illustrate the argument, which is, as far as it goes in relation to the corpus of his work, reasonably representative and, I guess, un-contentious. Throughout the book, Beer develops the cybernetic model of organisation emphasising “though it is impossible to predict any event at all, the pattern of interaction is predictable” (Beer, 1966, p. 178). His focus on the system uses an abstract or as Weber put it an “ideal-typical” approach and he notes that we must refuse to be blinkered by the conventions of description that actually exist for “reality is only one of the constraints on the optimal design conditions” (Beer, 1966, p. 182). But he straddles a wide epistemological river when he notes in the same passage that it is important to consider a situation “in its totality”. This is a large claim with profound practical and epistemological implications. Beer recognises the need to distinguish the roles of the scientist and the practitioner when he writes two pages later that “the OR man is paid by the management to help it make decisions: he is not paid to write PhD theses” (Beer, 1966, p. 184). There is in this formulation and indeed in all practical management an inescapable tension between completeness and relevance: Beer’s instincts as a scientist led him to the former, and his responsibilities as a manager drove him to the latter. Nonetheless it is tension that always engaged him. These concerns are amplified by his treatment of a key concept in the analysis of systems; that of redundancy. We first encounter this in Decision and Control in Beer’s account of the working of the brain under the influence of alcohol as related by von Neumann. He writes “It is now clear that the brain manages to achieve its high level of redundancy by multiplexing its nodes and its channels according to a very complicated logic. (But). . . the brain is in fact imperfect, from the very beginning of life to the very end, and its machinery is moreover constantly falling into disrepair; so the trick by which it produces a

Sequences of failure

523

K 33,3/4

524

reliable output is important” (Beer, 1966, p. 198). It was inspiring to see Stafford in his last phase, after the incident that so nearly robbed him of his life, continuing to come to terms with the reduction in his own capability to function cognitively, and devising new methods and competences to repair damaged circuits and impaired networks. It was brave and also rewarding to have the opportunity to help in the process of patching, upgrading and performing, substituting collective endeavour for individual sparkle. We can define “redundancy” in a system as the existence of a multiplexing of nodes and channels above the level adequate to permit an optimal pathway to exist. In many contexts, it can be understood as offering backup or alternative routes to the route identified as the “best”. Redundancy is important because without it most systems that have a finite capability would cease to exist as systems in the event of a breakdown of a major system component. In passing Beer notes with approval Revans’ perception that most actual organisations tended to appear opaque because “management systems need to be, can be and should be designed. . . they are stochastic, not formal structures. . . (but). . . whoever saw an organisation chart that was indefinite, redundant, or even asymmetrical?” (Beer, 1966, p. 201). I was myself interested in organisational opacity as a creative tool used by senior management to create, as one of the respondents in my study of white-collar clerical workers put it, a “wall of darkness” to obscure the workings of organisational hierarchy and the absence of clear, visible, structures of promotion and progression (Weir, 1973, p. 169). Beer notes that “redundant networks are inefficient in terms of the energy they consume or of their cost”, but the gains to the system are valuable in terms of the amplification of information. But there is more redundancy than this. It is through the components that are technically surplus to the optimal system requirements at any specific point in time that the system evolves utilising the slack thus available to create, test and prove new methods for attaining objectives. So redundancy in complex systems that apparently permit inefficiency may also be the source of adaptation and learning. However, these subsequent system states are inevitably sometimes superior, sometimes less adequate, but always different and thus not typically described appropriately in working system documentation which always lags behind the introduction of a new set of behaviours in the field. There are thus likely to be behavioural discontinuities between what should happen and what does happen. Even though the system may have been improved in technical terms, the actual performance may be inferior, sometimes dangerously so, because of the behavioural gap induced by the fact that the system is not where it is expected to be or in the position in which there are well-understood modalities of behavioural reaction.

A relevant example, well-documented in the airplane disaster literature, is that of the modifications in the locking mechanism in the rear cargo doors of the Turkish Airlines DC10, which crashed at Ermenonville shortly after take-off from Paris Charles de Gaulle Airport. Failure of the locking bolts in the rear cargo door had initiated an explosive decompression of the fuselage. The fuselage might possibly have survived this structural insult, if the main electrical and hydraulic controls all been severed simultaneously. This failure illustrated a basic design failure in the neglect of redundancy in planning back-up to these primary control conduits. The baggage handler charged with locking the door did not understand the instructions and their significance for the safe operation of the system, partly, it was alleged, because of lack of adequate competence in English-language operating and safety instructions. This failure illustrated the neglect of communication aspects of the system control documentation. Therefore, the second state of the system may well not and typically does not map on to the operating expectations in the minds of the systems designers, partly because it is unanticipated but always because events change systems. Former Prime Minister, Harold Macmillan knew the political aspects of this, for when asked on his retirement by an interviewer what had been the most difficult problems he had had to deal with during his premiership, he replied gnomically “Events, dear boy,. . .events”. Discrepancies between our maps of reality and the changing terrains to which these maps putatively refer inhibit managements’ ability to respond to critical deviations from “normality”. What managers take to be “normal” is invested with moral presumptions and is conditioned usually and sometimes fatally by what they conceive they have a right to expect. Sometimes these discrepancies can be seen to be the result of simple overload at critical nodes in the system, undetected until a pre-critical set of events occurs and leaves operators unguided as to which of several equally valid alternatives to select. Thus, at one time operating procedures on the Great Northern Railway for an on-board train fire in a main-line section enjoined the hapless guard to be simultaneously at the rear of his train laying track alarms to deter traffic, at the front of his train alerting the engine fireman to the situation, contacting the signalman to stop facing traffic from entering the section, fighting the fire and rescuing passengers from the burning carriages. It is evident that Superman himself would have difficulty in being as prescribed by Railway Operating Procedures to be in two entirely different locations simultaneously. The purpose of these regulatory injunctions was not however to compel smooth operation under normal operating conditions, but to ensure that in any subsequent enquiry at least one lower-level participant could be shown to have been in default of his obligations and the resultant disaster to be subject to regulation by Guilt and Blame rather than by system improvement.

Sequences of failure

525

K 33,3/4

526

A classic case of these competing expectations not being clearly met within the system as it actually operated in real time was the sinking of the Channel Ro-Ro ferry “The Herald of Free Enterprise”. The conflict between operating expectations requiring minimal time for in-port turnaround, conditioned by the leanest of manning schedules on the one hand, and the lack of an appropriate safety procedure compelling supervision and positive confirmation of the closing of the bow doors before encountering open sea conditions on the other crystallized in an overloaded task schedule for the bosun during the critical time-period. The bosun was blamed, but the system was to blame. In some other situations, it is the “expert” components within the system as for example, the highly experienced scientists at Chernobyl who introduced un-programmed error through deliberately distorting operating constraints just to see what will happen. Beer quotes from the work of another neglected giant of the early days of management teaching and research in Britain, the late Reg Revans, who describes some management systems in terms of their “opacity”. But Revans’ point out that “opacity” has still more implications for the student of risky systems. Beer quotes him thus: “Revans found that 28% of shop-floor operatives believe their own foremen understand the operatives’ problems; of these same foremen 88% believe they understand the operatives’ problems. Of these foremen 43% believe their own works’ managers understand the foreman’s problems; of these works’ managers 94% believe that they understand the foremen’s problems. Of these works’ managers 59% believe their own general managers understand the problems of work’s managers; and it can be assumed, says Revans, all general managers believe they understand the work’s managers’ problems” (Beer, 1966, pp. 202-3). These expectations not merely do not mirror. They are inversely related and Beer judiciously concludes that “The disparity between 7% and 83% is sufficiently large to indicate that the observer’s vantage point conditions his results (an outcome familiar enough to scientists) and to imply criticism of communications within the enterprise” (Beer, 1966, p. 203). Both Beer and Revans and indeed anyone who served in the Armed Services during World War Two would have been familiar within the acronym “SNAFU”. This is a polite abbreviation for “Situation Normal; All F***ed Up”. It is the “normality” implied in the phrase that packs the punch. These considerations also lead to another implication, that communication patterns within imperatively co-ordinated associations may differ significantly from those within systems where system objectives may be defined as universally agreed. This may be because there are no generally agreed goals or no goal-directing or organising supra-systems exist. But in reality where the components of the system are individual actors capable of forming goals at variance with those of the senior levels this expectation is unacceptable. The fact that organisational participants do not behave in ways anticipated or

predicted by system-designers or senior managers does not mean that they do not hear the messages beamed at them. Like Linda Ronstadt, they march as social actors to the beat of “different drums”. In short, organisational hierarchies are distinct species of systems; what happens in management work is normally opaque to the lower participants affected. Beer recognizes that the “real-life” situations in enterprises with which most practicing managers have to be concerned present challenges of considerable perplexity and he expands his framework and language to cope with this. Thus, he concludes “the enterprise as such is something more than the totality of operations: it is a whole organism. . .(and) should be regarded as an eco-system. . . by its very nature a complex organic entity, interacting with an environment. . . because of its immensely high variety, it has to be a self-organising system to a large degree. . . to handle it at all, we have to understand it as indivisible. The purpose is to determine. . . which principles of cybernetics can be invoked to institute and maintain control of the enterprise” (Beer, 1966, p. 371). But while we must agree with Stafford’s comment that “the task is almost terrifying. . .”, we may take the issue with him when he continues, despairingly, “and one which has very largely been given up by practising managers” (Beer, 1966). This is altogether too negative and in fact misses the mark. Most managers do not give up, but they develop routines, some simple, others sophisticated, for coping with the organisationally-induced complexity to which they are wilfully subjected. For this is not merely because as Stafford puts it, “they rely mostly on the self-organising properties of sub-systems, which they seek to link together by a macro-structure of relative simplicity which they can understand” (Beer, 1966). While most complex systems exhibit the characteristics of self-organisation, this by no means answers all the questions: the real issues concern whether systems planned to achieve certain goals will manifest self-organising modalities around those behaviours likely to promote system goals or around some others. The fact is that most managers do not understand it to be their job to manage complexity by enumerating and considering all possible outcomes, applying some consistent heuristic and selecting the optimal solution that will operate consistently to produce superior outcomes. Their task is to get on with things, achieve viable outcomes within the expectations of their superiors and get to the end of the week more or less in one piece without any major crises supervening. And this understanding is moderated always by the knowledge that something will go wrong because on the whole things don’t work. (SNAFU. . . situation normal. . . !)

Sequences of failure

527

K 33,3/4

528

1. How management systems really work Thus, we are not merely dealing with a question concerning the differing perspectives with which the system is viewed though this is an important consideration. The way the system is and its natural states and potentials are not merely different between the perspectives of the consultant, be he/she operational researcher or social anthropologist, and the CEO with overall responsibility for a complex and legally definite system. They are radically distinct also for those who are implicated in the middle of the system and whose personal choices for action are constrained by organisational history and culture as well as by operational science. Managers inhabit specific worlds of culture and language and these are by no means co-terminous with the “system” as defined by operational researchers. Spender (1990) has demonstrated that the very definitions of what is implied when middle managers speak of “market, cost, margin” and like vary among industrial contexts because the very “recipes” in these industrial cultures differ from one sector to another. Baldamus (1966) had showed earlier that the very elements in industrial work which are perceived by the researchers as degrading, dehumanising and restricting maybe those which through their very “tedium” provide the “traction” essential to getting the work done at all. There is a current concern in organisational research with the fact that knowledge is highly contextualised and much of it indeed may in practice be tacit. Moreover, it may in principle be inaccessible to the outside observer or the hierarchical superior, remaining implacably embedded within communities of practice or belief (Suchmann, 1998). But this does not mean only that the problem becomes one of what it is fashionable to describe as “knowledge management” as if once the right questions, if we knew what they would be, are asked, then the answers will honestly and quickly be forthcoming. There are strong reasons why some knowledge is inevitably tacit if not subterranean. Knowledge is power and power will not be shared unless objectives are also shared. Our models of organisational and enterprise “systems” need to be sophisticated enough to consider the factors of power, inequality, of knowledge wilfully misapplied or perverted by the very operations of the systems designed to promote them. Beer does not go quite so far as to imply that error is built in to the normal operation of apparently healthy systems nor would he have necessarily agreed with me that “all systems operate in degraded mode”. But he understood very well that mistakes happen, and that the consequences of a small irruption into the smooth operation of complex systems could provoke an amplified cycle of distress leading to system failure. At the time of writing Decision and Control, I believe that he was still optimistic about the potential for good systems

management, trained in cybernetic understanding, to get on top of these issues and bring them, as it were, back into the ambit of good professional management protocols. After Chile, his mood darkened, and he became more pessimistic that those in control of the levers of organisational command would opt for better system understanding as the path to improved control. If the normal experience of qualified and expert members of complex organisations is “things normally go wrong”, how does this come about and how may we deal with it? A first possibility, scouted by Beer, may be that the problems lie in the issue noted by Ashby that “homeostats consisting of a large number of inter-related sub-systems will not work unless they are very richly inter-connected” (Beer, 1966, p. 378). May the problems of system failure lie precisely in the additional difficulties of navigating the rich picture so that the greater availability of alternatives produces an increased liability to error? Are the sources of system failure more likely to be connected to the way in which human as distinct to other complex systems involve the operation of motivation, power, intention and action? Hierarchy and self-organisation are compatible for Beer, “providing the reason for doing this is borne continuously in mind” (Beer, 1966, p. 378). But while as Beer says “it is perfectly possible to super-impose a hierarchical organisation structure upon the self-organising system” it is not so clear that this will not change its operating conditions in unexpected ways and, perhaps, pervert it utterly.

2. Air crashes as failures of complex socio-technical systems Let us examine the case of a complex socio-technical system in which failure modes may be readily ascertained and causal sequences identified and on which comparative and time-series statistics in usable formats are available to see why this may be so. We use the term “socio-technical system” in the sense in which it is defined by Emery and Thorsrud (1976). The field we choose is that of catastrophic failures in complex systems involving the interaction of technology and human decision: to be more precise of catastrophic crashes involving big commercial passenger jet aircraft. There are a number of obvious reasons why these data are available and the topic clearly satisfies Beer’s criteria of urgency and importance: it is moreover a field in which information is available and knowledge has certainly progressed. Various agencies, the FAA and CAA among others, routinely analyse significant air crashes in terms of cause: these conclusions are normally based on comprehensive analysis by specialist investigators. A striking finding is that in a normal year between 80 and 85 per cent of big jet crashes involve “avoidable human error”: even more striking is that as a ratio this proportion has not varied significantly almost since records have been collected.

Sequences of failure

529

K 33,3/4

530

In other words, this is a ratio that is constant over several generations of operating technology. At one level, this could be interpreted as indicating that improving technology does not improve human behaviour: this can be parlayed into the folk adage that there is no technology so complex that human intervention cannot find a way to subvert it. While some of these events undoubtedly can be interpreted as the result of “human error” blaming the pilot who is often dead and thus in no position to offer any competing explanation is a convenient exculpation of other, more complex failures (Beaty, 1991). Pilots do make mistakes but so do controllers. Others see the prime field of research as constituting that of “human error” (Reason, 1990) or adopt a quasi-juridical framework that permits them to identify the blameworthy (Cook, 1989). But this search for error or guilt also requires an understanding of the way these systems typically operate to describe the circumstances under which errors of a certain type are likely to occur or those who can be subsequently identified as “guilty” to be able to commit their offences. It is generally accepted that there is a generic family of types of failure in complex socio-technical systems and these are well reviewed in a text written to accompany the Open University course on Complexity, Management and Change. (Bignall and Fortune, 1984). A considerable literature around these themes has built-up (Mars and Weir, 2000; Toft, 1992; Turner, 1978; Turner and Pidgeon, 1998; Weir, 1996). The most thoroughly articulated attempt at a general theoretical formulation is found in Perrow whose empirical data are taken from the operation of the nuclear power industry in the US and in particular from a review of the sequence of events that led to the accident at the Three Mile Island Unit 2 nuclear plant at Harrisburg Pennsylvania on 28 March 1979 (Perrow, 1984). Perrow notes that “interactive complexity” is a feature of many complex high-technology systems in industries like that of nuclear power and introduces the notion of “tight coupling” in which sequences of events, once started, lead apparently irreversibly to catastrophic outcomes even when only small or unimportant problems can be identified as the original “cause” of the chain-reaction. This tight coupling is a consequence of the very nature of the technology in a high-risk situation that prescribes well-defined, preset and invariant behaviours and tightly prescribed sequences of actions precisely because the risks of failure are objectively great. He concludes, “this interacting tendency is a characteristic of a system, not of a part or an operator” (Perrow, 1984, p. 4). Nonetheless the interactions prescribed in a tight-coupled system themselves become causes of system failure because once the first mistake occurs and the failure sequence is initiated, it is other behavioural characteristics of innovation, creativity, “thinking outside the box” and even simple-minded ingenuousness which may be required in the critical situation.

A relevant example comes from the Sioux City crash in which a fully laden commercial jet lost control of its hydraulic systems due to an unusual accident in mid-air. The pilots discovered that some control could be obtained by varying the pitch of the engines. Flying was uncomfortable, but progress was made and the plane did not crash immediately; but the pilots were dismayed by the technical team at the manufacturers to whom they referred that in their opinion the system was not one for which any experience data were available and that their best advice was to pray and to “try things”. They did so relatively successfully finally landing at well above the maximum safe speed only to be blown slightly off course just after the landing impact by an unexpected cross-wind. There was a crash and several people were killed. But the best system knowledge available had predicted catastrophic failure and total loss as inevitable. The ingenuity, craft skills and creativity of the on-board management team had improved on the scientists sum of knowledge. Perrow (1984, p. 5) notes that “for most of the systems we shall consider. . . neither better organization nor technological innovations appear to make them any less prone to system accidents. . . the odd term normal accident is meant to signal that, given the system characteristics, multiple and unexpected interactions of failures are inevitable”. Other analysts have concluded that the important thing is to stop the failure sequence from starting at all, anyhow, anywhere, and that therefore only more hard-wiring of the “correct” decision-pathways can provide the prophylactic against the initial error. They point to complex systems which do work regularly, routinely, consistently satisfying most system objectives, most of the time in the interests of most of their stakeholders and identify these as “high reliability systems”. Several of these high reliability systems are discovered in the military and include aircraft carriers and nuclear defence installations (Roberts et al., 1994). They argue that the two features differentiate the high reliability organisation. First, the goal of process reliability takes precedence over that of product reliability, so the outcomes are treated as of less importance than methods; doing it right is more valued than doing the right thing. Second, HROs are expected to perform at high tempo for sustained long periods of time and to do so repeatedly; as evidence of the latter they cite the air traffic control systems. Thus, they conclude that an organisational system meets these criteria, “if it performs some hazardous activity repeatedly without incident it is highly reliable”. High reliability systems are characterised as containing a culture which reinforces multiple communication channels: US aircraft carriers for example, design redundancy back into their operating systems with more than 20 communications devices to ensure instant communications to any part of the ship; the flight deck itself is linked to the control tower by five systems.

Sequences of failure

531

K 33,3/4

532

They reinforce a culture of safety by aggressively seeking to review and understand the causes of failure, by simulating critical and disaster situations in training and stressing a culture of safe operation, and supporting networked communication about how the effective performance of small tasks contributes to the safe operation of the whole system. These lessons are currently accepted in much of industry and in the UK for example, under the direction of the Health and Safety Executive, enterprises base their safety training on the creation of “Safety Cases” in which the vulnerabilities of the systems and potential weaknesses in preventative procedures are exhaustively reviewed and documented. On the day they are completed and for sometime thereafter these exercises undoubtedly do much to raise awareness of the ways in which failure can occur and the sequences in which small events can develop into large-scale crises and failures and it would be tempting to see these activities as somehow heralding that much-abused phrase “a change in the culture”. This, if true, would be a comforting conclusion. Unfortunately, as Stafford Beer would, if present, be the first to remind us it is likely not to be true at all. Perrow (1984, p. 11) points out that “we have produced designs so complicated that we cannot anticipate all the possible interactions of the inevitable failures; we add safety devices that are deceived or avoided or defeated by hidden paths in the systems”. Many of these systems are driven by the desires and ambitions of politicians and military leaders. To the natural inclination of these categories of decision-makers to presume that their objectives are superior to those of ordinary people, we add the complicating factor of the lack of competitive pressure that makes emulation possible and scrutiny more plausible. These are systems which are designed to be hidden; deliberately “opaque” in Revans’ term. Roberts et al. (1994, p. 158) argues that “the culture of high reliability demonstrates intriguing contradictions. These organizations are successful, yet some of their cultural norms encourage behaviours not commonly thought of as desirable”. There is a clear tension between the organisation’s insistence on the values of predictability and control and the orientations of the highly skilled practitioners within these systems for variety, self-expression, risk-taking and innovation, behaviours which challenge the overt values of the system. In the default case when there is a possibility of choice between these alternative values, a “risky shift” can occur and the high-risk alternative can be selected. A recent example concerns the death of some Canadian soldiers in Afghanistan under “friendly fire” from a US warplane operating under typical “high-reliability” constraints: the transcripts of the communications between the pilots and their military superior on the ground clearly show that they were aware that the ground forces might be “friendlies”, that the plane was never

under attack and that the decision to fire was taken without the appropriate safeguards in place. The point is not whether these events should happen or whether they happen frequently or rarely. The point is that they do happen. 3. Communication processes in the sequence of system failure Studies of big jet crashes as of other catastrophic events like the Flixborough explosion, the Bhopal gas discharge and the Piper Alpha fire, typically stress the vital significance of the communication process in the sequence of system failure and in particular, of the way in which “soft signals” can be ignored because of the failure of communication to proceed upward. In this context, it is necessary to be continually alert to the implications of Beer’s quotation from Revans. The “system” not merely appears different, but is differently conformed according to the hierarchical position of the perceiver. Downward communication falls on deaf ears while upward communications are filtered out because they contain messages that senior management finds unacceptable. In many cases, senior management compounds the error through reducing system variety by appointing lower participants in their own image. For those leading players in large organisations who appoint to a house model of appropriate competences and skills – “people like us” – end up by being told only what they know already to start with. But inevitably, perhaps, most senior people and technical specialists systematically under-rate the value of the knowledge offered on the basis of the experience of lower participants. Thus, in undervaluing these information sources because they may not be as strong as those available to the better qualified and more highly-paid is to miss the point. They are valuable because they are different; weak signals may give strong information if they emanate from new or unsuspected sources. In the case of Flixborough, the “safety system” could not operate as envisaged according to the organisation chart because a key role combining both technical and hierarchical power was unfilled at the time of the incident; in the case of Piper Alpha, vital information concerning a valve opened for routine maintenance was left open, but the information that this had happened was not positively passed to the next production team at shift handover. In most of these examples, there have been histories of near-failure and incorrect operation of which the details were known to lower participants, but not formally transmitted to those with the authority and technical competence to act on them. In the case of the “Challenger” space shuttle explosion there was an even more subtle twist in that while there had been an agreed history of near-failure of the crucial O-ring components, the significance of avoidance of disaster on these previous occasions was interpreted differently by the scientists and engineers on the one hand and the “senior line management” on the other. To the first group, the absence of experience data relating to the sub-zero

Sequences of failure

533

K 33,3/4

534

temperatures on the launch morning meant a negative indication for launch, but to the managers it meant that there was no negative experience so a launch could be justified. At a crucial phase in the late-night telephone caucus meeting, one of the senior engineers who had been very doubtful whether a launch could be justified was invited to “take off his engineer’s hat and put on his manager’s hat” ( Vaughan, 1996). So all information is indeed in the last resort management information even if it is information which is capable of being expressed in purely scientific terms. It will bear a twist or two, because managers are different from other system actors in that they have power. Moreover all decision-taking occurs under a-scientific conditions because in practice all systems operate in degraded mode. When we look at examples of overt deviance, disobedience, and criminal attacks on system elements we find a similar tale of system degradation; but surely, it may be argued, it ought to be possible to isolate the deviant elements and purge them from the system? Thus, industrial sabotage would seem to be self-evidently a system crime, the elimination of which must inevitably improve system conformance and thus efficiency. But again the evidence is not so clear. The classic study by Bensman and Gerver on semi-criminal perversion of production routines and craft standards in a wartime aircraft manufacturing plant indicates that because of the close inter-connection between craft skill and opportunity not merely to do wrong but also to put right relatively sophisticated assemblies it was normally the best workers who were involved in the “most deviant” activities (Bensman and Gerver, 1963). Likewise it is often experienced detectives with good arrest records who are likely to be implicated in police corruption enquiries: it is not possible in practice normally to separate out only those bits of the actual operating system which suit one’s value-preferences or those of the system-designers or of higher management. It is all of a piece, good with bad, and has to be accepted as such, as it really is, degraded but still, usually, under routine operating conditions operating well enough not to have to warrant exhaustive enquiry. But the secrecy involved in concealing the real operations of the system and the reduced effectiveness of the monitoring of the real data being produced by these operations amplify the original deviations from expected operating conditions in unknown and practice unchallengeable ways. This amplification produces new system states which are beyond the powers of observation of instruments designed to monitor expected states. This is not merely a question of a lack of requisite variety or of complexity per se: though many of the short-cuts which reduce the impact of complexity, introduced by experienced operatives are justified as “ways of making things work”. However, most of these informal routines will not typically be documented and incorporated in updated specifications because to do so would

admit the failure of some element of the original design which is normally “owned” by high-status or expert members of the system. Thus, while it is certainly possible to imagine an alternative outcome of the Challenger launch decision sequence involving a delay on the grounds that the operating conditions on that morning were sufficiently outside the experience envelope to justify caution, it is inconceivable from what we know about how these complex military-industrial high technology organisations are framed to expect a cancellation on the grounds that the system had been operating in a degraded mode all along and had in fact NEVER been safe. Yet, that is another truth. It was evident to most students of the Challenger disaster that the organisational system deficiencies which had led to this failure could predictably lead to another; these errors were reproduced in the Columbia catastrophe. These sequences of failure are now better understood by practitioners and regularly unearthed by objective researchers; their persistence indicates not merely the continuing relevance of Beer’s paradigm but the continuing challenge of understanding and controlling the behaviours of the human actors for whom, in his sanguine and optimistic moments, he had such high expectations. 4. Conclusion These considerations bear on another currently fashionable topic, Risk Management. While there is no doubt that the claims for this as a new discipline are considerably overblown, there is much insight to be gained from these approaches. But their potential for success is greater when expectations are reduced. Not all risks can be identified, measured, and quantified; but the greater risks accrue to the management teams which believe that a risk assessment, once successfully negotiated will remain relevant through changing internal and external conditions. . .”. “Events, dear boy. . . events” will derange the best-laid plans. But it may be argued that our present situation is certainly a positive improvement on the complacency that permeated the operating philosophies of the mega-bureaucracies in which Beer undertook the initial steps in the early days of Operational Research. But it is nonetheless advisable to conclude this review on a cautionary note. The problem with most of the complex systems in the socio-technical arena that researchers, consultants and practitioners have to deal with is that they are never in any meaningful sense tabulae rasae in terms of objectives, which is why power, intentionality and motivation remain important dimensions of their functioning. Organisations do not just exist; they are not “out there” as organisms evolving spontaneously from unknown life-forms; they have been created to serve the interests of groups or individuals who plan and intend to benefit from

Sequences of failure

535

K 33,3/4

536

their outcomes. The ends of research are usually planned to support actions that sustain the existence of one set of stakeholders while disadvantaging those of another. The risks of failure are normally greater for little people; the benefits for the powerful who see in complexity means to ensnare the ingenuous. So the final lesson to be gleaned from the rich harvest of ideas and analysis that comprises Decision and Control is the reminder that it is only through continual and persistent critical awareness that we not merely keep up with the evolving reality of complexity. This “organisational cynicism” is indeed the only prophylactic against what Stafford described as the villainy of the tenacity which “produces too slow an adaptation to cope with the revolutions that the world is undergoing in every sphere” for “new thinking everywhere is blocked by the method of tenacity (this idea has not yet been tried out, let someone else make the mistakes). Conversely, when new ideas about management and control have been discussed for a sufficiently long time, they too will be generally adopted, not by logic but by the method of tenacity. But this time, perhaps, it will be too late” (Beer, 1966, p. 21). Complex systems need to keep evolving and our critical awareness to keep pace with their mutations: otherwise as Stafford implied it will indeed be “too late”. References Baldamus, W. (1966), Efficiency and Effort, Tavistock, London. Beaty, D. (1991), The Naked Pilot, Methuen, London. Beer, S. (1966), Decision and Control, Wiley, London and New York, NY. Bensman, J. and Gerver, I. (1963), “Crime and punishment in the factory: the functions of deviancy in maintaining the social system”, American Sociological Review, Vol. 28. Bignall, V. and Fortune, J. (1984), Understanding Systems Failure, Manchester University Press, Manchester. Cook, J. (1989), An Accident Waiting to Happen, Unwin, London. Emery, F. and Thorsrud, E. (1976), Democracy at Work, Martinus Nijhoff, Leiden. Mars, G. and Weir, D.T.H. (2000), Risk Management, Vols. 1/2, Ashgate-Dartmouth, London and New York, NY. Perrow, C. (1984), Normal Accidents: Living with High-risk Technologies, Basic Books, New York, NY. Reason, J. (1990), Human Error, Cambridge University Press, Cambridge. Roberts, K.H., Rousseau, D.M. and La Porte, T. (1994), “The culture of high-reliability, quantitative and qualitative assessment aboard nuclear-powered aircraft carriers”, The Journal of High Technology Management Research, Vol. 5 No. 1. Spender, J-C. (1990), Industry Recipes, Blackwell, Oxford. Suchmann (1998), “Constituting shared workspaces”, in Engestrom, Y. and Middleton, D. (Eds), Cognition and Communication at Work, Cambridge University Press, Cambridge. Toft, B.A. (1992), “The failure of hindsight”, Disaster Prevention and Management.

Turner, B.A. (1978), Man-Made Disasters, Wykeham Press, London. Turner, B.A. and Pidgeon, N. (1998), Man-Made Disasters, 2nd ed. updated, Butterworth-Heinemann, London. Vaughan, D. (1996), The Challenger Launch Decision, University of Chicago Press, Chicago, IL. Weir, D.T.H. (1973), “The wall of darkness: subjective operationality and the promotion system among clerical workers”, Men and Work in Modern Britain, David Weir: Fontana, London, pp. 169-80. Weir, D.T.H. (1996), “The role of communications breakdown in plane crashes and business failure”, in Hood, C. and Jones, D.K.C. (Eds), Accident and Design, UCL Press, London. Further reading Lagadec, P. (n.d.), States of Emergency, Butterworth-Heinemann, London and Boston. Revans, R. (1982), The Origins and Growth of Action Learning, Chartwell-Bratt, Bromley. Smallman, C. and Weir, D.T.H. (1999), “Communication and cultural distortion during crises”, Disaster Prevention and Management, Vol. 8 No. 1. Toft, B. and Reynolds, S. (1994), Learning from Disasters: A Management Approach, Butterworth-Heinemann, London. Weir, D.T.H. (1993), “Communication factors in system failure or why big planes crash and big companies fail”, Disaster Prevention and Management, Vol. 2 No. 2. Weir, D.T.H. (2002), “When will they ever learn? The conditions for failure in publicly-funded high-technology projects: the R101 and Challenger disasters compared”, Disaster Prevention and Management, Vol. 11 No. 4.

Sequences of failure

537

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

538

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

The market economy unchecked Another version of fundamentalism? A polemical tribute to Stafford Beer Allenna Leonard The Complementary Set, Toronto, Ontario, Canada Keywords Cybernetics, Market economy, Social benefits, Control systems Abstract A case is made that the market economy imposes constraints and makes demands that mirror the rigidity of fundamentalist religions and that this makes answering one fundamentalism with another more likely. The denial of the world’s complexity, the dismissal of the suffering of those not among the select and the obliviousness to other perspectives that characterize fundamentalist economics or religion, create a dangerous situation for all. Systems models and tools, created by Stafford Beer and many others, provide practical means to address complexity and to achieve a fair balance of stakeholder needs and interests. These tools, and the holistic philosophy underpinning them, can and should be applied to questions of public importance. To do so, we have to follow the footsteps of Stafford Beer and other pioneers in the systems and cybernetics field.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 538-546 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523553

Argument I can think of no more fitting or challenging way of honouring Stafford Beer’s memory than to attempt to continue his practice of regarding the world from a holistic perspective and linking the observations to the underlying cybernetic concepts. Indeed, he often called upon us to address the problems of the world (Beer, 1993, 1997). The current environment of terrorism with its links to Islamic fundamentalism and the fear of terrorism with its focus on security would certainly have claimed his attention. He invited us to use his yo-yo model (Beer, 1966) and progress from a metaphor to an analogy and to continue as far as it seemed useful and predictive to go. I have been struck for sometime by analogies between fundamentalist religion of any sort and the extreme effects of capitalism. To a great extent, commerce and the free market economy seem to be the supreme arbiter of many questions and their dictates can be as rigid as those of any ayatollah. I find their demands irritating and distasteful as a “cradle capitalist” living under a liberal government. So, it is no surprise that some with access neither to the comforts of a consumer economy nor the rights of a citizen of an elected government might find these demands to be intolerable. This seems especially likely in those places where ruling elites flaunt their wealth while the majority of people live in poverty. This may seem to be a strange connection to make. We are accustomed to think of free markets as liberating, not constraining. But it is not always so when you read the fine print. We might ask: “What are the features of

fundamentalism?” I suggest that they are variations on a theme – the rejection of variety. The following are the five aspects of this rejection. (1) Non-negotiable world views – often anchored in the past or in an idealized present – a social version of the frictionless pulley. (2) Management of surrogate worlds that leave much human experience outside the frame. (3) Refusal to acknowledge the validity or legitimacy of other points of view, preferred goals or measurement criteria. (4) Literal interpretation of dogma with certain “facts” and assumptions closed to debate. (5) An us/them mentality. All of these may be seen in the constraints that can be imposed in the name of religion or the market. Think, for example, of the one-size-fits-all prescriptions of the International Monetary Fund. Although they work for some countries, they have made the situation worse in others. Or think of the shouts in the wilderness about the risks to the food supply posed by monoculture or the risks to the total economy posed by hackers when almost everyone depends on Microsoft whether they want to or not. One of the activities that Stafford and I enjoyed on our off hours was touring the cathedrals in Britain. Stafford’s visiting professorship at Durham Business School gave us a number of opportunities to explore Durham Cathedral and its 900 years of history. What impressed me most was the sheer reach of the Church. The Prince Bishops of Durham had their own army; minted their own coin; directed trade and industry; provided education; sponsored the arts; and determined social relations as well as pronouncing on the theological, moral and ethical questions of the day. The influence of the Church and its powerful bishops was felt everywhere and rivalled that of the king until the Reformation and the clash with Henry VIII led to the break with Rome and the establishment of the Church of England. At present, the same level of influence appears to be held by the market. It is so much a part of our environment that we often do not notice how inescapable it has become. Sometimes, the market’s economic considerations conflict with others such as the preservation of the natural environment or human rights and social justice. Small countries have to struggle to support their agriculture and industry to remain viable. Large countries are threatened when they try to protect their populations from perceived risks such as genetically modified foods. When this sort of thing occurs, we are seldom surprised when economics prevail. International trade agreements such as GATT and NAFTA protect commercial interests, but there is no comparable international protection for nature, human beings or their societies. Yes, we will protect the environment as long as it is not too costly; even better if we can develop new industries that will earn higher profits than the

The market economy unchecked 539

K 33,3/4

540

polluting ones. Yes, human rights are important, but not important enough to turn away from cheap oil and other resources and opportunities to market our products. Yes, it would be nice if the small banana producing countries in the Caribbean could earn a living but we must not allow exceptions to trade agreements that favour large plantations even if that tilts farmers toward growing illegal drugs to survive. Yes, it would be nice if ordinary people in countries run by despots had legitimate means to address their grievances and achieve their dreams. But it really is not our job to intervene. If that leaves an open field for extremists who are implacably opposed to modernism or opportunists who would use religion and its promise of rewards in the afterlife to motivate people to violence, that is too bad. The West has a poor record of supporting informed and progressive leaders in developing countries. In the Guardian of 1 March 2003, there were articles on two such men. Mohammed Harbi, exiled from Algeria to France, has received death threats from the Algerian Secret Police, the Radical Islamicists and some ex-settlers who have not forgiven him for his role in Algeria’s National Liberation Front. Ezatollah Sahabi of Iran became a dissident after the CIA replaced Iran’s President with the monarchy. He was imprisoned by the Shah and then again by the Islamic Republic after he criticized them for their repressive practices. With power held by dictators, beholden or not to the West, and no possibility for democratic dissent, the opposition that forms is likely to be among radical elements. With no alternatives and little to lose, it is easy to recruit among people who have little to lose. Stafford was fond of saying “the purpose of a system is what it does”. The system we have is perfectly designed to get the results we are getting. Although I would not deny the improvements in human life and the standard of living that have been achieved through the enabling of market economies (or religion for that matter) they simply do not have the requisite variety to address the full range of human concerns. To some extent, the great institutions of the market economy have realized this. Recent World Bank Reports (2002, 2003) sound the alarm about the bleak picture of development where many people are trapped behind political boundaries in desperate circumstances. They lack political and economic infrastructure as well as basics such as access to health care, clean air and water and education. If one compares some of the expected behaviours of powerful religions to those expected of participants in a market economy, one finds a number of similarities. Fundamentalist versions of religion are known for brooking no deviation from their central concepts and for taking many of their own precepts literally even in the presence of disconfirming evidence. Opposition to government regulation and opposition to labour union influence, for example, are articles of faith in extremist capitalism. Then there are the rituals. Some religions call people to devotions many times during the day. As consumers in a market economy, we too are

called – to attend to advertisements. In North America, TV commercials may run four or five times an hour for 4 min intervals. It is not possible to open a newspaper, and rarely a magazine, without going through the ads. Nor can one walk down a main street without seeing signs and billboards. There are ads on subways and before feature films. You cannot dry your hands in some ladies rooms without being presented with noisy commercial messages on a video screen. Similar screens are being piloted in the back of taxis – one of the few commercial spaces left where people enjoyed some privacy. And the Internet! Well, my ratio of Spam to message is about nine to one. Some people have told me I am lucky. How they communicate with children is another way to distinguish a fundamentalist approach. While modern religions teach children their precepts through examples that reflect their own experience, fundamentalist religions ask children to take on concepts and make promises that are far beyond their comprehension. In television advertising, they actually talk about “the two to eleven year old market”. At the younger end of this “market” children do not even know the meaning of a commercial transaction. Older children and teenagers do but they can be manipulated into making choices that are not in their best interests. Commercials for snack foods and sugar lay the groundwork for a lifetime of poor nutrition and compromised health. Nor do expensive branded toys and clothing necessarily promote good values or represent good value. We are also familiar with fundamentalist religions that regard people’s value primarily as adherents serving a religious purpose. One of the irksome aspects of commercialism is while we think we are customers because we have bought a ticket or subscribed to a cable service or a magazine, we are also a “product” and a more valuable one based on our age, gender, race or income. This has led to bizarre outcomes such as people from the wrong postal code having their magazine subscriptions discontinued or to popular television shows, such as Lawrence Welk in the United States, being cancelled because the viewers were too old for their commercial sponsors. Religions exalt saints and prophets and it may be blasphemous to joke about them. Post-modern commerce exalts brands. It is very difficult to make fun of them too – at least in any official capacity such as in print. Stafford once drew a cartoon of the moon with a Coca-Cola sign. It was not much of a stretch. We could watch the moon rising over the water from our hotel in Montevideo, but the view was dominated by a Coca-Cola sign covering most of the side of a large building. The lawyers would not let it be included in “How Many Grapes Went into the Wine” for fear that the publisher might be sued. Of course, places like London or Toronto have regulations preventing at least some of their scenic vistas from being used as commercial space. But that did not apply in Uruguay. Fundamentalist religions have strict rules about what you can and cannot talk about where they have jurisdiction. In the mall where I used to live, the Salvation Army wanted to continue their practice of soliciting contributions for

The market economy unchecked 541

K 33,3/4

542

the needy at Christmas time. This was a progressive community so it was permitted, although that was not the case in many malls. The condition was that they substitute real bells that ring for cardboard bells with ding and dong printed on them. Must not have any extraneous noise to distract the shoppers. Regulation of what you can say in a commercial space is far more restrictive than most governments would even think of imposing. Unfortunately, there are fewer and fewer public spaces where public speech is feasible. When I lived in the States and was active in local politics, this was a problem. We ended up holding signs and passing out leaflets at traffic lights like squeegee kids at election time, as television advertising was beyond the reach of most state and local candidates. Fundamentalist religions often have an “elect” with privileges that ordinary members do not share. In the market economy, some well known politicians have engaged in raising “cash for questions” or, even worse, sold their attention to those of their constituents who paid hundreds if not thousands to attend their receptions. Probably the worst example of this was when the Bush Administration, early in its term, invited the energy companies, including the since disgraced Enron, to formulate policy for the Department of Energy – and were not even ashamed to admit it. Such actions turn what is supposed to be public speech into private speech and communication with one’s tax supported public servants into a commodity. Fundamentalists of any stripe often place less value on those who are outside their own circles. The west seems to have little interest in resolving conflicts where there is no oil or other valuable natural resources at stake or in developing drugs, for tropical diseases, which we do not get in the west if there are few profits in sight for pharmaceutical companies. People in developing countries have little money to buy branded drugs. But, their herbal remedies have been patented and they have served as human subjects for the trials. It has taken an unconscionable amount of time to agree on steps to implement promises made at the Doha talks of the World Trade Organization to do something about this problem – and at this writing it is not certain how long it will be before the needed drugs are available. Are the poor and the conflict ridden the infidels of the market economy? Where do these observations leave us? One does not need to be a cybernetician to be aware of the situation although it does help to see the pattern. The first step is to acknowledge complexity and to be clear that any type of fundamentalism is a cruel trick because it denies the variety that is actually there. Of course, this is what makes fundamentalism attractive to some. They are overwhelmed by variety and happy to believe it when someone says “It’s all very simple, what’s good for us is good for everyone who counts.” or “Everything will be fine: just follow these rules and punish anyone who doesn’t obey.” But, we know from Ashby that only variety can destroy variety. The variety that is not dealt with may lead to disaster.

When vital communication channels are blocked or distorted even life or death information may not get through. The hierarchy of wealth and power frequently precludes technical people and workers on the ground from having authentic communication with executive and political decision-makers. The people at the top do not have the capability of understanding the engineers’ and the workers’ concerns because they do not have the experience or the knowledge base. Beck (1992) proposed that we begin to think of ourselves as a “risk society”. It is characteristic of such a society that it is almost as much a matter of distribution of “bads” as of “goods”. As Stafford was fond of reminding us, the president of an organization employing hundreds and thousands of people does not have a larger brain than anyone else. That person has no means of making good decisions at the top without well-designed filters at each lower recursion that amplify and attenuate the flow of information. The consequence when this does not take place is that undesirable things can happen and no one is accountable. Well-documented cases of this include the Bhopal, the Challenger and sadly, now, Colombia, where workers’ e-mails requesting the ground photographs taken of its launch were traced back – to investigate why the proper channels were not followed – rather than to address the merits of the concern (Globe and Mail, 27 August 2003). The market has been widely accepted to be inadequate in figuring and accepting the full costs of its actions. Indeed, it has been approved practice to upload the benefits while downloading the costs. We are most familiar with this phenomenon in the environmental field but it also has impacts on questions of human rights and social justice. It is interesting that the market drives behaviour in this direction even when individual business people might prefer other alternatives. For example, it is very difficult to resist saving money on maintenance or accepting a new practice, such as adding unwanted animal products to cattle feed, when to hold back might make one’s costs higher than the competition. Or perhaps a practice becomes standard and changes slip without much notice or examination. What has changed is the perception that the great and good (and this includes scientists) can be depended upon to consider the full range of factors that could result in harm. The notion that experts can prescribe for others, when it is the others and not themselves who will suffer the brunt of the consequences, is becoming discredited. The hierarchy, whether religious or commercial, can be and has been wrong. Neither group is free of individuals whose primary aim is the accumulation of power and benefits for themselves. Ordinary people have become, for good reasons such as power outages and mad cow disease, suspicious of unregulated market practices. Too many of them have suffered – from mild inconvenience to lethal harm. Beck puts the case that neither government nor business can be counted on to address enough of the variety in complex situations to be, or seen to be, legitimate

The market economy unchecked 543

K 33,3/4

544

decision-makers for the whole. He calls for a number of new and ad hoc organizations in the realm of sub-politics to serve as interfaces where all relevant viewpoints can be considered and integrated. Nor can any religion claim ethical leadership if it makes exclusive claims to the truth and dismisses any other position as heresy. Pulling together disparate elements with different distinctions is inherently a systems effort. There is complexity; there is variety to be absorbed; there are outcomes to be encouraged or discouraged; there are different stakeholders for whom the situation has to work; and, there are many observers who contribute different perspectives on any topic worth serious study. The standard tools of logic, modelling, probability and statistics are part of this, as are group process methodologies and other means of enquiry. But it will not happen until the contribution of systems thinking becomes public knowledge. It is imperative that people who know about systems become involved in politics and public debate and that they make common cause with those whose positions include tolerance of other viewpoints. It does not matter which party – or indeed if it is within a party structure at all. What is needed is a commitment to endorse the reality of complexity and insist that public processes can still be made to work. It is comparable to the arguments that were made thousands of years back in Athens and hundreds of years later in London, Paris and Boston for logic and reason to guide political decisions. Although these arguments still have not convinced everyone, the argument for systemic thinking takes public debate a step farther. Reductionist “either/or” thinking is fine for some questions, but most of the important issues are better addressed with a “both/and” approach. We do not necessarily have polar opposites, or, if we do, we might have several different opposites of the same pole. Kelly (1955) addressed this phenomenon in his personal construct theory. He observed that there was a great deal of difference in the way people reacted to the common circumstances of the Depression. From these observations, he deduced that people constructed their worlds from the patterns they reflected internally. He developed a tool, called the Repertory Grid, which asked people to list important elements in their lives (significant people, activities, beliefs, etc.) and to make structured comparisons among them. This tool is very effective to elicit and compare the perspectives of individuals, and to come to grips with the variety of perspectives. We need to recognize that the existence of multiple perspectives is as limiting and as enabling as Heisenberg’s uncertainty principle. We need to take McCulloch’s (1989) notions about the heterarchy of values at face value. If an organism with six neurons is too complicated to be predictable on any single scale of values, then we can hardly expect that a single scale will serve social humans with all their complex histories and considerations. (You might remember McCulloch’s rats. If offered a choice between sex and food, they chose sex. Given a choice between sex and avoiding electric shock, they chose to avoid the shock. When the alternatives were avoiding electric shock

or eating, they chose the food.) McCulloch pointed out that circularities of preference were not evidence of inconsistency but of a consistency of a higher order. There is no summum bonem that will arise naturally. If one exists it will have to be imposed by a massive reduction of variety. President Bush’s statement “You’re with us or you’re with the terrorists” denies variety and complexity and reduces choice to two unacceptable alternatives. Many people who find terrorism unacceptable resist this simplification. Interestingly, many people, ordinary and extraordinary, express similar resistance with respect to their unhappy countries. They do not like the government and the opposition either. In other places, people might say “I don’t like everything to be run by big business, but I don’t want everything to be run by religion or any other ism either.” We have got plenty of approaches, practical and theoretical, in the system’s field to enhance the appreciation of multiple perspectives and promote the implementation of democratic decision-making. Critical systems theory (Flood and Jackson, 1991) and socio-technical systems theory (Passmore, 1988) are useful bases for analysis. The viable systems model (Beer, 1979, 1981, 1985), with its sensitivity to incipient instability, is a powerful argument for autonomy and consideration of requisite variety at all levels. System Dynamics (Forrester, 1968) can show what is likely to happen if rates and levels continue as they are, or if they are changed. It is especially effective in mirroring the “no intervention” option. Ackoff’s (1981) Interactive Planning and strategies for achieving democratic organizations is another. On the participatory side, Stafford’s own invention, Team Syntegrity (Beer, 1994) is a very powerful process for enabling groups of people to address complexity but there are others. The publication “Participation Works” (New Economics Foundation, 1998) describes 21 group processes. There are also a number of technically supported formats that either provide for anonymous voting on propositions or for people to interact anonymously on discussion boards. Stafford set a good example of engagement in using cybernetics to try to make the world a better place and to benefit society. He was in a tradition that first entered the public eye with Norbert Wiener’s “Human Use of Human Beings” (Wiener, 1954). We could follow that tradition too – and maybe we will succeed in making some good things happen. Or to recall a topic in the very first Syntegration held at the Manchester Business School, we might at least be able to identify “states to avoid” and use our tools to look for alternatives. A good start has been made in the professional societies. The American Society for Cybernetics has for a number of years placed a great deal of emphasis on the cybernetics of “observing systems” and the International Society for Systems Science has been promoting the idea of global Agoras with multiple methodologies to enable democratic discussions. But these and other

The market economy unchecked 545

K 33,3/4

546

system societies are small and do not attract much public notice. It may be hoped that recent high profile events, where the complexity gap between the situation on the ground and the means being used to address it is particularly apparent, will get people thinking. In a problem-ridden world, we may be sure of one thing. Simple and simplistic solutions put forward by fundamentalists of the marketplace, of religion or of any other stripe are dangerous. This is so because they exclude so much variety from their pictures of the world. To succeed, we need to build inclusiveness. Cybernetics and systems tell us that there is no objective reality, separate from the different observers who define it. Listening to the multiplicity of their voices, with all the tools at our disposal, is an important first step. References Ackoff, R.A. (1981), Creating the Corporate Future, Wiley, New York, NY. Beck, U. (1992), Risk Society: Towards a New Modernity, Sage Publications, London. Beer, S. (1966), Decision and Control, Wiley, Chichester and New York, NY. Beer, S. (1979), Heart of Enterprise, Wiley, Chichester and New York, NY. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester and New York, NY. Beer, S. (1985), Diagnosing the System for Organizations, Wiley, Chichester and New York, NY. Beer, S. (1993), “World in Torment: an idea whose time has come”, Kybernetes, Vol. 22 No. 6, pp. 15-43. Beer, S. (1994), Beyond Dispute: The Invention of Team Syntegrity, Wiley, Chichester and New York, NY. Beer, S. (1997), “The Culpabliss error: a calculus of ethics for a systemic world”, Systems Practice, Vol. 10 No. 4, pp. 365-80. Flood, R.L. and Jackson, M.C. (1991), Critical Systems Thinking: Directed Readings, Wiley, Chichester and New York, NY. Forrester, J.W. (1968), Principles of Systems, MIT Press, Cambridge, MA. Kelly, G. (1955), A Theory of Personal Constructs, Norton, New York, NY. McCulloch, W. (1989), “A heterarchy of values determined by the topology of nervous nets”, in McCulloch, R. (Ed.), Collected Works of Warren S. McCulloch, Intersystems Publications, Salinas, CA. New Economics Foundation (1998), Participation Works: 21 Techniques of Participation for the 21st century, New Economics Foundation, London. Passmore, W.A. (1988), Designing Effective Organizations: The Socio-technical Systems Perspective, Wiley, New York, NY. Wiener, N. (1954), Human Use of Human Beings, Doubleday Anchor, New York, NY. World Bank Report (2002, 2003), World Bank, Washington, DC.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Tigers at play: Stafford Beer’s poetry David Whittaker Charlbury, Oxfordshire, UK

Tigers at play

547

Keywords Cybernetics, Poetry, Literature Abstract Considers Stafford Beer as a poet. Discusses Beer’s interest in a wide variety of poetic forms. Gives some technical examples of how Beer employed Welsh and Sanskrit metre and rhythm to writing in English. Also includes a translation from the Spanish; a ballade and selections from his longest poem One Person Metagame. You liked the formal beat of ancient forms The discipline they call for in the brain Ballade for Warren

For all of his adult life, Stafford Beer wrote poetry (Beer, 1983). This was not some eccentric quirk, a way of passing the time like doing a crossword puzzle. Poetry was an essential channel for Stafford’s creativity; a means of expressing and exploring vital aspects of the human condition (love and death), which was not always possible in the management texts. Stafford was a latter day Renaissance man personified; a transdisciplinary thinker who moved with ease across the sciences, philosophy, religion and aesthetics. The dedication and epigraphs to each chapter in Decision and Control (Beer, 1966) alone amply demonstrate the breadth and lively curiosity of his reading. Considering his many other commitments, the poetry is of outstanding quality. The French poet Paul Valery made a distinction between prose and poetry when he said that prose may be characterised as words marching and poetry as words dancing. Stafford was particularly interested in how to make his words dance. (Moreover, as a prose writer he had a distinct and enviable eloquence.) In poetry, he looked beyond the English poetic tradition to find ways of articulating more ambiguous sentiments. All languages have developed some kind of metrical rules to provide rhythmical structure and movement akin to a musical score. The rules involved in the many different forms of poetry particularly appealed to Stafford as a cybernetician. Confronted with an enormous variety of words and grammar these rules help to attenuate the almost infinite permutations of how to say anything. Paradoxically these constraints actually liberate creativity and prompt a more resourceful search for an appropriate pattern of words, offering an effective way of organising how to say something. Apart from the many classical forms of Western poetry (including a sonnet and a ballade for Warren McCulloch) Stafford found that Sanskrit and Welsh offered him a notable challenge in applying their principles to writing in

Kybernetes Vol. 33 No. 3/4, 2004 pp. 547-553 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523562

K 33,3/4

548

English. These two languages were important to Stafford. He had done his National Service, in 1946, in India and it was here that he was introduced to yoga and Vedantic philosophy; Sanskrit is inextricably yoked to these ideas. For this example, Stafford chose a classical metre known to Indian scholars by the evocative name of Tigers at Play (Beer, 1983), this is also the title he gave to the poem. It consists of four stanzas, each of which scans to one lengthy line of Sanskrit. The long stressed syllables are in bold. TIGERS AT PLAY Sea birds stand on a rocky wall evening is here. Farewell to sunshine today. Walk please walk with me now and come into the gloom where sea and wind counter play. Talk dies hard in the mouth in that resident’s lounge. They haven’t that much to say. Come please come with me now and walk over the sand out where the tide turns at bay.

Stafford said this metre demonstrated the startling effect that quantity can have on quality in poetry. As for the Welsh, Stafford lived for many years in an area of Wales, Ceredigion, where there is a good deal of national pride and he duly learned the language to feel more at home and show some solidarity with the locals. Like all Celtic peoples the medieval Welsh liked to talk and sing and their Bards maintain one of the great oral traditions. For this they developed a wide range of techniques including cynghanedd, which means “harmony”. This poetry was meant to be heard; the sounds of the words are crucial (Welsh being a strongly stressed language) and the pattern of sound can sometimes convey meaning and mood. Cynghanedd is a system of sound correspondences involving alliteration, accentuation and internal rhyme. It was a major influence on the English poet Gerard Manley Hopkins in the 19th century; he called them “chimes” from which his own formal experiments in sprung rhythm evolved. It is the most sophisticated method of sound patterning practised in any poetry of the world. Stafford sometimes applied these rules with rigor and sometimes they are distributed throughout a piece as a kind of guiding echo. In Behold a Cry (Beer, 1983), written in 1982, he deploys these rules with great care and skill, though this short paper allows only a cursory illustration of the elegant pattern involved. The items listed for further reading

(Brough, 1968; Conran, 1992; Preminger and Brogan, 1993) will provide the interested reader with a more detailed account. Two types of cynghanedd are on show here: sain (sound) and gytsain (consonantal). With sain the line falls into three sections: the first two sections are linked by terminal rhymes, while sections two and three are linked by alliteration of the rhyme words of two and the final stressed word of three; in gytsain the line is in two sections where the consonants in the first half must be repeated in the second half. A sound that follows the stressed vowel that ends the half-line does not count. The sain is in italics while gytsain is underlined. BEHOLD A CRY The soul ’s begging bowl, branded with its greed or singular need, renounces alms; rage collapses its ribcage: who can seek peace, escapes.... Now as the tyrant knows that truant and has mastery of his mysterious will to prize the epoch to oppress the picked out hostage with his two-edged sword, tremble: stem the stream of blood, the flood of flowing words. In the silence is more anguish than for dying: dumb agony could see that saints who deflected death’s arrow had afflicted this sorrow. Curb in still blood your carbon steel blade.

Even this crude and simplified schema on the page (remember it needs to be heard), and there are further subtleties expounded in the collection Transit (Beer, 1983), displays the complexity of the structures Stafford was prepared to pursue in his finely wrought craft as a poet. He often spent months on finding the right word. In addition to wrestling with the complexities of some of the forms listed above, he wrote in free verse. One Person Metagame (Beer, 1983) is a terrific tour de force and may well come to be seen as his overlooked masterpiece. It consists of 1,000 lines and took 4 years to write (1971-1974 – mainly in Chile). Ezra Pound’s Cantos were an acknowledged influence, but T.S. Eliot’s Waste Land is not so far removed either. This is an autobiographical epic and philosophical work divided into six games: immaturity, maturity, love, politics, knowledge and death. Along with his unpublished Chronicles of Wizard Prang and his Requiem (a large installation of non-figurative paintings), this is perhaps Stafford’s most revealing work. Rather than one strict form the poem ranges over various

Tigers at play

549

K 33,3/4

550

techniques in a playful manner. Indeed wordplay was an important ingredient in Stafford’s poetry (the first game opens with the pun “sun and air”) and in conjunction with the flexible capacity of metaphor (which is a kind of metalanguage, not so much transcending paradox as revelling in it) enriches his vocabulary throughout. There are also many allusions from his wide reading including Shakespeare, Nietzsche, Jung, Hesse, Sartre, Engels, Ashby, Spencer Brown, The Book of Job and The Bhagavad Gita (as a younger man Stafford translated part of the latter). A few quotations (selected with random precision) offered almost as aphorisms provide something of the unique flavour of the work; ideally it should be read straight through to fully appreciate the journey across the varied landscape of an abundant and busy life: ONE PERSON METAGAME Stir the cinders of flamboyant fires that burned life to this charcoal sun and air fresh wind on fresh life strip off the mask and still the mask is masked prison is the calculus itself everywhere old men rule it seems that some are almost thirty here was the knowledge lost from gnostics alchemists that continent distinction builds the universe in a plane space make a mark the world will follow from this whose science is a glass bead game then tumble in the hay with time enjoy her body years ago awake a century from now to kill a clock or two and ride back on the pendulum that stopped you started it the other way backwards and forwards once it swung who noticed just be honest now

that for this trip it swung instead forwards and backwards

Tigers at play

crispin crispian tock tick

Stafford translated poetry as well. This fine example is from the Spanish Cantares of Antonio Machado (1875-1939) and is in the unpublished collection On the Move. Stafford had a recording of the poem sung by Joan Manuel Serrat. Being aware that translation is a problem of transduction he said his own English rendering was “transmitted from the Spanish to preserve the rhymes and rhythm”: ANTONIO MACHADO’S CANTARES Everything passes, everything stays. Our lot is to pass – always to be making pathways in passing a road on the sea. I never chased after glory nor hoped that mankind’s story would immortalize my song. I love my worlds subtle: weightless worlds, gentle. . . soap-bubbles floating along. . . I like to see them dispersed in sun and scarlet: they fly under an azure sky – suddenly tremble and burst. I never chased after glory. . . Walker, your footprints are the road – nothing else. Walker, there is no road: the road is made as you walk. As you walk the way is opened. Glancing back, you’ll see the path you’ll never tread again. Walker there is no road only a wake on the sea. There was a time, just hereabout – forests now with brambles clothed – the voice of a poet was heard to shout:

551

K 33,3/4

552

walker there is no road: the road is made as you walk – knock after knock verse after verse. The poet went far from home to die; foreign dust is his abode. As he left they heard him cry: walker there is no road: the road is made as you walk – knock after knock verse after verse. When the troubadour can’t sing the poet lifts a pilgrim’s load, when prayer becomes a useless thing. . . walker there is no road: the road is made as you walk – knock after knock verse after verse knock after knock verse after verse knock after knock verse after verse

Stafford never lost an opportunity to enthuse about his foremost mentor, the neurocybernetician, Warren McCulloch. McCulloch, furthermore, was a blacksmith and a poet. In the role of the latter, he wrote in classical 16th and 17th century forms. Stafford deliberately chose one of these more difficult modes, the ballade, for his moving elegy to McCulloch. This is a poem he particularly enjoyed reading aloud; it has something of a prophetic tone and I personally now read it as a kind of epitaph for Stafford himself. BALLADE FOR WARREN You liked the formal beat of ancient forms The discipline they call for in the brain; You liked the flash of honesty that warms The protocol of each austere refrain. Because we made no parting, you and I, To all my friends I say this fond goodbye. Here’s one you left religiously performs The rituals of friendship. Is it vain To conjure you greathearted through the storms That lifted you above the common plane To spar with angels? And is it profane Because they inconsiderately die

To all my friends I say this fond goodbye? The lively young define the living norms By which to soil the earth and light the lane That leads into the sky. And their reforms In worlds outside our room may yet explain The marks we left upon the windowpane. Because I’ll leave them too without a cry To all my friends I say this fond goodbye. So much we know through pleasure, thought and pain Will perish in the future’s hurricane: Because there’s nothing to indemnify To all my friends I say this fond goodbye.

References Beer, S. (1966), Decision and Control, Wiley, London. Beer, S. (1983), Transit, Mitchell Communications, Canada. Brough, J. (1968), Poems from the Sanskrit, Penguin, Harmondsworth. Conran, T. (1992), Welsh Verse, Seren Books, Glamorgan. Preminger, A. and Brogan, T. (Eds) (1993), The New Princeton Encyclopedia of Poetry and Poetics, Princeton, New Jersey. Further reading Beer, S. (1993), “Requiem”, Kybernetes, Vol. 22 No. 6, pp. 105-8. Whiltaker, D. (2003), Stafford Beer – A Personal Memoir, Wavestone press, Charlbury, Oxon, U.K., (This publication is available on the author’s website: www.wavestonepress.co.uk).

Tigers at play

553

File name{KYB}Issues/33_3-4/Part 2.3d 09:52 – 1/4/04 – Total page(s) 3

PART II Illustration by Stafford Beer (II)

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

City planning “Dissolving” urban problems insights from an application of management cybernetics

City planning

557

Markus Schwaninger Institute of Management, University of St Gallen, Switzerland

Markus Koerner InterAktion, Pirmasens, Germany Keywords Cybernetics, Project management, Control systems, Urban areas Abstract Urban development in fast-growing cities is one of the huge challenges of our time, and several projects of technical cooperation are dedicated to this issue. The aim of this paper is to help project managers to enhance their capability of dealing effectively with the formidable complexities inherent in this kind of project. For this purpose, we explore the potential of Organizational Cybernetics and Social Systems Theory in a relatively new area of application. We have developed a set of conceptual tools that are helpful in coping with dynamic complexity in change and development projects. These tools have in common an inherent logic deriving to a great extent from Stafford Beer’s Viable System Model and the St Gall framework for systemic management. The application of the tools is illustrated by a state-of-the-art case study from the realm of Technical Co-operation – the revision of the Urban Master Plan for the City of Addis Ababa, the capital of Ethiopia. However, the toolkit is in principle also applicable to any complex project of change or development.

1. Introduction Project Management is the basic approach to Technical Co-operation (TC). TC sets out to assist developing countries mainly through projects, which are jointly defined by the “donor” and the “recipient” countries [1]. Up to the 1980s, the notion of “development” emphasised the transfer of knowledge or the implementation of advanced technology. At present, however, “development” is increasingly seen as an issue of managing social and institutional change and even the objectives of these intended change processes become evermore demanding. Instead of limiting itself to straightforward performance improvements, TC also aims at enhancing sustainability, organisational learning, equal opportunities, political participation, etc. Thus, TC projects have become much The theoretical foundations of this paper are Organizational Cybernetics, especially the work of Stafford Beer (1979, 1981, 1985), Social Systems Theory, Luhmann (1987, 1997, 2000) in particular, and the methodological work related to modeling and simulation with System Dynamics (e.g. Sterman, 2000), and finally the Management Framework developed at the University of St Gallen, which is one of the authors’ academic bases (Bleicher, 1999; Gomez and Probst, 1999; Schwaninger, 1989, 1994).

Kybernetes Vol. 33 No. 3/4, 2004 pp. 557-576 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523571

K 33,3/4

558

more complex, and coping with this complexity has become the main challenge for project managers and their advisors. For project managers, increased complexity means a growing uncertainty faced in decision-making on the project’s course of action and/or on the advice to be given to project partners and clients. A system is “complex” to the extent that it can assume different states, or exhibit a variety of patterns of behaviour. System complexity is determined on the one hand by the variety of elements and relationships within the system and in its environment [2] and in the system-environment interrelationships, and on the other hand by the degree of uncertainty and ambiguity (with regard to technologies, objectives, etc.) that limits or expands the range of options for decision-making ( Williams, 1999). Management is the key to coping with complexity (cf. Baecker, 1997; Schwaninger, 2000). However, classical project management approaches [3] are often of little use in institutional change management. Many TC project advisors already employ elements of a systemic management approach which contravene established procedures. In fact, we observe a growing disparity between project methods as standardised and prescribed in procedures and handbooks of TC, and the reality of project implementation. This disparity in turn leads to insecurity and even confusion in the management of these projects, which will inevitably impair their performance. In the following, we shall outline a parsimonious set of five basic and interrelated conceptual tools that capitalise on systemic principles of management. This toolkit enhances project managers’ and advisor’s capabilities to deal effectively with complex change and development projects. It also helps those concerned to adjust the instruments of TC to the new challenge of managing change. We are not making a proposition invented in some academic ivory tower. First, the proposed toolkit is based on theories which have been submitted to extensive empirical tests, as corroborated in the literature we refer to. Secondly, we ourselves have applied the whole set of conceptual tools in a pilot project. This paper provides an account of this pilot project in the form of a case study. There is no scarcity of definitions of a “project” (cf. Duncan, 1996; Wideman, 2001). For our purposes, we shall define a project as an undertaking which is in principle unique, usually to some extent innovative, and subject to a closing-date. Mostly, projects are also characterised by complex tasks and the participation of a multiplicity of actors (individuals, teams, enterprises, institutions, etc.). Also, projects are often identified by goals, start-up dates and deadlines [4]. We present each tool and then exemplify its use with the case-study, ours being a large urban project whose goal was to develop a City Development Strategy and Master Plan for Addis Ababa, ultimately leading to an advanced and foresightful Urban Management. This initiative was named the

“Master Plan Revision Project”. One of the authors (MK) assisted in this project as a resident advisor to the Addis Ababa City Government, seconded by the Deutsche Gesellschaft fu¨r Technische Zusammenarbeit (GTZ). The other (MS) participated in the project as an advisor on issues of systems methodology and organisational design.

City planning

2. First tool: primary processes and systems model Occidental thinking is founded upon bipolarity. Typical examples are the notions of “cause and effect” as well as “problem and solution”. These are also the root concepts on which current planning methods are usually based. With regard to concepts of causality, determinism has played a key role in the Western worldview (Luhmann, 1987; Wagner and Zipprian, 1985). Bipolar and deterministic approaches, however, often mislead us into “putting the cart before the horse” [5]. As an alternative to bipolarity, systemic thinking introduces circularity as a basic concept. This involves seeing projects and organisations from a viewpoint which is quite different from the traditional one. “Circularity” means that the output of a process is re-used as an input to that process (directly or indirectly) [6]. This creates a causal linkage, for which we then have two options: Either, more of one factor/variable also increases the other (for example: increased product quality increases staff self-confidence). Or, more of one factor/variable leads to a decrease in the other (for example: increased qualification of machine operators reduces accidents). Depending on the combination of these relationships in a causal network, we then obtain either self-enhancing (“reinforcing”) or self-attenuating (i.e. “balancing”) loops. Figure 1 shows these two different types of loops in a simplified form. Complex systems can in practice always be modelled on the basis of these two different types of loops [7]. Our central tenet is that the leaders of a change project should conceive of their project issue as a network of processes which are linked in loops. It is also especially helpful to identify a few basic processes which are central to the creation of value by the system. These basic value-creating processes are the engines/drivers of the entire system (Gomez and Probst, 1999). Such modelling, first, emphasises the dynamic character of project work, which traditional project planning perceives as rather problematical. Second, it

559

Figure 1. Two kinds of process loops

K 33,3/4

560

focuses the attention of project managers on those processes which are critical for the sustained operation of the system. The crucial tasks in developing a TC project are, then, to discover or construct – and then reinforce – circular and value-creating processes. More specifically, project managers need to: . Discover extant circular value-creating processes which are the bases of the system’s operation. . Construct new processes whereby virtuous self-reinforcing dynamics can be created; this may take the form of transforming an open causality chain into a loop. . Reinforce processes that embody virtuous modes of operation. . Correct harmful processes whereby problems and conflict have become self-sustaining, by analysing and breaking up such pathological self-reinforcement [8]. Often, this is achieved by introducing a balancing loop. . Link up processes to strengthen the overall performance of the system. We do not conceive of the relationships between components of basic processes as deterministic. In systems thinking, causality is not rejected but seen as probabilistic in nature [9]. The emphasis is on the indeterminacy of complex systems. In practical terms, this underlines the actors’ responsibility to conduct themselves responsibly. It also highlights the significance of responsiveness and flexibility, – a prerequisite to seizing opportunities. Viewing and modelling the issues of concern in this way help project planners and managers to: . focus their attention on the critical “drivers” of their projects, . ensure that sustainability is already “built into” a project from the outset and in a systematic – and in a sense also systemic – manner, . base their strategies on a view of their project which is comprehensive and which considers the importance of interfaces as well as the connectivity between the actors of the system-in-focus, and finally, . identify points of maximum leverage for interventions into the system. Example: Addis Ababa Addis Ababa has long been growing in a disorganised manner. At present, the city has 3 million inhabitants, expected to reach anywhere between 6 and 9 million within the next 10-15 years. Many people fear that such unconstrained growth could make it unmanageable. How can governance (leadership, policy-making, urban management) be enabled to initiate and foster a more gradual and benign evolution of the capital? To answer this question, a project was launched to develop a long-term strategy and “Master Plan” for the development of the city.

In line with the propositions mentioned earlier, the project team developed a general model of urban governance (Figure 2). The causal relationships depicted on the right-hand side of this model deal with politics, planning and finance. Urban policies influence the activities of urban actors through plans, programmes and regulation. These in turn are put into effect through the institutional set-up of the city administration. Good policies and effective implementation promote investment, which in turn enhances revenues from taxes and fees. Increased revenues, due to a bigger municipal budget, provide more options for policy-making. The left-hand side of the model is about governance and control. It shows that better policies and their efficient implementation also lead to more options for policy-making, through increased acceptance and support by the public [10]. In a highly abstract version, this model is made up of two basic loops (Figure 3).

City planning

561

Figure 2. Model of urban governance

Figure 3. Abstraction – two self-reinforcing loops

K 33,3/4

562

Representing the urban management of a city as a dynamic and circular system brings several important insights to the fore. . Both of these loops depict self-reinforcing circular causalities and, hence, an inherent potential for development. These loops reveal potential win-win situations for the stakeholders involved. . The qualitative aspect of development becomes manifest in a broader range of options (repertory of behaviours, opportunities to make impact, leeway) for stakeholders throughout the system. . The diagram emphasizes relational aspects. The focus of optimization is on both components and their interrelationships, often with emphasis on the latter. . The diagram shows that unidimensional measures are unlikely to have a significant impact on the performance of the system. It adverts rather to the need for balanced combinations of multidimensional measures, to which many actors can make their contributions. . Often, political constraints block the most direct and formally “correct” path to problem-solving. The model depicts already existent, self-sustaining interrelationships. While the cause-and-effect model often leads to frustration (because “necessary and sufficient preconditions” for intended results cannot be brought about), the concept of circular processes emphasises the multiplicity of opportunities available to enhance virtuous circles. These opportunities can be taken creatively and more at one’s leisure. In addition, leverage is provided through the inherent dynamics of the system. 3. Second tool: recursive structures Typically, a project management is concerned with three organisational levels as follows. (1) the overall organisation in which it is embedded; (2) the project for which it bears responsibility; and (3) its subprojects. Traditionally, these levels are mostly conceived of as a hierarchy, where the overall organisation controls the project and the project, in turn, controls its subprojects. In large and complex projects, however, one-dimensional hierarchical relationships become increasingly inefficient. The lower level depends on decisions taken by its controlling superior level, but the latter is often challenged to provide these decisions with sufficient speed and precision. As a result, the lower level tends to wait, i.e. delay its operation. Conversely, the upper level needs more and more and ever more detailed information from the lower level to prepare and justify decisions, but the lower level is, in its turn,

hard pressed to provide this information with sufficient speed and precision. As a result, the upper level tends to delay decision-making. Traditional organisational hierarchies are more often not caught up in such a “Catch 22” situation of decisions and information requirements – a self-sustaining, harmful causal loop. Systemic thinking offers the principle of recursion as an alternative: every primary unit (which is essentially a basic, self-sustaining and value-creating process with its own management) is considered to be a system “in its own right”, while at the same time it may be conceived of as a subsystem of a larger one. Hence, we regard projects as “wholes within wholes”, or systems within systems. In this way, control and organisational intelligence [11] are not concentrated at the “head” of the organisation. Rather, they are distributed throughout the primary units at all organisational levels. Each sub-unit is given an identity and the corresponding autonomy to be managed for sustainability in its own right, while at the same contributing to the sustainability of the greater whole. Recursive relationships combine control [12] with self-organisation in a complex pattern: control from “above” is accepted, but constrained by the principle of the autonomy of the lower level. Autonomy of units at the lower level is accepted, but constrained by the principle of contribution to the “overall” needs and goals of the higher level. In complex settings, recursive structures perform better, because the above-mentioned self-created impasse of the hierarchical model is avoided (cf. Espejo et al., 1996). In project management, recursive structuring leads to a more efficient use of resources and a reduced load of work and complexity on project managers. It is therefore important for the management of recursive project structures to look at the interfaces between the different levels’ “overall organisation – project – subprojects” (Figure 4). The relationship between overall organisation and project is usually set out in the formal “Terms of

City planning

563

Figure 4. Recursion model of a project organisation

K 33,3/4

564

Reference” of the project. These should be regarded primarily as a tool to manage this relationship, and not – as is usually the case in conventional project management – primarily as a framework for planning detailed activities of the project. The interfaces between the project and its subprojects can best be structured according to the principle of “Management by Objectives”. The leadership of subprojects should be given autonomy to manage its “own affairs”, while it is held responsible for contributing to the overall needs of the project. This responsibility is formalised in “Task Assignment”, which spells out the expected results and responsibilities of each subproject. What are the consequences for project management? (1) Ideally, units at each level should be conceived so as to: . have their own “identity” and structure; and . be able to manage, from start to finish, the processes for which they are the raison d’eˆtre . . . . (2) The project management’s attention should focus on the interfaces between the levels of recursion. It will neither try to interfere directly in the overall system, nor will it micro-manage the subprojects. (3) With regard to the interface-located “upstream” project, management first acknowledges the necessity for the project to contribute to the distinctive needs and objectives of the overall system. These goals are derived from the need to maintain the overall system’s primary processes intact. The management acknowledges its responsibility to structure the project as an autonomous unit within that system. These different requirements are usually not explicitly laid down in the written terms of reference attached to projects. However, interaction between the project and the representatives of its superior level of recursion is to a large extent a matter of how to balance the project’s contributions vis-a`-vis both the needs of the overall system and its own needs for autonomy. (4) Looking “downstream”, the project management will make use of management by objectives, leaving the internal control of the subsystems to their respective capabilities of self-organisation. However, tasks assigned to the subprojects must directly contribute to goal attainment at the level of the project as a whole. Example: Addis Ababa (see Figure 5) The three recursion levels in this case are as follows. (1) Overall system. The City of Addis Ababa and the City Government as its decision-making body. (2) Project. The Master Plan Revision Project, set up by the City Government. The MPRP project as an organisation has developed its own “corporate

City planning

565

Figure 5. Recursion model – Addis Ababa application

identity”, in separate offices, with its own staff, rules, goals, and budget. It has a high degree of autonomy and can, therefore, adapt optimally to the needs of its specific target groups and partner organisations. The project leaders manage the interface between the project and the city level through continuous negotiation with the city government on the interpretation of the MPRP’s “terms of reference”, i.e. its goals and the degree of the project’s autonomy. This leads to a dynamic alignment of goals and priorities between the two levels. (3) Subprojects. The project’s internal subdivision follows a “thematic” approach, disregarding sector or departmental boundaries. Rather, the project management establishes task-specific groups on a temporal basis to achieve comprehensive benefits with regard to real-life issues, as a direct output of the overall project. For example, the issue of “inner-city upgrading” is seen as just a “real-life” issue and is being handled by a group of professionals made up of planners, sociologists, architects, and economists. Since there is no internal division of labour to shield the task group from interacting with the environment and representing the whole project, each task group is confronted with the whole of the environmental complexity related to the issue it is working on. Interfaces between project and subprojects are embodied in task assignments which specify targets to be achieved. Control is essentially concerned with these targets, mostly disregarding the internal affairs of the groups (management by objectives). 4. Third tool: three-level model of management Additional leverage to deal with complexity is obtained by applying the well-proven three level model of the St Gall Management Concept ( Bleicher, 1999; Schwaninger, 1989, 1994). For a unit to achieve excellent performance, its

K 33,3/4

566

Figure 6. Three-level model of management

management must simultaneously meet criteria of success at three logical levels of management: (1) operative management – “efficiency”; (2) strategic management – “effectiveness”; and (3) normative management – “sustainability”. This model (Figure 6), which was originally designed for dealing with management at enterprise level, is equally valid for project management. Operative project management is about the project’s realization. It aims at excellence in execution, leading to high standards of value generated by the project “here and now”, in terms of economic, social and ecological benefits, but also in terms of the productivity (cost-benefit-ratio) achieved. Therefore, the overall criterion of operative project performance is efficiency. Strategic project management is about the orientation of the project in a larger time- and space-related horizon. Here, the focus is on building up value potentials, i.e. the prerequisites for generating value in the long run. The respective control parameters are critical success factors to be mastered (e.g. knowledge of the target groups’ and partner organizations’ fundamental problems) and core competencies (e.g. advisory and implementation capabilities paired with collaborative capabilities). In sum, the overall criterion of performance from the strategic perspective is effectiveness. Normative project management is about the founding principles of the project organisation. The purpose of normative management is to ensure the viability and development of the project. The principles of normative management embody the ethos of the project and are for the most part invariant over time. The ethos, the vision, as well as structural and cultural properties characterising a project organisation, indicate how viable that project is. The systemic view implies that the viability of a project can be maintained only if it is aligned with the greater whole in which it is embedded, i.e. as long as it creates a net benefit for that greater whole. Otherwise, the identity of the project must be further developed (adjusted). In sum, the overall criterion of performance from the view of normative management is sustainability.

Specific goals and orientators correspond to each of these logical levels. This allows for a focused management approach, but also brings to light sources of structural conflict, for example on the question as to whether resources should be allocated to short-term or medium-term goals. However, there is a logical pre-control relationship between these three levels, the superior levels creating preconditions and frameworks for the lower levels. This helps to establish priorities. Applying the three-level-model helps the project team to: . sort out complex issues of management, concentrating on the essential variables (mostly by applying the distinction between “operational” and “strategic” issues); . ensure that (normative) long-term issues are identified early on and pursued consistently; . resolve conflicts, which often arise from contradiction or competition between the different logical levels; . attain better overall consistency in its work, by analysing issues as operational/strategic/normative and then aligning different proposals on how to address these; mostly by taking recourse to the higher levels [13]. Example: Addis Ababa The project leaders shared experience gained from earlier cases: given the many aspects from which a project may be regarded, team discussions are often confusing, and decision-processes tend to be complicated, wearisome and time-consuming. In the end, the interests of the long-term are often sacrificed to a few short-term advantages, because these are more tangible and therefore easier to assess. In the MPRP project (Figure 7), the managerial issues are sorted out by distinguishing and balancing all three logical levels. In this way, the goals and therewith the criteria for assessing results, as well as the necessary measures, can be conceived more comprehensively. . Operative level. The project management ensures that the project continuously creates visible benefits – social, ecological and economic – for various target groups in the City of Addis Ababa. For example, the project team worked out a proposal on how to allocate scarce land for burial grounds between the different, heatedly competing religious communities. This task was initially not part of the terms of reference, but was taken on because it provided opportunity to demonstrate operational success. . Strategic level. The project management invests in human resource development and strengthening relationships between stakeholder groups (for example between the public sector and the business

City planning

567

K 33,3/4

568

Figure 7. Three-level model of management – MPRP project

.

community) by constantly developing the project’s resource base as well as space for manoeuvre in the political sphere (negotiation of the project’s mandate). Normative level. The project management aims at instilling a culture of result orientation, of effective decision-making and of transparency. It also cultivates an ethos of collaboration with stakeholders. At the same time, it ensures that the social values that guide the project are acceptable to the population, e.g. fairness, social responsibility for the poor, and the common good.

The principle of recursion and the logical three-level model are complementary to each other, so they can and should be combined. Figure 8 visualises that all three logical levels of management are functions inherent in each recursive whole, – project, super-project, subproject, etc. – provided that these are to be conceived as viable units. In other words, not only an operative management for efficiency but also strategic and normative management are functions to be distributed recursively across the whole system (cf. Beer, 1979, 1981, 1985). We shall return to this aspect in Section 5. 5. Fourth tool: process control model Traditional project planning and management approaches are static, in that they basically describe unwanted current and desired future states. However, they shed relatively little light on the processes required for the ( potential) change between these states. In practice, however, projects are driven by these processes rather than by the planning documents. As a result, traditional project management is often caught up in a dilemma – either to act according

City planning

569

Figure 8. Model of recursive management

to its plan, with foreseeable marginal or even negative results, or according to what is required by “the situation”, even if this is contrary to its plans. To make planning and management more flexible and adaptive, process orientation has emerged as a crucial concept. However, the concept appears to be somewhat elusive. What does process orientation mean in the context of managing complex change and development projects? The following model of process control is based on two of the tools hitherto presented, namely, the notion of primary, value-creating processes and the logical three level model [14]. It helps to structure the complex process of project management into a straightforward but limited set of modules/activities. Figure 9 focuses on the project itself, i.e. the middle level of our recursive set “system-project - subprojects”.

Figure 9. Process control model based on the viable system model

K 33,3/4

570

To put it in a nutshell, to manage a project as a process means designing, controlling and developing (only) four components and their interrelationships (cf. Figure 9): (1) The primary, operative, value-generating process. The activities which accomplish the purpose of the project (including their respective regulatory functions). (2) An operative control unit. The unit which regulates (via feedback) and steers (via feedforward) the ongoing process; this unit also deals with irritations (e.g. disturbances, difficulties, hindrances, disruptions) which originate from the project’s environment, and their prevention (see broken arrow) [15]. (3) A strategy, systems design and development function. This ensures the attainment of strategic management goals as defined above (tool 3), mainly through strategic controlling; this function also deals with the environmental changes: goals must be revised and the system design successively adapted. (4) A normative function. The ethos of the project – which is normally a project of both emergence and design – manifests itself in basic norms and lived-up-to values. It is insufficient to name or “fix” these at some point, since they need incessant reflection and discourse to improve them. Systemic process control, in contrast to traditional project management, does not require a comprehensive and detailed “operations plan”. This does not imply that plans can be dispensed with. In contrast the more uncertain the process is, the more important a good plan becomes. However, planning effort is focused on the critical variables/issues and hence, reduced to a reasonable and efficient level of detail. In this sense, planning no longer means constructing rigid tracks into the future. Monitoring becomes easier and more sensible, while planning itself turns out to be a vital instrument for continual adaptation and learning [16]. Example: Addis Ababa The management process is designed on the basis of the process control model. (1) Primary, value-generating process. The causal loops of “more legitimacy” and “more money” introduced in Section 2 are the focus of project management. Whether there is “more legitimacy” for urban governance is monitored inter alia by checking local news. Whether the virtuous circle of “more money” is enhanced can be observed by gauging the City’s income from taxes and fees. (2) Operative control is basically performed through assigning tasks to subprojects and through monitoring task achievement. Indicators for

task achievement are expressed taking both “hard” and “soft” data into account. (3) Strategy, systems design and development: the crucial instrument for these is the causal network model of city management already introduced. This model represents key elements of both project and its environment with a view to the mid to long-term. Based on this, the management team performs a periodic and “holistic” review of project status which includes its assessment in terms of goal achievement and the identification of key constraints. From this review, milestones are set, and their achievement is regularly checked. Process control at the normative level is performed through a steadily albeit slowly flowing stream of feedback from outsiders, and through the continuous display of presence by the project leaders, and an open debate on the project with their staff and clients. This leads to a continual adjustment of the project’s vision and mission as expressed by its management, the management styles displayed and the project’s long-term priorities. 6. Fifth tool: relationship model Finally, we have to address the relationship of clients and advisors. TC usually places international long-term advisors, who are provided through “projects”, with senior staff of the recipient’s country public offices. The advisors help the clients to do things they could not do before, or help them to improve upon what they are already doing, but “sustainable”, which means that they are able to do it without their advisors, and, as everybody knows, “help” has a tendency to disempower those who receive it. We therefore have a somehow paradoxical situation: first, to maintain the relationship in which clients are expected to do things which they cannot do (or maybe even: do not want to do). But officially, the clients are the sole beneficiaries of this relationship. Second, the advisors are expected to stay close to their clients and at the same time to keep their distance from them. In practice, it is difficult to maintain an optimal balance between remoteness and proximity between advisor and client. The optimum balance is achieved if the advisor is close enough to be able to provoke or unsettle (i.e. “irritate”) the client and give him food for thought, without getting so close as to become biased or unable to criticise him (cf. Ahlemeyer and Ko¨nigswieser, 1997; Su¨lzer and Zimmermann, 1996). In order to disentangle these complicated relationships, we again use a systemic approach in which we conceive of the client and his institution on the one hand, and the advisor and his “project” on the other, as two distinct, autonomous systems. Both systems are complex, self-sustaining through core value-creating processes, and operationally closed (cf. Luhmann, 1987, p. 55).

City planning

571

K 33,3/4

572

The relationships can then be separated into three types of causal reference as follows. (1) Self-reference of the partner organisation/target group (“client system”). (2) Self-reference of the project (“advisory system”). (3) Reference of the project to the partner organisation/target group. These three types of references can be located in the relationship model (Figure 10). The advisory system impacts on the client system through “interventions”. The client system, in its operational mode of self-reference, can either “accept” or “refuse” each intervention. Therefore, it is crucial that the interventions should be individually customised and monitored, and that the advisory systems learn from such cases of “acceptance” or “refusal”. The advisory system appears “alien” to the client system. This is a constraint. But it is also an advantage in that it is often easier to suspend taboos of the client systems. Furthermore, by providing multiple perspectives on current issues of the client system, the latter’s conceptual horizons are expanded, and innovation is more likely to occur [17]. The model cannot structurally dissolve the above-mentioned paradox of “help to self-help”. However, illustrating it helps the actors involved to deal with the inevitable insecurity and conflict which are thereby incurred. Thus, the model makes “help to self-help” a more operational and realistic concept. The model highlights the need to respect the autonomy and self-responsibility of both parties, which is a critical building-block for collaboration in a client-advisory setting. Example: Addis Ababa The client in this case is the Master Plan Revision Project of the City Government, whereas the advisory role is performed by the German technical co-operation, GTZ. Contrary to conventional practice, there is no joint plan of operation between the international project advisor and the client system. Rather, the GTZ team acts with a considerable degree of operational autonomy, which is, however, constrained by its obligation and commitment to contribute to strategic project goals which were agreed between the two governments involved. It assesses ongoing processes and tries to enhance or restrain them through various types

Figure 10. Relationship model

of interventions, which may include critical questions or even “unsolicited advice”. For example, if at an operational level everybody seems to agree that hawking and begging are punishable offences – an unrealistic approach and a form of denial – the external advisors may justifiably question this. GTZ’s role as an “outside advisor” has been instrumental in accessing information that – owing to a high level of mistrust within the administration – is as a rule not easily transferred between organizational units. Moreover, GTZ as an external partner can often convey bad news and criticism of high-level decisions more easily and in some cases more acceptably than a member of the client system. The relative “independence” which is granted to GTZ as an advisor also provides an option for staff of the City Government, who operate within the client system and otherwise would feel bound to adhere strictly to established standards and practice, to mobilise additional support for activities that might be considered too innovative or risky by the majority of city decision-makers [18]. Finally, GTZ is again and again confronted with a request to solve the City Government’s problems, i.e. to work out and implement a proposal for participatory planning procedures. However, the definition of the relationship as shown in the relationship model helps to clarify the self-responsibility of the client [19]. 7. Synopsis and outlook The aim of this paper was to help project managers to enhance their capability for dealing effectively with complex issues, with a focus on urban planning and development. For this purpose, we have explored the potential of Organizational Cybernetics and Social Systems Theory to support the management of complexity, from the specific perspective of project organization and leadership. To this end, we have designed a set of conceptual tools. Despite the parsimony of these tools, the fact that they are bound together by an inherent logic makes them powerful devices for dealing with the dynamic complexities confronted in change and development projects. We have tried to outline this cogent logic along the lines of our presentation, and illustrated each tool by reverting to the manifold aspects of one single, highly complex case. The purpose of the project-in-focus is a revision of the Urban Master Plan for the City of Addis Ababa, Ethiopia. A project of that stature cannot “solve” all the problems at hand. But it can achieve a joint creation of a “vision” – a coherent scenario of a desired and viable future, and a design to bring it about. If not all the problems can be “solved”, a process can be initiated, which has the potential to “dissolve” problems through the redesign of the system (Ackoff, 1999). Given the fearsome challenge of its endeavour, this Urban Master Plan project is certainly a good case in point. The toolkit presented here deals with the invariances inherent in any complex organizational project. Therefore, it is

City planning

573

K 33,3/4

574

probably valid for a great variety of change, development or innovation projects. The formidable complexity of the issues tackled in our pilot project makes this a good example of the new vistas for applications of Management Cybernetics and Social Systems Theory. To name just few relevant fields: . local planning, problems of megalopolis and regional development; . creation of viable economic and social systems; . care of resources, ecology and sustainability; . design for health care and education; and . issues of conflict and peace. These new prospects have only just started to emerge. There is plenty of work to do for Stafford Beer’s disciples. Notes 1. The following thoughts were developed in the context of a project implemented by the City Government of Addis Ababa with the support of GTZ, Deutsche Gesellschaft fu¨r Technische Zusammenarbeit. At any given time, more than 1,000 projects are being supported by GTZ, with the sponsorship of the German government. 2. When writing about the “environment” in this paper, we are referring to the specific environment here, which is relevant for a project under study. 3. See for example the Project Management Body of Knowledge (PMBoK) and the AMA Handbook of Project Management. 4. We differentiate the notion of “project” from that of “process” in that the latter – designating a “business process” – in business administration is usually applied to a mode of structuring routine operations. From a systemic perspective, however, project management always implies process management. We also subscribe to the “fractal” perspective, which understands a “programme” to be a project at a higher level, which is made up of a bundle of projects. Accordingly, projects can also be divided into subprojects. 5. A critique of causal models was delivered, e.g. by Musto (1987). See also Su¨lzer and Zimmermann (1996): 306 ff. 6. Kim (1992) and Richmond (1998). 7. To achieve stable development, a complex system must always be made up of both kinds of loops. A system with self-reinforcing loops only would necessarily move away from equilibrium uninhibitedly, and sooner or later drift into a catastrophe (Kim and Anderson, 1998; Senge, 1992). 8. See, for example, the systems archetypes proposed by Senge (1992) and Wolstenholme (2003). For applications, see Schwaninger (2003). 9. Grint (1997), Levy (1994), Stacey (1993), Sterman (2000) and Treadwell (1995). 10. This qualitative model concentrates on two self-reinforcing loops, which can drive the evolutionary process of the city. In other cases, both reinforcing and balancing cycles may be necessary for an appropriate mapping, the latter for example in relation to the reduction of pollution. It may also be indicated to elaborate more detailed models and to formalize important issues in quantitative models. In other words, the simplified qualitative model exhibited here cannot represent or simulate the dynamics of the system under study with its

11. 12.

13.

14. 15.

16.

17. 18.

19.

often counterintuitive patterns of behaviour and unexpected side-effects. This would require a complete simulation model (Schwaninger, 1997). For the concept of, “Organizational Intelligence” as used here, refer Schwaninger (2000, 2001). “Control” has two components: “regulation” (based on feedback), and “steering” (based on feed forward). In the English terminology, the adjectives “controlling” and “regulatory” are often used synonymously. The management based on this three-level framework can be supported effectively by dynamic simulation models, e.g. System Dynamics models (Schwaninger, 1997; Sterman, 2000), as well as evolutionary or agent-based models (Allen, 1997). This model of process control is based on the viable system model (Beer, 1979, 1981, 1985). A further function represented by the two punctuated arrows in Figure 9 is the provision of coordinating, self-organising and attenuating functions, as well as an auditing function. For a detailed discussion, see Beer (1979, 1981, 1985). This also implies: planning serves primarily as a discourse of project members for the sake of a joint invention of a desirable future and how to bring it about, and less as an anticipation of “the”future (Ackoff, 1981; GTZ, 1997). This concept revolves around what in German is termed “beraterische Distanz”. In early 2000, an exhibition about the Masterplan Revision Plan was held. It was visited by 50,000 inhabitants, and about 1,200 participated in hearings on city planning in Addis Ababa. Both opportunities were used, according to a broad survey among citizens, to inform themselves about their ideas and needs. By the end of 2001 over 120 workshops had been held for that purpose. “Self-responsibility” refers to the autonomous handling of problems. This does not exclude recognition of the interdependencies with other units.

References Ackoff, R.L. (1981), Creating the Corporate Future, Wiley, New York, NY. Ackoff, R.L. (1999), Ackoff ’s Best. His Classical Writings on Management, Wiley, New York, NY. Ahlemeyer, H.W. and Ko¨nigswieser, R. (Eds) (1997), Komplexita¨t managen – Strategien, Konzepte und Fallbeispiele, Gabler, Wiesbaden. Allen, P. (1997), Cities and Regions as Self-organizing Systems, Gordon and Breach, Amsterdam. Baecker, D. (1997), “Einfache Komplexita¨t”, in Ahlemeyer, H.W. and Ko¨nigswieser, R. (Eds), Komplexita¨t managen – Strategien, Konzepte und Fallbeispiele, Gabler, Wiesbaden. Beer, S. (1979), The Heart of Enterprise, Wiley, Chichester. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester. Beer, S. (1985), Diagnosing the System for Organisations, Wiley, Chichester. Bleicher, K. (1999), Das Konzept. Integriertes Management, 5th ed., Campus, Frankfurt/New York. Duncan, W.R. (1996), A Guide to the Project Management Body of Knowledge, Project Management Institute, Newtown Square, PA. Espejo, R., Schuhmann, W., Schwaninger, M. and Bilello, U. (1996), Organisational Transformation and Learning. A Cybernetic Approach to Management, Wiley, Chichester. Gomez, P. and Probst, G. (1999), Die Praxis des ganzheitlichen Problemlo¨sens, Haupt, Bern etc. Grint, K. (1997 ), Fuzzy Management – Contemporary Ideas and Practices at Work, Oxford University Press, Oxford. GTZ – Gesellschaft fu¨r Technische Zusammenarbeit (1997), Leitfaden zum Project Cycle Management, Gesellschaft fu¨r Technische Zusammenarbeit, Eschborn, Germany.

City planning

575

K 33,3/4

576

Kim, D.H. (1992), “Using fixes that fail’ to get off the problem-solving treadmill”, The Systems Thinker, Vol. 3 No. 7. Kim, D.H. and Anderson, V. (1998), Systems Archetype Basics: From Story to Structure, Pegasus Communications, Waltham, MA. Levy, D. (1994), “Chaos theory and strategy: theory, application and managerial implications”, Strategic Management Journal, Vol. 15, pp. 167-87. Luhmann, N. (1987), Soziale Systeme. Grundriss einer allgemeinen Theorie, 2nd ed., Suhrkamp, Frankfurt am Main (also published in English: Luhmann, N. (1995), Social Systems, Stanford University Press, Stanford, CA). Luhmann, N. (1997), Die Gesellschaft der Gesellschaft, Suhrkamp, Frankfurt am Main. Luhmann, N. (2000), Organisation und Entscheidung, Westdeutscher Verlag, Wiesbaden. Musto, S.A. (1987), “Die hilflose Hilfe: Ansa¨tze zu einer Kritik der manipulativen Vernunft”, in Schwefel, D. (Ed.), Soziale Wirkungen von Projekten in der Dritten Welt, Nomos, Baden-Baden. Richmond, B. (1998), “Closed-loop thinking”, The Systems Thinker, Vol. 9 No. 4. Schwaninger, M. (1989), Integrale Unternehmungsplanung, Campus Verlag, Frankfurt/New York. Schwaninger, M. (1994), Managementsysteme. St. Galler Management-Konzept, Campus Verlag, Frankfurt/New York, Vol. 4. Schwaninger, M. (1997), “Integrative systems methodology – heuristic for requisite variety”, International Transactions in Operations Research, Vol. 4 No. 4, pp. 109-23. Schwaninger, M. (2000), “Managing complexity – the path towards intelligent organisations”, Systemic Practice and Action Research, pp. 207-41. Schwaninger, M. (2001), “Intelligent organizations: an integrative framework”, Systems Research and Behavioral Science, Vol. 18 No. 2, pp. 137-58. Schwaninger, M. (2003), “Modeling with archetypes: an effective approach to dealing with complexity”, in Moreno-Dı´az, R. and Pichler, F. (Eds), Computer Aided Systems Theory – EUROCAST 2003, Springer, Berlin, pp. 127-38. Senge, P.M. (1992), The Fifth Discipline, Century, London. Stacey, R. (1993), “Strategy as order emerging from chaos”, Long Range Planning, Vol. 26 No. 1, pp. 10-17. Sterman, J.D. (2000), Business Dynamics. Systems Thinking and Modeling for a Complex World, Irwin/McGraw-Hill, Boston, MA. Su¨lzer, R. and Zimmermann, A. (1996), Organisieren und Organisationen verstehen – Wege der internationalen Zusammenarbeit, Westdeutscher Verlag, Opladen. Treadwell, W.A. (1995), “Fuzzy set theory movement in the social sciences”, Public Administration Review, Vol. 55 No. 1, pp. 91-6. Wagner, G. and Zipprian, H. (1985), “Methodologie und Ontologie – Zum Problem kausaler Erkla¨rung bei Max Weber”, Zeitschrift fu¨r Soziologie, Jg. 14 Heft 2, pp. 115-30. Wideman, M. (2001), Wideman Comparative Glossary of Common Project Management Terms v2.1, available at: http://www.pmforum.org/library/glossary/PMG_P08.htm#Project Williams, T.M. (1999), “The need for new paradigms for complex projects”, International Journal of Project Management, Vol. 17 No. 5, pp. 269-73. Wolstenholme, E.F. (2003), “Towards the definition and use of a core set of archetypal structures in system dynamics”, System Dynamics Review, Vol. 19 No. 1, pp. 7-26.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

The role of information and models in regulating complex commercial systems

The role of information and models 577

Stephen Brewis British Telecom, Towcester, UK Keywords Cybernetics, Information, Modelling Abstract The unsophisticated bartering systems of the past have been replaced with money and highly complex business systems. This has also required the need for management systems to direct and steer these complex business systems through commercial space seeking profitable and sustainable lineages. However, not all lineages will continue to be survival worthy, imposing pressures on businesses to seek out alternative pathways. Unfortunately, the management structures of these businesses fail to manage the complexity of the evolving operational substrate. These structures are hierarchically based and use metrics that cannot be adequately disposed through the structure that created them, resulting in the inability to effectively direct and steer the business. This paper looks at the evolution of these structures and the primary role of information and regulation to sustain value propositions in modern day commercially oriented enterprises.

Introduction In the beginning, there were no reasons only causes. The universe was intentionless. Nothing as such had a purpose. The laws of physics dominated. There was no teleology in the world. The explanation for this was simple. There was nothing that had interests. However, over the millennia born out of dying stars, complex atomic structures were created that enabled other complex compounds to form. Through mere chance these compounds evolved a structure within the primordial soup that was capable of a behaviour that prevented its own dissolution. This behavioural dynamic consisted of repeated iterations of selection from the non-random survival of random variants. Replication was the mechanism that nature had found to sustain these special types of systems. These replicators are units of information with the ability to reproduce themselves using resources from some material substrates. This was the bootstrap required for evolution to take hold. These systems had the ability to maintain and evolve structure; indeed they evolved the ability to build their own temporal survival machine to safely transit their structural information with reformulation and slight modification into the next generation. The early replicators were dumb survival machines, implementing simple survival heuristics, and as such did not care too much about their own survival. In actual fact, they were not complex enough to care at all. The survival responses to environmental perturbations were prescribed genetically, by the accumulation of wisdom over many generations. These autonomic responses enabled these very primitive replicators to deal with their world and the objects

Kybernetes Vol. 33 No. 3/4, 2004 pp. 577-589 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523580

K 33,3/4

578

in it. They walk on objects, they avoid them, they eat them and they mate with them. But they only become “aware” of these objects when they make physical contact with them. What a tremendous benefit would come to these very early replicators if they had an awareness that could prepare them of the presence of these objects. Some sort of remote sensing technology would have its obvious advantages, an awareness of an object before hitting it, of a predator to escape its attack, of food that was not already in its immediate vicinity. What might this technology be? Over time these systems became extremely complex evolving beliefs, desires and intentions. They became aware. They could sense themselves as well as their environment. They spoke not only to tell others, but also to tell themselves what they were thinking. Through such feedback systems they evolved a conscious awareness. Their reason and sole purpose was to stay alive long enough to reproduce. These systems were able to create a point of view into which external events could be partitioned into the favourable, unfavourable and the neutral to secure their safe transit into the future. This point of view evolved through a cybernetically based structure that was capable of directing and steering the system through survival space. At the heart of this structure lay the information and regulatory structures required to do this. Information and regulation were the essence of these advanced complex systems and indeed Ashby (1964) proved that for a complex system to survive and prevent its own dissolution, it required a model both of itself and its environment. Only through this model could it answer the most rudimentary of questions – faced with this predicament what do I do next. These systems did care because they were complex enough to care. They had evolved a pain and pleasure system that could guide behaviour through survival space. They were able to store and acquire knowledge through their culture and their brains. Lamarckism replaced Darwinian evolution as the mechanism that could evolve knowledge at a rate that was many orders of magnitude faster. This paper looks at how commercial systems can exploit the properties evolved by these natural systems to improve their survivability in commercial space. The role of information in system regulation Does a ship sail no better than a drifting log? To be in hell is to drift to be in heaven is to steer. George Bernard Shaw

Born out of the need to answer such questions as “faced with this situation what do I do next” evolved the need for information, information that could inform these systems to avoid the unfavourable and attract them to the

favourable. Information enabled them to choose the line of greatest advantage instead of yielding in the direction of least resistance. But what is information and what is its source. What is it that enables us to steer towards the favourable and avoid the unfavourable? Information is what changes us. It is a difference that makes a difference. Data with meaning can be viewed as knowledge, but what exactly is meaning. Computer scientists draw a distinction between Syntax and Semantics. Syntactic information is simply raw data, perhaps arranged according to the rules of grammar, whereas Semantic information through some sort of context has meaning. Information per se does not have to mean anything. It can be enjoyed as well as actioned. Snowflakes contain information describing their specific configuration. They may look beautiful under a microscope, but these configurations have no semantic content, no meaning for anything beyond the structure itself. By contrast, the distinctive feature of biological information is that it is replete with meaning. DNA stores the instructions needed to build a functioning organism. It is a blueprint or an algorithm for a specified, predetermined product whereas snowflakes do not code for or symbolise anything, DNA most definitely does. Dawkins (1989) introduced the notion of the “meme” as an “idea” that was commonly shared through social transmission. The meme, analogous to the gene, is propagated by mind through mind space. With the gene being viewed as a unit of genetic inheritance, the meme by analogy can be viewed as a unit of cultural inheritance. However, the lack of subsequent development of this concept is conspicuous with no intellectual campaigns to produce a theory of cultural replicators. The problem seems to be that the meme is a semantic classification, not a syntactic classification that can be directly observed. What then is the source of this semantic information, the information that we use to make sense of the world, and the information that is the kernel of our mental models that direct us through life? words don’t mean anything . . .

Words don’t mean anything? Information and meaning has to come from within. Information comes from the purposeful system that interprets the syntactical constructs. For a commercial system, it is the business model that provides the commercial perspective and provides the answer to the question “what does this mean? The models are the interpreters. They bridge the gap between where we want to be, the intention, the output and where are we now, the input. Models extract meaning from the data. They represent the context. They determine the meaning of the data for the systems purpose. The business model is the businesses survival model. The two models are not just similar, they are identical they are same thing. It is the business model/survival model that enables us to steer clear of the unfavourable and

The role of information and models 579

K 33,3/4

580

steer towards the favourable. It is the business model that enables us to traverse survival space to seek out new lineages before dissolving existing ones. It is through the business model that we can play out the future and determine which lineage provides investment returns commensurate with risk. It is the business model that through its current intention can answer the most rudimentary of questions, as answered correctly by our biological ancestry for over 2 billion years, faced with this predicament what do we do next? It is therefore, the business model that determines the ultimate meaning of the data; it is the business model that determines the business value of an action. It is the value propositions, prescribed by the model, that survive. Businesses remain viable by continuing to provide value into a market space that continues to select it. In this respect, Business Systems and Biological Systems are considered similar. They operate under competitive pressures, compete for limited resources and are selected by their relative fitness to others. To remain survival worthy they strive to stave of their own dissolution. They are viable systems. We will see, due to the pioneering work of Ashby (1964), Beer (1979, 1981) and Weiner (1961) not just how similar they are, but how identical their informational and regulatory structures are. The role of models in system regulation Models enable us to predict outcomes given the current state of affairs. Weather forecasting models enable us to predict the weather. Economic forecasting models enable us to predict the future economic state. But there is a fundamental difference between these two types of models. One is based on Aristotelian logic and the other on Hegelian logic, i.e. the future state is dominated by informational feedback. In other words, weather forecasting does not affect the weather, but economic forecasting can affect and does affect the economy. Hegelian systems respond to information, whereas Aristotelian systems do not. Physical events are factual, they exist and are causal whereas informational events are interpreted through a model; they may be wrongly informed or misinterpreted but generally are acted upon, creating the future. Models are created through the learning process. Learning involves a fundamental shift or movement of the mind (Senge, 1990). Learning has become synonymous with taking in information, yet taking in information is only distantly related to real learning. A learning organisation is an organisation that actions learning through the model created during the learning process. Models though are not reality. Our reality is an illusion, but it is an illusion that serves us well. What is out there is a vast resonating cacophony of waveforms, a “frequency domain”, that our senses are tuned to interfere. Indeed it is this interference that fixes our reality. In this context the real world as we know it, does not exist independently of us. It is this act of observation that brings our reality into existence. What is “out there” cannot be accessed

directly. There is no direct access to “reality”. Everything that we perceive is mediated through our senses and thought processes. Even if that reality has some kind of independent objective existence we have no means of accessing it. We are forever cocooned within our biological survival machines that evolved for the sole purpose of safely transiting our genes into the next generation. A positivist philosophy enforces these limitations in terms of what we can and cannot know. Our reality is through our sense experience. In science, there is always a temptation to say, “is this real, do these twelve dimensions or do these imaginary numbers really exist”. These questions have no meaning. All one can ask is whether mathematical models with 12 dimensions or mathematical models that use imaginary numbers provide a good description of a given situation with a specific purpose in mind. From the viewpoint of positivist philosophy, one cannot determine what is real. All one can do is to find which mathematical model best describes that situation. But how do models enable us to understand the infinite complexity that the environmental continuum presents us. We cannot deal with infinite models. To understand the complexity of this continuum we have a natural tendency to chop it up and classify it into manageable chunks (Rumbaugh, 1991). This is convenient, but we must be aware of the potential problems this will cause. Abstraction is the selective examination of a given perspective of the problem. The goal of abstraction is to isolate those aspects that are important for some purpose and to suppress those aspects we deem to be unimportant. Abstraction must always be for a purpose for it is for that purpose that what is and what is not abstracted are determined. By implication, either by design or ignorance, all abstractions are attenuators of variety. In other words we leave behind information, which may or may not be important. All abstractions are therefore incomplete and inaccurate. This reductionist approach has its advantages, but it also has its pitfalls. The abstractions we make in an attempt to simplify the complexity of the world out there are artificial and we need to be aware of the problems that may result. Abstraction is therefore attempting to decouple what is taken from what is left. We then make the assumption that any behaviour we derive from the abstraction will not be significantly affected by what has been left behind. However, we cannot remove the fact that anything we derive from the abstraction, anything we say about it, is an abridgment of reality, it is incomplete. Language is an abstraction providing us with an incomplete description of the real world. However, this does not destroy its usefulness. The purpose of abstraction is to enable us to deal with a complex situation more effectively, to limit the universe and to limit infinity so that we can do things. To elucidate some salient features from this seamless web of reality, some pattern can be identified from the ambient noise around it. We must not search for the absolute truth, but for the adequacy of the purpose. There is no single correct model of a situation other than the situation itself, which by

The role of information and models 581

K 33,3/4

582

definition is not a model. Models are either adequate or inadequate for the purpose with which they are being used. A good model captures the crucial aspects of the problem while omitting other aspects. Models help us to do things better, to sustain our lineage through survival space by reducing our future uncertainty. It is tempting to think of these models as being real, but they have evolved for the sole purpose of safely transiting our genes into the next generation. Our models are survival models, not truth models. Take for example, the process of walking through a room. Our nervous system gives us the illusion that we are walking through the room whose walls are painted in yellow. This is an illusion. What we are actually doing is walking through a spatial model of the room, which is fortunately aligned with the room with patterns on the wall that the brain through the process of evolution finds it convenient to process as a colour. This type of model enables us to move through three-dimensional space to seek the favourable and escape the unfavourable. How can such models enable us to manage massively complex and dispersed systems that are a manifestation of our culture? Before we attempt to tackle this question let us first consider the nature of this social complexity.

The evolution of social complexity About 15,000 years back, we led a very basic lifestyle in the form of the hunter-gather. Much of our time was spent sleeping and searching for food. Most hunter-gatherers were nomadic, continuously on the move, looking for suitable prey or gatherable food. Such communities were very small with woman taking on the role of the gatherer and man the role of the hunter. The homeostat in these communities represented the balance of energies between the hunting and gathering and the benefits reaped. Such was the effort required to maintain this balance that consumed all of the resources of the community. This also meant that all of the cerebral energy was solely focused on these activities and how food would be obtained the next day. For gathering to be effective choices had to made from what to gather, what not to gather, when to move on and when not to move on. These were not trivial choices, the quality of life depends on getting these choices right. Moving location to early would mean that a valuable food source was being left behind and unnecessary energy being expended to move on to a new location. It might have been woman who made one of the two epochal discoveries that began the total transformation of man’s economy from about 10,000 years back (Jay, 2000) namely, cultivation in the sense of deliberate weeding, sowing and harvesting. For the first time specific crops, which thereby became domesticated, were naturally selected to favour man’s uses. The second of these epochal discoveries was the domestication of animals. This ability to productionise food negated the need to think about the next meal and the need to travel. This

released the communities’ energies from having to think where the next meal was coming from. The agricultural era had started. This change of lifestyle released the communities potential into new areas of specialisation. If a village with its surrounding fields are generating a surplus of food above and beyond the needs of their own community then this provides a convenient place where a person supplying some other service other than food production can both live and find a market for his services. Since it was not always necessary to remain on the move, skills could be developed that enabled these communities to build more substantial infrastructures to support their needs and improve the quality of their lifestyle. Carpentry and stone masonry skills evolved were being supported by the increase in productivity of food production within the community. As these communities evolved, complexity was an inevitable consequence. Through the need to manage this complexity these societies discovered structure as a means of introducing order into the community. With this order, communities were able to grow even larger crops, which also introduced the means for further specialisation. Communities were able to trade with each other, however the bartering process was slow and clumsy and did not provide the flexibility required by traders and therefore limited the size of these communities. About 3,000 years back money was invented. That provided the trading flexibility required by these communities. This was the catalyst to further accelerate the evolution of social complexity, taking us through the industrial era to our current information age with extremely complex societies and an infinite capacity to interact. This pace, unabated, has left behind the informational and regulatory structures needed to manage these systems, resulting in social systems that are highly unstable and businesses that struggle to adapt to such volatile and complex environments. In the recent past, these environments had been relatively sympathetic to the businesses that had serviced them. These markets were very local and hence only accommodated a few players with little influence from the outside. Businesses operating in these markets were able to manage the situation with relatively simple information and regulatory structures with a form that was advanced at a relatively sedate rate. Indeed many businesses were perhaps ignorant of the information and regulatory structures they were operating with. They had emerged by Darwinian principles by the need to keep the system working. Such systems were born out of crisis but sufficed in these circumstances. They were not properly understood and therefore not intentionally designed. Markets at present are not so accommodating. They are open, heavily regulated, serviced by many players and greatly influenced by world economies. This dynamic is manifested by market behaviour that is both

The role of information and models 583

K 33,3/4

584

highly erratic and uncompromising. Today’s commercially orientated systems, the extended phenotypes of our biological systems, are faced with the challenge of sustaining value propositions, metabolised by money, within these very complex and volatile markets. Yet, the management structures of these businesses fail to manage the complexity of the evolving operational substrate. These structures are hierarch based and use metrics that cannot be adequately disposed through the structure that created them, resulting in the inability to effectively direct and steer the business. This should not be surprising considering the enormity of the task at hand. Such complexity cannot be handled within the mind of a single individual. The best that we can achieve is to break-up the problem space into a number of management domains. The problem faced by doing this is that managers will work for the betterment of their particular domains, leading to sub-optimal enterprise behaviours. This is often seen by fluctuating inventory levels and service cycle times being significantly greater than the cycle activity time; the number of inter-domain hand-offs dominates such behaviour. This is often the result of skill imbalances, poor resource scheduling and co-ordination. It can be argued that the reason behind such behaviours is the absence of a higher-level logic. Such higher-level logics can be exemplified by the recursive levels within Beer’s viable system model (VSM) (Beer, 1985), where the higher-level system “comes into being” as a result of the need for the residual variety, created by the emergent behaviour of the lower-level interacting system, to be absorbed. How can such massively complex systems be managed, controlled or regulated? How can we unfold such complexities in a formal way that we can make sense of what is going on without losing the essence of what we are controlling? Owing to the pioneering efforts of Ashby and Beer the complex organisational structures of these systems can be understood and brought under control. Beer disillusioned with classical management practices, proposed a cybernetic approach to improve organisational effectiveness based on the information and regulatory structures found in natural systems. Beer’s focus concerned the adaptation of the entire system including the operational machinery of the business to fit the environment it is serving by concentrating on how these systems were able to adapt themselves to new information from its environment. Beer’s methodology begins by recognising that systems have something – and in this case that something of interest he was seeking was viability. This special class of system seemed to possess a very special property that prevented their own dissolution they were autopoietic (Maturana and Varela, 1992). We notice such resemblances first at the levels of the similes or metaphors that management communications are like a nervous system and

the real muscle of the plant is the steel press. Such insights can be taken a stage further into the domain of conceptual modelling and may result in the formulation of a scientific analogy that can be tested. The more rigorous we make the comparison the more we may approach the specifics of homomorphism, that is a mapping relationship, i.e. one to many can be established between the systems, animal and machine, and the mathematical model. Now suppose the models can be related to each other isomorphically then we have a one to one correspondence. Then we shall have a scientific method demonstrating those aspects of the two systems we deem to be invariant, i.e. the informational and regulatory structures in the animal and the machine. In this context, the machine is any non-biological system that is deemed to be viable. What are commonly called laws in science are statements about systems of various classes that are content free. That is to say regardless of any particularity that is unique to a given system, such and such a property always obtains, i.e. it is an invariance of the systems concerned. If the work discerning how systems are viable is well done, then any viable system maybe mapped onto the model. Born out of such an inquisition is the Beer’s VSM model. A fractal-based structure showing how a complex entity can be decomposed into self-similar structures connected via informational and regulatory channels to ensure cohesion and common identity. Owing to the regular appearance of such fractal forms it would appear that the universe around us is not linear but fractal in nature. That is, we see the same pattern appearing time and again, no matter what the scale it is examined on. Look at a river with its tributaries. Each tributary is itself a river, with smaller tributaries, which in turn are themselves rivers with smaller tributaries, and so on down to creeks or a fern frond, the main frond of which consists of a two rows of sub-fronds, each of which consists of two rows of smaller fronds or leaves. Or the pattern of a coastline: bays and peninsulas contain smaller bays and peninsulars, right down to grains of sand. This invariant recursive structure provides us with the power to manage complex systems by providing the means to unfold complexity and still retain control of the whole. But there are limitations to the degree of unfolding that can take place and the system remaining viably engaged with the environment. Unfolding business complexity In complex systems, such as organisms or organisations, goals are typically arranged in a hierarchy, where the higher-level logic controls the settings for the subsidiary goals. For example, the primary goal of survival entails the lower order goal of maintaining sufficient hydration, which may activate the goal of drinking a glass of water. This will in turn activate the goal of bringing the glass to your lips. At the lowest level, this entails the goal of keeping your hand steady without spilling water.

The role of information and models 585

K 33,3/4

586

Using the fractal nature of the VSM model, the monolithic structure of enterprises can be unfolded into a federation of manageable business units within a cohesive framework working towards a common goal. Unfolding the business serves a number of purposes. Besides being used as a mechanism to manage complexity, it serves as a means of connecting the business system through a greater number of information channels to its environment through its own divisional model. Connecting to the environment in this manner makes the business highly sensitive to environmental disturbances. However, these models are logical and consist only of a functional embryology. The business logic needs to be implemented upon a suitable substrate that will provide the necessary dynamics to manifest the required behaviour. This non-trivial task involves the wide range of skills we would expect from the world of business management with special attention to organisational structure and human behaviour. The required number of organisational levels depends on the regulatory ability of the individual organisational units: the weaker that ability, the more hierarchy is needed. This is Aulin’s (1989) law of requisite hierarchy. Increasing the number of levels can have a negative effect on the overall regulatory ability, since the more levels the perception and action signals have to pass through, the more they are likely to suffer from noise, corruption, or delays. The effect of such delays is often seen in production lines where the process cycle time can be much larger than the activity time. Queuing theory tells us that it is the number of hand-offs between organisational units that dominates this type of behaviour. It is best to maximize the regulatory ability of a single layer, and thus minimize the number of requisite layers. This principle has important applications for complex organisations, which have a tendency to multiply the number of bureaucratic levels. The present trend towards the flattening of hierarchies can be explained by the increasing regulatory abilities of individuals and organisations, due to better education, management and technological support.

Business models – the consciousness of the enterprise Complex businesses are replete with information. Unfortunately a large amount of the information does not code anything, i.e. it is meaningless. For the information to be meaningful, the interpreter must specify it in advance since it is the role of the interpreter to dispose the information. It is the business model that performs this role. It is the business models that are the consciousness of the enterprise. Only through the business model can we make the enterprise aware of itself and its environment. The business model is the anatomical map of the enterprise. It is the business model that feels pain and seeks out pleasure. It is the business model that guides the enterprise through commercial space. Without the business model we are lost.

For a business to dispose requisite action, it requires a body map that is capable of representing the anatomical parts of its enterprise so the actions it disposes can avoid pain and seek pleasure.

Not all business systems have body maps. Simplex business systems comprising of not more than one person will almost certainly have. Body maps make us aware of our anatomy. They enable us to avoid pain and seek pleasure. They direct and modify our behaviour. Complex business systems are not normally business aware. They are not aware of the value of their actions upon quality business survival. The business model is not granular enough to create a point of view into which information can be actioned. They do not feel pain and cannot detect pleasure. They lack any form of emotional behaviour. They are not able to learn from their actions. They are not able to adapt their behaviour to ensure that their actions are survival worthy. We use pleasure to make all types of decisions, all decisions are made to maximise pleasure including apparent altruistic actions. However our mental models prevent us from being slaves to instant gratification. We can plan for the future without succumbing to immediate pleasure, but even this decision can be viewed as a strategy to maximise long-term pleasure. These strategies are played out through models that maximise long-term success. models without practice are dangerous Practice without models is fatal

Businesses, the reward systems provide their stakeholders with pleasure in return for their personal investment. They all therefore have an interest in the business surviving. If the business achieves they achieve, if the business fails they fail. Pain, pleasure our motivation, emerged for one simple purpose – to guide actions. Pleasure is the common currency we use to prioritise our actions. How can we guide our businesses through commercial space if we cannot feel our way? Where are the pain trajectories that we should avoid and where are the pleasure trajectories that we should seek. The commercial landscape offered up by the environment is infinite. The only way of avoiding pathways that are hazardous and seeking out pathways that are pleasurable is through the business model. The business model is not an add-on to the business, it is the business. The business model must guide every action, only through the business model can we ask ourselves the most basic of all survival questions: faced with this predicament what do I do next. Conclusions Every living organism alive today has at least one thing in common, not one of their ancestors died in their infancy. If they had they would not be here today. All their ancestors were able to continue their lineage through survival space. They were able, as vehicles of genetic propagation, to stay alive long enough to

The role of information and models 587

K 33,3/4

588

fulfil their purpose. They were able to successfully transit their genetic code with some minor random modification into the next generation before they dissolved back into their environment. They possessed body maps that enabled them detect, locate and avoid pain and a pleasure system to reward and guide behaviour that improved their survival chances. They had what it takes to stay alive long enough to accomplish this task. This was their sole purpose; this was the meaning of life. At the core of all of these living structures was information and regulation. As these organisms became more and more complex to deal with more hazardous environments and competing gene pools these structures become more and more sophisticated at making a living and staving of dissolution. To become bigger, faster and more agile required a nervous system that was able to control all aspects of the organism. Commensurate with control is information. Without these two major functional ingredients the organism would cease to exist. Theories destroy facts. Mendeleyev brought order to the world of the physical elements; Darwin brought order to the world of living things. Weiner and Beer have brought order to the apparent chaotic world of survivable systems. They recognized the primacy role of information and control in the animal and machine. They spent a lifetime providing new insights of how information and regulation was used to control these systems. Business, Enterprises, Governments and Social Organisations are all types of complex systems and in order to survive in a competitive environment they need as part of their embryology, informational and regulatory structures. Ashby provided us with insights concerning requisite regulation that we achieve through our body maps and mental models of our environment. We use these models to seek out pathways through survival space. Beer provided us with a fractal based invariant structure for any survivable system. It may sound a little contrived, pretentious even, to describe a commercial system in terms of the world of the living, but through the work of Beer and others they have shown us that they do indeed have a lot in common. DNA the molecule of life responsible for its own destiny and its unending pilgrimage through geological time has got us where we are now. The journey has taken billons of years and will no doubt continue well into the future. Business models by contrast are the DNA of commercial systems they are responsible for their own destiny. It is the business model that is the blueprint for the commercial system. But for Business Models to be as successful as their biological counterparts they need to pervade every nook and cranny of the commercial enterprise providing a point of view into which we can answer the most rudimentary of questions “faced with this situation what do I do next to maintain a quality survival”. Through the work Beer undertook in the area of Managerial Cybernetics, we can ensure that commercial systems have the informational, regulatory

structures as part of their embryology that has sustained the biological world for over billions of years References Ashby, W.R. (1964), An Introduction to Cybernetics, Methuen, London. Aulin, A. (1989), Foundations of Mathematical Systems Dynamics, Pergamon Press, Oxford. Beer, S. (1979), The Heart of Enterprise, Wiley, Chichester. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester. Beer, S. (1985), Diagnosing the System for Organizations, Wiley, Chichester. Dawkins, R. (1989), The Selfish Gene, Oxford University Press, Oxford. Jay, P. (2000), Road to Riches, Orion Publishing, London. Maturana, H. and Varela, F. (1992), The Tree of Knowledge, Shambhala, Boston, London. Rumbaugh, J. (1991), Object Modeling and Design, Prentice Hall, Englewood Cliffs, New Jersey. Senge, P. (1990), The Fifth Discipline, Century Business, London. Weiner, N. (1961), Cybernetics, 2nd ed., MIT Press, Cambridge, MA.

The role of information and models 589

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

590

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

A self-organizing network for the systems community Jose´ Pe´rez Rı´os E.T.S.I. Informa´tica, Universidad de Valladolid, Valladolid, Spain Keywords Cybernetics, Control systems, Information networks Abstract The number of approaches, researchers and practitioners in Systems Thinking has been growing over the last decades and particularly in recent years. This wide and deep evolution within the systems community that is a sign of vitality has also increased the communication difficulties among its components. If the systems community wants to overcome these difficulties and help cross fertilisation, adequate structures are needed to facilitate the communication processes. In this work, we present a communication and exchange of information system specifically designed for this aim. It tries to use the possibilities offered by Internet to create a systems thinkers/practitioners virtual community with the capacity to facilitate the creation of as many new partial communities as the members of the systems community wish. The design of the network system is based on Beer’s viable system model and his recursive conception of viable systems.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 590-606 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523599

1. Introduction Systems thinking and practice have experienced a rapid growth over the last decades and particularly in recent years. This growth gave birth to different systems communities working in relative isolation. The systemic approaches also experimented a big change and evolution. For instance, we have seen how the development of the functionalist approach, the interpretive and, more recently, the critical approach gained in diffusion and number of adopters. This growth in the number of schools, researchers, practitioners, methodologies, methods, tools, etc., increased the fragmentation of the field and made communication among its members and access to the information produced much more difficult (Schuhman, 2003). If we take the case of the systems approaches related to Management just as an example of this growth process, we can witness the deep evolution that took place through the second half of 20th century. Midgley (2000) discusses three main waves within Systems Thinking. Examples of representatives of the first wave are Systems Engineering, Systems Analysis and the early version of System Dynamics. The first wave approaches were later criticised for considering the models of problematic situations as representations of reality and for an insufficient consideration of individuals’ perception of that reality. The second wave of systems thinking put emphasis on dialogue and inter-subjective construction of realities. Some approaches representative of this wave are strategic assumption surfacing and testing (SAST), soft systems methodology (SSM), interactive planning, etc. In parallel to the development of new approaches, some of the methodologies that appeared during the first wave experimented an important evolution that included adoption of some

characteristics of the second wave. This is the case, for instead, of system A self-organizing dynamics (Lane and Oliva, 1998; Schwaninger and Pe´rez, 1998). By the end of network the 1980s, a third wave of systems thinking comes to light. It does so as a result of the critics made to second wave approaches for what was considered an inadequate treatment of the power relations. An example of this approach is critical systems heuristics (CSH). Another aspect that appears within this wave 591 concerns the increasing interest in the use of multiple methodologies in the same study. This subject has been treated by Jackson (2000), Jackson and Keys (1984), Midgley (2000), Mingers and Gill (1997), Pe´rez and Schwaninger (1996), Schwaninger (1997) and among others. This methodological pluralism has raised a strong debate about its feasibility. Discussions about paradigm incomensurability came to light, as well as theoretical developments attempting to deal with this question like the “virtual paradigms” concept of Yolles (1999). This wide and deep evolution within the systems community at the same time that signals its vitality also increases the communication difficulties among representatives of different paradigms/methodologies. If the systems community wishes to overcome these difficulties it may be convenient to help cross fertilisation by developing adequate structures that can facilitate the communication processes. This paper describes a communication and information exchange system specifically designed for this aim. It pretends to use the possibilities offered by the Internet to create a systems thinkers/practitioners virtual community with the capacity to facilitate the creation of as many new partial communities as the members of the systems community wish. In terms of viable system model ( VSM ) of Stafford Beer (Beer, 1979, 1981, 1985), we could be talking of developing something like a recursion zero for the systems community that will contain many other recursion levels. The process of unfolding the complexity (Espejo, 1989) represented by the variety of the fields of study and applications and systemic approaches striving for their viability can be done according to many criteria (dimensions) and can generate multiple recursion levels, but all of them must fulfil the conditions for viability identified in the VSM. Since the system we are proposing has been designed according to VSM we think it can help to increase the viability of the systems community as a network of networks. In what follows, and after a brief consideration about the conceptual foundation of the system proposed, we will show the structure of the system and how it can be applied to this scientific community. Then, we will review its main features and possible uses, and will end, after a short comment about new developments in progress, with some final conclusions. 2. Conceptual foundations of systemsnet.org The design of systemsnet.org is based on the VSM of Stafford Beer. We take his reflections about the conditions for viability as a very useful guide in the

K 33,3/4

592

process of development of this system. VSM is one of the contributions of Stafford Beer to organisational theory. In it he sets the conditions that a system must fulfil to be viable (that means to be able to keep a separate existence). These conditions imply that the system will have the capacity for self-regulation, adaptation and evolution. One of the conditions for a system to be viable is to have the capacity to respond to the relevant complexity in its environment. If we take variety (number of possible states that a system may present) as an indicator of complexity, controlling a situation means being able to respond to its variety. In this sense, Ashby law states, “only variety can absorb (destroy) variety”. The way to do that is by developing attenuators of variety (from the environment to the system) and amplifiers of variety (from the system to the environment). This process is known as “variety engineering”. After these considerations about variety, let us briefly review the main elements of the VSM that we will use later to explain the structure of systemsnet.org. According to VSM, a system is viable if and only if it has the five functions identified by Beer as Systems 1, 2, 3, 4, and 5. System 1 is made up of the productive processes that allow the organisation to produce the services and/or products that justify its existence. As we will see later, each network (within the network of networks that is systemsnet.org) is an element of System 1 at a certain level of recursion. Each of these elements (networks) has their own connection with the environment and their own management metasystem. All the other systems (from 2 to 5) have the mission of serving, regulating and directing System 1. System 2 takes care of the coordination activities, its main function being to damp oscillations that may take place as a result of the activity carried out within the elements constitutive of System 1 and their interactions. System 3 is responsible for the “here and now” of the organisation in real time. Among its functions we can signal resources bargaining with the primary operations (System 1), transmission of instructions, auditing and intervention when coordination has not been able to solve conflicts among the operations composing System 1. The audit function is carried out by what is called System 3*. The function of System 4 is to continually watch the evolution of the environment (external to the organisation). It should take care of the “outside and then” of the organisation to guarantee its adaptability. Ideally, System 4 should be composed of an “operations room” where different scenarios for the future can be continually explored to help make the decisions that will increase the probability that the future for the organisation is the desired one. Something that can strongly help this activity would be the availability of a model of the organisation itself built with System Dynamics (Pe´rez, 1998). Such a model would facilitate the communication between the representatives of functions belonging to Systems 3 and 4, whose interests are initially quite different.

Finally, System 5 provides the logical closure to the viable system and A self-organizing monitors the three-four homeostat. It is a function of this system to take care of network the ideological, normative aspects for the organisation. It will also define its mission and style. System 5 has to watch that an equilibrium exists between the adaptation to the changing environment and an adequate internal stability. Another essential aspect of VSM is the consideration of the “recursive” 593 character of viable systems. All viable systems contain viable systems and belong to systems that are themselves also viable systems. A particularly interesting feature of this characteristic is that any of these systems must have the five systems identified in the VSM that characterise viability. The viability of the system requires that all five functions are present in all levels of recursion of the organisation. This conceptual framework will be used in the design of systemsnet.org as will be shown later. 3. Structure of systemsnet.org The general structure of systemsnet.org is composed of three main components, namely, technology, management of systemsnet and new developments. The first one will take care of the design and update of the physical elements (hardware, security, etc.). The management component will handle all the activities related to the administration of the system. The function of the new developments component is to continually explore and develop new functionalities for the whole system, trying to keep, and if possible increase, its usefulness through time by its adaptation and evolution. By new developments, we mean the addition of contents. For example, the supply of systems thinking related doctoral programs worldwide. An example of this is the work done by the University of Valladolid in identifying the doctoral programs actually available in all the Spanish and Portuguese speaking universities. This includes all Spanish-speaking countries of America, plus Brazil, Portugal and Spain (Sebastia´n, 2001). This information has been made accessible through Internet in the addresses www.grupotordesillas.org and www.iberfora2000.org (Pe´rez, 2001). But, new developments also refer to new applications that will be made available to the members of the systemsnet (e.g. a research exchange Internet-based system, computer supported collaborative work tools, etc.). The design of systemsnet.org is based on the VSM and its recursive character. We pretend that each network of systemsnet.org will have all the components necessary and sufficient for its viability (independent existence, self-regulation, learning and adaptation capacity). In Figure 1, we have a general representation of systemsnet.org using the VSM where we can see, just for illustration purposes, a network containing three subnetworks. In Figure 2, we show an example of what could be a second level of recursion for the earlier network. When we access, through Internet, a specific network the initial screen will show us general information, news, links, etc., concerning this network. It is a place where the members of the network can put information related to the

K 33,3/4

594

Figure 1. VSM of systemsnet, network/subnetworks

vision, mission, ethical code, etc. Also on this screen we will have a button, which, when pressed, will take us to the cooperative space exclusive of this network. Once inside, all functionalities, which will be described later, will be at our disposal. Having made these comments about the structure and design of systemsnet.org we will review some of the content which has been included within this system and which covers some of the main functions identified in the VSM as a requirement for viability (Figure 3). All that follows applies to any network, at any level of recursion. System 1 will be made of networks, subnetworks, etc., each of these corresponding to a methodology, a research group, or any other community of

A self-organizing network

595

Figure 2. VSM of systemsnet, showing second level of recursions network/subnetworks

interests within the systems community. Criteria for creating these networks are up to the members of this community. Among the tools included in systemsnet.org to facilitate the operation of System 1 we can mention a knowledge capturing system, which allows uploading, downloading or consulting information in each network. Other tools that are also included are a messaging system and a chat-room, which will complement and facilitate the communication activities that can be carried out through other means like telephone, personal meetings, etc.

K 33,3/4

596

Figure 3. VSM of systemsnet, networks/subnetworks showing some tools and content

The content of System 2 in systemsnet.org can be formed, for instance, by the declaration of codes of behaviour for the members of a network. As an example of tools included in systemsnet.org to help System 2 functions, we can mention a scheduling system specific for each of the networks. As examples of functions related to System 3 (“inside and now”), we can signal the web master of systemsnet.org, the managing editors of Scientific Journals, etc. Tools included in systemsnet.org to facilitate the operation of

System 3 are an information handling system, a messaging and chat module A self-organizing and a computer supported collaborative work system (to help group problem network study, etc.). Examples, of what could be considered activities related to System 3* are the ones carried out by Scientific Quality Boards. System 4 is in charge, as discussed before, of the “outside and then”. It must watch what is happening in the environment at present, but also “explore” the 597 future to make the adaptation of the whole system possible. In addition to managing the “outside and then”, System 4 must also provide self-awareness for the system-in-focus (the network). It can be done by developing the model of the system-in-focus and also the model of itself. There are many tools which can be used to help these functions. System dynamics (SD), for instance, allows the exploration of dynamic behaviour and alternative developments through time. In systemsnet.org, besides SD we will add other tools like Delphinet (an application which allows making Delphi studies through Internet), VSMod (a software tool which facilitates the application of VSM and which will help the network to see itself ) (Pe´rez, 2003), a research exchange (which will facilitate the identification of who is doing/wishing to do what), and a computer supported collaborative work system (to help the collective exploration of the “outside and then” and modelling of itself). The content for System 5 will be formed by the vision, mission, ethical codes, etc., agreed upon by the members of the network. Finally, we say that in order to facilitate the interaction between Systems 3 and 4, systemsnet.org includes: a computer supported collaborative work system, a Fora creation system, and a specific software created to facilitate the organisation of meetings. This interaction can also be facilitated by using team syntegrity (Beer, 1994). Team syntegrity is a specially designed method developed by Beer to help the interaction, among other uses, between Systems 3 and 4.

4. Recursion in systemsnet.org Systemsnet.org makes use of the recursion concept in its design and way of operating. Different systems approaches emphasise different aspects of the problems under study, and also the interests of the people making the study may vary (research, consulting, teaching, etc.). This means that the process of creating networks/subnetworks/sub-subnetworks will be quite diverse. Systemsnet.org provides the flexibility to deal with this variety of approaches and interests. Its users can create networks, subnetworks etc. as a process of deepening into a certain approach (in this case we will be making use of the recursion concept) or they may create different networks which are relatively independent of each other. In either case, all networks will have access to the functionalities available within the system. Figure 4 shows the representation of part of the structure of systemsnet.org. The figure also shows

K 33,3/4

598

Figure 4. Partial structure of systemsnet.org

that all functions available within the system are accessible to any of the networks/subnetworks/sub-subnetworks. In Figure 5, we show, just as an example of how the system can be used, how different systems approaches can generate networks at different levels.

Figure 5. Partial structure of systemsnet.org with examples of systems approaches

5. Features included in systemsnet.org A self-organizing In what follows we will comment on some of the functionalities available network within the system. When we visit the Internet address www.systemsnet.org, the first screen we see is that of Figure 6. Some of the possibilities offered to the users of the system are listed there. As it is shown they can create networks, subnetworks, sub-subnetworks, etc. according to any criteria (thematic, 599 geographical, methodological, etc.). They can share knowledge (load, download and/or see information), have access to colleagues’ data, and exchange messages with them. They can also create fora to discuss subjects of interest or to plan work or events. A scheduling feature included in the system will help them to organise and follow up activities to be carried out within a network. All these features are available within any particular network and at any level of recursion. Some additional features, described later, will be incorporated into the system in the future. Besides the general information just described, the main screen also has the entry point to the operating system itself. The name of the application is “Collaboration Nets”. Pressing there will take the user to the space where he/she can make use of all the functionalities available. Access is restricted to the system community members, so a login and password will be necessary. 5.1 Searching networks Before entering into the system we can explore the networks that have been created. To facilitate the search, we have included a search engine based on UNESCO codes that will give us all the networks corresponding to a certain code. We can also see the list of all the networks available. This feature is useful

Figure 6. Main screen

K 33,3/4

600

for investigating whether other groups working on a subject we are interested in exist. If we find a network that we would like to join, the system will facilitate the request to be admitted (and the admission itself if the request is accepted by the creators of that network). Once we have decided to enter the system and introduced our login and password, we will see a screen (Figure 7) where all the networks to which we have access are listed, either because we created them or because their creators accepted us to become a member of them. After selecting the network we wish to get into, pressing enter will take us there. 5.2 Features Once inside the network we have access to all functionalities available within the system. The first thing we see is a general description of the network (Figure 8). Next we can, if we wish, consult the information available in this network. We can consult all the information available, or just the one uploaded by a certain person, by a certain institution or by the user. In addition to just consult what information there is, we can download the information we wish or load a new one. Another feature of this system refers to the communication among the members of the network. To help this, the first thing we can do is checking the list of all of them (Figure 9). We have all the personal data in it (address, e-mail, telephone, etc.). We can also make use of the messaging system embedded in the network (Figure 10). The selection of addressees is straightforward (just mark them in the screen). After the selection we just write the content of the message, add any attachment if desired and send it.

Figure 7. List of networks with access

A self-organizing network

601

Figure 8. General data of a network

Figure 9. List of members in a certain network

The system also includes the possibility of opening fora. Within a certain network we can open as many fora as we wish or we can consult the ones opened already, as well as the tree of answers that have been produced in each of them.

K 33,3/4

602

Figure 10. To send messages to members of this network

In addition to everything that has been discussed above we have the possibility, also within a certain network, of using a scheduling facility. This can be useful for organising events or projects and for keeping track of the degree of execution of the different milestones. Besides all the possibilities we have mentioned related to a particular network we can also move to other networks, where we have access permission, and use them in a similar way as described. In case we wish to enter into a network where we have no access permission, we can make a request for it. To join, we just fill in the application form (Figure 11) and send it. This will reach the creator of that network who will have the capacity to either accept or reject our request. The system also facilitates this task. Finally, if we cannot find any network that fulfils our interests we can create it ourselves. Figure 12 shows how to create it. We consider this facility as an important one because it allows the evolution of the systems community by creating, in a self-organising way, the subnetworks that satisfy the emerging interests. Any member of the system is allowed to create the networks he wishes. Having done that he/she will have the privilege to accept other members to join the newly created network. 5.3 Possible applications for systemsnet.org and future developments The scope of applications of systemsnet.org is quite wide. Some examples of these are: . contribute to the viability of the systems community; . create and facilitate activity of research groups;

. . . .

identify possibilities of collaboration and its follow-up; facilitate the organisation of encounters/conferences/symposiums etc.; identify, create, develop doctoral/post grade teaching programs, etc.; create thematic groups and facilitate activity within them;

A self-organizing network

603

Figure 11. Form to join another network

Figure 12. Form to create a new network

K 33,3/4

604

. .

open and facilitate debates; and publish news, activities, events, etc.

The system presented here has been in constant evolution since its first version created back in 2000. Since then a number of structural changes have been made and new features added. We plan this evolution to continue in the future. Some of the more immediate improvements will be related to the possibility of handling different types of users and the improvement in the way the information is handled and organised. This will be necessary as the number of users and the amount of information loaded into the system grows. Other improvements in the system utility will come from the addition of several new features as well as from the possibility of using software developed to satisfy specific needs of the systems community. Some examples of this software are: . computer supported collaborative work tools (Internet-based collaborative knowledge capturing system); . VSMod (software to facilitate the application of the VSM; . Delphinet (software to make Delphi studies through Internet); and . software to facilitate the organisation of meetings, conferences and similar events; . research exchange system (meeting point through Internet between supply and demand of research). Adding specific contents of interest to the systems community can also enlarge the value of systemsnet. For instance, the information about the worldwide supply of doctoral and master studies and availability of teaching materials is particularly interesting. Other areas where there is more room for research and improvement in systemsnet are the ones related to the behaviour of its users. It would be very convenient to know how they can be motivated to participate, exchange information and develop communications beyond information (Espejo, 1996). 6. Conclusions The rapid growth of systems thinking in the last decades increased the difficulty for communication and cross fertilisation among the different systems approaches. The new information and communication technologies, together with systems theory, may help to overcome this problem by facilitating the creation of adequate recursion levels to face the complexity of the field of study, including a possible recursion zero level. In this work, we have presented an Internet-based tool (www.systemsnet. org) that, using cybernetic principles and the VSM in its design, facilitates the creation of these recursion levels and the exchange of information and communication among their members. The use of a recursive design makes it

possible that all levels have identical access to a set of facilities included in the A self-organizing system. Among these we can mention the possibility of loading or unloading network information, opening fora for debate and using a specific messaging system. A particularly interesting feature of this system is the possibility offered to any member to create new networks as desired fitting his/her needs (research, consulting, etc.). Since this facility is entirely decentralised, we think it will help 605 the self-organisation of the systems community. In addition to the information and communication features mentioned, some other facilities can be included in the system which, we think, will add value to it. Examples of these are a computer supported collaborative work tool, the VSMod software for the application of the VSM, Delphinet (an application for making Delphi studies through Internet), among others. Finally, we can say that this system may also act as a repository of systems thinking education materials (supply of doctoral, master or other studies available worldwide) making them more accessible and facilitating its diffusion. References Beer, S. (1979), The Heart of Enterprise, Wiley, New York, NY. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, New York, NY. Beer, S. (1985), Diagnosing the System for Organizations, Wiley, New York, NY. Beer, S. (1994), Beyond Dispute, Wiley, New York, NY. Espejo, R. (1989), “The VSM revisited”, in Espejo, R. and Harnden, R. (Eds), The Viable System Model: Interpretations and Applications of Stafford Beer’s VSM, Wiley, New York, NY. Espejo, R. (1996), “Requirements for effective participation in self-constructed organizations”, European Management Journal, Vol. 14 No. 4. Jackson, M.C. (2000), Systems Approaches to Management, Kluwer Academic/Plenum Publishers, Dordrecht/New York, NY. Jackson, M.C. and Keys, P. (1984), “Towards a system of systems methodologies”, Journal of the Operational Research Society, Vol. 35. Lane, D.C. and Oliva, R. (1998), “The greater whole: towards a synthesis of system dynamics and soft systems methodology”, European Journal of Operational Research, Vol. 107. Midgley, G. (2000), Systemic Intervention: Philosophy, Methodology, and Practice, Kluwer Academic/Plenum Publishers, Dordrecht/New York, NY. Mingers, J. and Gill, A. (Eds) (1997), Multimethodology. The Theory and Practice of Combining Management Science Methodologies, Wiley, New York, NY. Pe´rez, R.J. (1998), “Diffusion of cybernetics: a preliminary system dynamics model”, in Espejo, R. and Schwaninger, M. (Eds), To Be and Nor to be that is the System: A Tribute to Stafford Beer, CD-ROM, Carl Auer-Systeme Verlag, Wiesbaden. Pe´rez, R.J. (2001), “Formas de colaboracio´n en la red Grupo Tordesillas”, II Encontro de reitores Universidades do Brasil, Espanha e Portugal, Recife, Brasil. Pe´rez, R.J. (2003), “VSMod: a software tool for the application of the viable system model”, 47th Annual Conference of the International Society for the Systems Sciences (ISSS), Heraklion, Crete, Greece.

K 33,3/4

606

Pe´rez, R.J. and Schwaninger, M. (1996), “Integrative systems modelling: leveraging complementarities of qualitative and quantitative methodologies”, 1996 International System Dynamics Conference, Boston, EE.UU. Schwaninger, M. (1997), “Integrative systems methodology: heuristic for requisite variety”, Int. Trans. Opl Res., Vol. 4 No. 4. Schwaninger, M. and Pe´rez, R.J. (1998), “ISM – integrative systems methodology: a hybrid methodology for applying cybernetics cybernetically”, in Espejo, R. and Schwaninger, M. (Eds), To be and nor to be that is the System: A Tribute to Stafford Beer, CD-ROM, Carl Auer-Systeme Verlag, Wiesbaden. Sebastia´n, J. (2001), La formacio´n doctoral en Ame´rica Latina y la colaboracio´n de las universidades espan˜olas, Proyecto IBERFORA, Universidad de Valladolid. Schuhman, W. (2003), “Observing experiences with the VSM”, Symposium about Stafford Beer Life and Work, 3 March 2003, London School of Economics, London. Yolles, M. (1999), Management Systems. A Viable Approach, Financial Times/Pitman Publishing, Great Britain. Further reading Espejo, R. and Schwaninger, M. (Eds) (1997), To be and not to be that is the System: A Tribute to Stafford Beer, CD-ROM, Carl Auer-Systeme Verlag, Wiesbaden. Sa´nchez, M.P. and Pe´rez, R.J. (2001), “Desarrollo de investigaciones delphi en Internet”, IV Congreso de Ingenierı´a de Organizacio´n: CIO-2001.

File name{KYB}Issues/33_3-4/Part 3.3d 09:51 – 1/4/04 – Total page(s) 2

PART III Illustration by Stafford Beer (III)

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Observing experiences with the VSM Werner Schuhmann Practitioner in Cybernetics, Germany

Observing experiences with the VSM 609

Keywords Cybernetics, Management, Control systems, Systems theory Abstract This paper offers a senior manager’s reflections on experiences with the viable system model (VSM) in different situations of its application. These reflections are based on second-order observation; that is on observing the forms of first-order distinctions. Recently, systems thinking developed by Niklas Luhmann serves as an epistemological background. The VSM is recognized as a valuable tool to study and to change organizations, however, its “upgrade” is recommended in the light of insights, such as the protologic of distinctions or the more consistent realization of systems properties, such as operational closure and self-reference.

1. Introduction The German philosopher Hans Georg Gadamer once stated that experience is one of the least enlightened concepts. This sentence is astonishing, but only at first sight: Luhmann called his Systems Theory a labyrinth. If that is an adequate metaphor, then the current confusion of tongues in the systems community may be properly characterized by the Tower of Babel. Many important aspects of current theories are controversial and thus indicate many white or opaque spots: for instance, logical foundations of the epistemology (trans-classical logic), the theory of observation and self-reference of social systems, the modus operandi of autopoietic systems, their multiple structural couplings or the evolution and co-evolution of social systems. Basic terms of terminology such as operation, observation, system, organization, element, structure, information, communication and many others are quite often confusing. However, not only the “sciences of communication and control” have problems with communication amongst themselves. If we have a look at the whole range of social sciences (e.g. economics or business-administration, which were of special interest for me as a manager) then it is even necessary to invent a metaphor for the description for their state: a “Collapsing Labyrinthic Babylonian Tower”! With this situation I was already confronted, when I was looking for scientific support for my everyday tasks. I developed my current position during a long-term learning process based on switching between the complementarity of theory and practice. My systemic foundations are an amalgam of studying the “Greats” such as, Stafford Beer, H. von Foerster, Niklas Luhmann, a few scientists such as, e.g. Dirk Baecker, Peter Fuchs, a close cooperation and friendship with Raul Espejo and Markus Schwaninger and reflections upon my learning and doing.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 609-631 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523607

K 33,3/4

610

Having in mind the above-mentioned metaphors, I think it is useful to offer a brief summary for a better understanding of my argumentation. (1) “Complexity is the very stuff of today’s world” (Beer, 1975, p. 15). Stafford Beer’s sentence is my point of departure. It suggests that we have to give up the ontological view of “one reality” and to replace “identity” (ontological entity, whatever it may be) by differences. Consequently, the recurrent theme in this paper is the “Logic of Distinctions” (Schuhmann, 1998; Spencer Brown, 1997) and the “Theory of the Observer”, that is, the concept of “Operationally closed self-referential Systems” (von Foerster, 1993a). In my opinion, both fundamental topics are still in their infancy. (2) The “world”, the “reality”, does not contain any information. Information about complex “worlds” can only be achieved by observation that is drawing a distinction by indicating one side of it and not the other (Spencer Brown). The outside of a distinction remains unmarked, empty and blind; yet it serves as a background (context) that is necessary for any distinction. The inner side of a distinction (the marked space), the outside of the distinction (the unmarked space) and the boundary between are called “the form” of the distinction. The form of a first-order observation can be seen only by a second-order observation, which in turn has again its unmarked side, its blind spot. Second-order observation is observation of forms. Of course also the blind spot of a second-order observation can be observed by another observation, but again only at the cost of an unmarked side (Baecker, 1993a). It is unavoidable, that the generation of any information (“seeing”) simultaneously produces blindness. (3) This process of constructing our realities by drawing distinctions and making indications (e.g. explanations, descriptions) leads to a fundamental paradox of cognition: to explain the birth of the first information without parents, e.g. the first distinction and indication. Spencer Brown begins his book by starting with an injunction: “draw a distinction” (Spencer Brown, 1997, p. 3), therefore taking as given the idea of distinction and indication (but only at the very beginning!). Later in his book, he resolves the paradox (drawing a distinction, without having one and no observer, who could note this injunction!) by introducing the re-entry, the form within the form. Luhmann follows this approach with his famous dictum: “The following considerations assume that there are systems” (Luhmann, 1984, p. 30). (4) The logical innovation of the “form within the form” leads to a totally different viewpoint: no longer can systems be regarded as ontological units (e.g. as holistic wholes consisting of related parts) but should be taken as “units of a difference”, as forms! The “form” is produced by the inside of the distinction, that is the system, the outside, that is the specific environment (as seen by the system) and the (also self-produced) boundary. The form of the system is generated by the same kind of internal operations. The form within the form, the re-entry, allows observation of both the system itself and its environment.

Thus, the same distinction of “system/environment” that generates the system Observing is copied into the system and produces a model of its form. Again we have an experiences with outcome of the fundamental paradox: the circularity that systems are produced the VSM by observation (observers), which are generated by their system. A system is always an observer that gives birth to itself and vice versa. Enlightenment enlightens itself. The re-entry is a logical way to formally unfold this paradox. 611 (5) An operation is the reproduction of one system’s element (e.g. one thought in a consciousness-system or one communication in a social system). However, although we call it an element, one must not think of it as prefabricated or available in an outside storage. The system itself produces the elements of which it consists with the help of the elements it consists of. This is the modus operandi, the autopoiesis, of systems. As the elements of systems are defined as instants, systems “exist” only in the very moment these instants emerge and disappear. (The same system is different every moment.) Time coherence of systems is a property seen by an observer, applying the difference before/after. Any operation of a system will be identified by the following next operation, which in that case acts as observation. Then, this operation (at present in its capacity as operation) will be identified by the next one. This is to say that any operation is an observation and any observation is an operation. One operation carries out two functions. There are no first or last operations: the first one would not be an observation and the last one could not be identified as an operation. There is no beginning and end; there is only the circularity of operations, autopoiesis, and ouroborus. (Fuchs, 1995, p. 22). The unity is a difference, the one is many and the same (distinction) is not the same (as re-entry)! There are several ways to resolve the paradox of self-referential forms. But we are confined to language (Niels Bohr referring to Quantum Theory) and therefore, urgently need more formalization of second-order cybernetics. (6) Points four and five present in detail a few of the essential concepts of Luhmann’s theory. The reason for this is to emphasize the many difficulties in debates with systems scientists, for instance, about organizational learning, development of culture (how does the memory of a system work?), management and the structural coupling between consciousness-systems and organizations or the unfolding of complexity. Many points in Luhmann’s theory may be controversial, but if we do not go down to this micro-level of systems theory, debates lack the necessary complexity and therefore, in many cases are not very productive. (7) Complexity leads to the undecidability of fundamental questions (Beer, 1975, pp. 13, 374; von Foerster and Bro¨cker, 2002, p. 178) and forces us to decide for a specific answer for which we have to take the responsibility. Different distinctions lead to different answers, which result in different constructions of “realities” (e.g. different social or economic theories). Therefore pretending wholeness, completeness or general truth is an observer’s illusion.

K 33,3/4

612

Experience emerges as an eigenvalue of recurrent observations within a specific system, a combination of perception, cognition and operations. (8) Different “realities”, e.g. different definitions of problems and solutions, imply the risks of dissent about questions such as: what is the case, what is relevant, what is behind the situation, what to do, and what is possible and what is preferable? (9) Dissent mostly ends up in conflicts. Whether we like it or not, the driving forces for different distinctions are different interests (as motives). Both irritations of outside systems and information or communications within social systems, quite often camouflage these interests behind the distinctions. We cannot read other’s minds; therefore, we are intertwined in networks of communication. But communication cannot be governed by a democratic consensus-machine. Therefore we need a heuristic, how to deal with dissent as an unavoidable consequence of complexity. (10) Systems thinking offers such a heuristic: the second-order observer. The viable system model (VSM) is one systemic language (amongst many others) with an adequate variety to communicate the results of idiosyncratic distinctions in the realm of organizations. (11) In order to cope with the variety of the world, to bear the inconsistencies and ambiguities of the numerous observable realities and to live with undecidable questions, trust and tolerance are needed as the cybern-ethical foundation of communication (von Foerster, 1993b). They are also the “gravity” for autonomous social system’s dynamic equilibrium. Information is also formed (understood) by the listener (addressee), who could be right as well. (12) My favorite insights from systems theory are: . Beer’s dictum: “absolutum obsoletum”, . application of this sentence also to system theories and to itself, . seeing implies blindness, and . “Ought to” implies capabilities and resources to do. It took me a long way to arrive at this temporary position and those insights. This ongoing learning process changed fundamentally my thinking and feelings. Stafford Beer’s work was a trigger, offered signposts and was my intellectual companion throughout this journey. I sincerely hope that this paper will reflect adequately my admiration for Stafford’s work. At first, I would like to report on some episodes of this journey as a first-order observer, according to the motto: What was (is?) the case? Secondly, I shall comment on these episodes as a second-order observer: What do I see as behind the cases through both VSM and, unavoidably, my experiences with other system theories. Finally, I shall make some proposals as results from observing my experiences with those “glasses” itself: how viable is the VSM? Should it be

improved by integrating recent theoretical achievements to make it more Observing viable? experiences with

the VSM 2. First-order observation: what was (is) the case? Stafford defined “fact” as “that which is the case” (Beer, 1979, p. 282). But what was the case when I read “Brain of the firm” (Beer, 1981) in Orwell’s year 1984? What happens, when we report on long past stories? Can that be first-order observation already being influenced by systems theory for almost 20 years? In 1984, I did not understand my introductory points. This suggests that my systems thinking in 1984 was considerably different. This difference in distinctions and indications leads of course to different interpretations of the VSM as well. To sum up: in complex situations, also the so-called “facts” are constructs (as the latin word “facere”, to make, indicates); my facts are present models of my own past. I am the observer who observes in a reflective mode the forms of observation of 1984 with the current forms. Learning requires an observer comparing the states of a system at different times; if I observe myself, I compare two snapshots of the permanent reinvention of my own re-entry. Be careful with facts; consider the potentialities! 2.1 Episode one: my personal Eureka or about the difficulties of unlearning Organizational transformation starts with individual transformation! I experienced my own personal “Eureka” purely by coincidence. I was asked to deliver a lecture at the University of Mannheim on the topic of “Management of Change” and I knew that this meant above all: change management! Naturally, I wanted to relate my experiences of many years as a senior manager to a particular theoretical framework, in order to communicate these experiences in a language that the students could relate to. I knew of course that the curricula at German departments of Economics and Business Administration were only of limited value for my lectures. Management is an activity in its own right (sui generis). Many successful managers are chemists, engineers, or lawyers or simply did not attend a university at all. After a long search, I was left with only a few books that I thought could be of some help. One of these books was “Brain of the Firm”. When I skimmed through it, I was astonished to see detailed pictures of the human brain and I was startled to find strange loops leading to a completely unknownn “anastomotic reticulum” with such confusing peculiarities as a variety of ð2n Þ2 (Beer, 1981, p. 47). My first inclination was to put this book aside. But then I discovered an exciting motto on page two: “ABSOLUTUM OBSOLETUM! If it works it’s out of date”. These sentences were my personal Eureka and my motivation to study “Brain of the Firm” in-depth.

613

K 33,3/4

614

Stafford’s book was very hard reading for me. It led me to completely unfamiliar grounds in a foreign language. It set free strong emotions and was troublesome and fascinating at the same time. I had to change a great deal of my epistemological foundations (I had studied economics), and had to write off a good deal of my previous knowledge, which like most people I regarded as a treasure. I had to replace the usual search for the “cause-effect-blame-relations” and the imputing of responsibility either top down or from left to right (and vice versa) by circularity and recursion. However, repeatedly, the routine of being bogged down in the daily interactions and colloquial business-language triggered those culturally established “ontological reflexes” such as managing the organization in the usual way as a trivial machine. Deliberate forgetting hurts. Later, when I read “The Heart of Enterprise” for the first time, I discovered Stafford’s unique ability of connecting cogent cybernetic theory with sophisticated practical insights and experiences. Quite often, I thought he had taken many examples from my company. But a most impressive aspect that Stafford brought to my attention was that cybernetics and the VSM were the science of intellectual self-defense against a culture, that stubbornly refused to consider the consequences of the world’s complexity embodied in numerous idiosyncratic realities. 2.2 Episode two: teaching at the University of Mannheim (1985-2000) However, the learning process that I wanted to initiate in my students with the VSM as a model of effective organization was even more difficult for them than it had been for me. As a manager with 20 years of experience, I could take a pragmatic attitude. In many occasions, I had an amalgam of patterns of behavior (theories in use) as a back up position, which proved more or less useful. Often students would find it difficult to behave pragmatically, because they have been “trivialized” (von Foerster), as they have to repeat what was lectured earlier. They are confronted with divergent theories, mostly without specifying their different epistemological foundations. In the fields of economics, business administration and management sciences, the trivial machine is still the dominating but albeit implicit research background. The clash of paradigms produced by my lectures led to delicate situations. Most of my students found the VSM useful, but some of my colleagues did not. As a side effect of this insufficient paradigm-mix prevailing in most economic departments of universities, I could observe an even more severe problem when young graduates joined my company. Chemists for instance, were mostly hired according to the subjects they had concentrated on during their studies (e.g. doctoral thesis). Thus, the most recent knowledge of interest was continuously imported into the company. This way a mutual learning process emerged: new scientific knowledge in exchange to technical know how

(e.g. how to run a machine) or management expertise (e.g. how to deal with Observing shifts on the shop floor). This process also accelerated social integration of experiences with young people. Graduates in economics mostly did not have this opportunity; the VSM their theoretical background was only of minor interest to practitioners. 2.3 Episode three: German society for social cybernetics (GWS) My experiences within this society were very similar to those I had at the university. Its name already reveals its program. Practitioners founded the GWS after World War II in order to capitalize from cybernetic theory. For about 10 years, I was the Deputy Chairman and member of the Program Committee that also selected the papers offered for the conferences. When I joined the GWS, the discussions were driven by first-order cybernetics: centering on terms such as complexity and feedback. Most of its members were professors of different disciplines, and as expected, again the clash of paradigms and different underlying system theories were the main cause for dissent. Since the VSM looked almost tailor made for a deep understanding of quality management, organizational learning and information systems, I made persistent efforts, often supported by Markus Schwaninger and students from St Gallen and Mannheim, to present the VSM in different application contexts. However, I am afraid that it was difficult for the GWS to see itself disseminating the ideas of VSM and second-order cybernetics against the mainstream: the status quo defends its state! 2.4 Episode four: the Hoechst project The objectives of the project, its organization, problems, resistances, progresses and results, when it formally ended, are documented comprehensively in our book “Organizational Transformation and Learning” (Espejo et al., 1996). Nevertheless, I would like to use this opportunity to make the following two comments. 2.4.1 Cyberfilter/Cybersyn as a trigger for cybernetic thinking. In the preliminary meetings before launching the project, my colleagues in senior management were not so much interested in “theoretical discussions” about cybernetic management. But they were keen on the concepts of Cyberfilter/Cybersyn (Beer, 1979, p. 499; Espejo et al., 1996, p. 261). This package ingeniously avoids the shortcomings that managers suffer from most information systems. It draws attention not only to the usual operative figures, but also to the deviations from normative and strategic references. Management is challenged to articulate the close coherence of normative, strategic and operative objectives and measures and to introduce their decisions into the communication process. Furthermore, with the recursive architecture of Cyberlogic, (it follows the logic of the VSM), this information system supports the autonomy of departments (e.g. Business Units) without concealing their responsibility. This way, one of the most annoying and costly

615

K 33,3/4

616

problems in organization, the control-autonomy-dilemma (Espejo et al., 1996, p. 114, and 296), can be mitigated. 2.4.2 Project duration. The Hoechst project lasted for 5 years compared to the average project duration of consultancy projects this seems to be a long time. But, as I see it today, the external supervision was perhaps rather too short. Organizational change by means of communication and not of command (as it is the case with the usual reengineering projects!) is a slow process. Irritations generated by consultants and closely related managers towards the organization have to change the organization’s members and the shared mental models in a way that self-organizing energies are released. Management can do without relevant external consultancy only if the capacity of appropriate internal knowledge and, more importantly, if the readiness for new patterns of behavior, communication and decisions grows and establishes themselves. 2.5 Episode five: Festschrift-Syntegration The Festschrift-Syntegration in celebration of Stafford’s 70th birthday (Espejo and Schwaninger, 1997) offered an opportunity for a team to work together under very favorable conditions. . All participants had a strong motive for a high commitment to the project: their admiration and affection for Stafford Beer. . The objective of the task was clear-cut: to produce a Festschrift in his honor. . The team itself decided both topics of different working groups and assignments for the participants to the different working groups by applying a democratic procedure (i.e. team syntegrity). . Most members of the syntegration were knowledgeable experts in systems thinking and the VSM. Thus they could communicate the complex matters in a language with adequate variety. . The communication was structured according to Stafford’s syntegration model. At the end, I was surprised with the outcomes to say the least. I shall explain the reasons for that in the next section. 3. Second-order observation: what is behind the case? How shall we ever conceive however express a new idea, if we are bound by the categorization that delivered our problem to us in the first place? (Beer, 1994, p. 8) To avoid repetitions, I shall not comment on the individual episodes, introduced in Section 2, but I will try to work out common characteristics based on recurrent observations. Most typical experiences in these episodes were fundamental shortcomings of second-order observation and therefore, in Stafford’s words, the loss of its power and logical capacity to explain (Beer, 1975, p. 71).

Any second-order observer observes a first-order observer (who may be Observing himself) trying to determine which “form” was used for first-order observation. experiences with This way he can realize the unmarked space, the blind spot of the first-order the VSM observer, who cannot see, what he has excluded, as he indicates an “object with properties” and by this very operation, he blindly distinguishes it from “something other”. 617 The second-order observer observes forms based on distinctions and looks for answers to the question: what is behind the phenomena generated by first-order observations, what explanations can I give. It does not matter, whether those distinctions are drawn for instance, on the basis of . different systems concepts, . competing economic theories, or . the so-called “evident facts” of every day life. Of course, the second-order observer has an unmarked space too, a blind spot, which could only be seen by another second-order observation. Blindness is a co-production of seeing or seeing implies blindness. But the second-order observer is aware of his blind spots and tries to avoid second-order ignorance: that is being blind of the own blindness. Second-order observers are very nice “persons” or systems, but extremely rare to find. 3.1 Biological and cultural blinders Stafford defined the fundamental obstacles to systems thinking very clearly: The brain has been programmed to deal with a vanished world. Man is a prisoner of his own thinking (Beer, 1975, p. 15)

These obstacles are the source of all the others! (1) The first quotation refers to the evolutionary limitations of our cognitive apparatus. According to theories of evolution, since the emergence of mammals it has taken over 500 million years to generate today’s capabilities of the human brain. (2) The second quotation deals with cultural limitations. Cultural and societal conditions influence this apparatus as well. They inform and deform the capabilities of judgment; they tell us what we have to regard as “logical” or “reasonable”, “good or bad”, “true or false”. We can look back at approximately 20,000 years of cultural evolution. The biological apparatus, the engine, that generates the consciousness-system (the recursive network of thoughts and feelings) is connected to social systems (networks of communication) that irritate “the self” of this consciousness, to a large extent, and this in turn the social systems. The time-ratio of 2 : 50,000 already indicates that till date, culture had only little success in civilizing the biological autopoietic beasts!

K 33,3/4

618

These two blinders generate phenomena with different names. I would like to comment on some of them, which occurred in all episodes. 3.2 The difficulties with meta-management In the VSM, management is a meta-systemic (to the operational units) and distributed function within an organization (Beer, 1979, p. 284); Management has to act as a second-order observer. As such, management has to respond with decisions to the differences of its expectations and information. It has to discover differences as symptoms (e.g. high efficiency, but poor quality, or an absenteeism ratio above average in a public office), and explanations for those symptoms. It receives conflicting interpretations for the same differently defined problem-situations from various departments and hierarchies (production, research, sales and works-council) or from external stakeholder, and it has to deal with those conflicts as well. In Stafford’s words: management “does not deal with the stuff of the system at all – it deals only with the managerial consequences of what the system does” (Beer, 1979, p. 289). But what it does is seen differently by different observers (e.g. stakeholder-dilemma). Managers are all too familiar with Murphy’s law. However, sometimes if Murphy’s law itself goes wrong, success cannot be reduced only to self-organization. The paradox of maintaining identity through change has to be resolved also by management via recursive communication. This is not at all a trivial transformation because the self-description of the organization must change. Communication has to reflect the organization’s identity, the inner model of both organization and its environment and to reinvent its re-entry. Management is able to change the patterns of communications and decisions only slowly and gradually. Cybernetics and the VSM may serve as epistemological and structural models, but they have to be translated into the idiosyncratic language of the organization in order to enrich its language and change its operations. Management has to consider the multiple structural couplings with systems outside, with other organizations, with differentiated social systems and consciousness-systems (e.g. the members of the organization or other stakeholder). In all kinds of organizations, managers have problems with their own meta-management. Within organizations it is regarded as “soft” and passive; employees do not take it seriously or do not pay respect. The public and the published opinion mostly do not like entrepreneurs and managers. Yet, following a mechanistic and trivial idea of an organization’s modus operandi, both of them expect managers to be tough and straightforward. There is a widely held belief that with good intentions, capabilities and sufficient resources, any target could be reached by a consistent course of command and control. However, this control-optimism often criminally ignores the characteristics of complex systems.

Thus, the public rewards activism as well; command looks much easier, and Observing therefore managers (also in politics) prefer direct intervention, dressed up with experiences with pleasant names: we have to take care of and give help; we are striving for the VSM risk-absorption, for public welfare, good education. (The funnel-model is still regarded as a royal road to learning and the exorcism-model the shortest way to unlearning. Disillusion with politics is another result of politician’s 619 disregarding complexity.) It is extremely difficult to change the dominant image and attitudes of managers, their “espoused theories” that they have to manage a trivial system called “organization”. I could tell numerous stories about that.

3.3 Theory and practice Managers “theories in use” have to be different. How otherwise could we explain that, for instance, many business-organizations maintain their viability quite often for a long time or that non-profit organizations are almost immortal? This means, that management has to have at least an implicit knowledge about the viability of their organization even if it is not yet sufficiently described in theory. Business-managers are forced by market-mechanisms to permanently strive for effectiveness and efficiency of their organizations. Therefore, they also look for new approaches to organizational change. Available concepts are evaluated by using the distinction “practice/theory”. One problem with organizational learning is, amongst other things, caused by the fact that only knowledge and practice approved by senior management is usually regarded to be useful. But if one asks for detailed descriptions for “practical knowledge” there may be as many different answers as managers. With good theory, it is different. The results of social sciences, their theories and knowledge-offerings, are often treated with suspicion and caution. That is not amazing! But in companies with an established culture, dominated by natural scientists (e.g. chemists) members with outstanding theoretical knowledge for instance, in quantum theory or bioinformatics have a very high reputation. It does not matter to senior-management, whether or not this kind of knowledge is used in the laboratories. Cooperation with universities is usual. It seems not very promising to convince senior managers to study system theories intensively. A preferable way could be to make apparent the implicit systemic knowledge and behavior of successful managers, by explicitly applying systems theory in projects of common interest to a company, such as the development of business strategies or improving quality management and information systems. It is not important to use scientific terminology, but to translate the underlying theoretical concepts into a language that can be understood by as many members of an organization as possible.

K 33,3/4

620

Of course, also many managers are often bound by the specific categorizations that delivered the problems (in Stafford’s words). This is one reason for the success of traditional consultants and the small market share of systemic consultancy. Even the Protestant church in Bavaria hired McKinsey as a consultant! A frequent misinterpretation of the VSM regards the recursively embedded systemic functions (Systems 1-5) as representations of the existing formal structure of an organization or institution. Thus, for example, the supervisory board of a company or the president of a country have been baptized as Systems 5. Another practical experience seems to be important for organizational transformations based on the VSM (as a design for freedom). It concerns the asymmetry between freedom and responsibility and the paradoxical effects of hierarchies. The demand and the desire for freedom are high, but the willingness to bear the risks and take the responsibilities of the own autonomy in decision-making are much smaller. Many people like to wait for instructions and to pass on responsibility bottom-up or socialize it under the magic label called “solidarity”. The chain of command and control is undesirable and desirable at the same time. On the one hand, it determines tough limitations and on the other hand, it offers safety and promotes a strong motivation for people to work their way up. Often in system-theories command and control hierarchies are regarded only as obstacles, harmful and unnecessary. However, history indicates that these hierarchies can be found in all kinds of organizations, in tribes and monasteries, churches and states, universities and enterprises. The distinction “top/down” seems to be a universal invention of social evolution. Generally, hierarchies separate (e.g. tasks), connect (e.g. communications or decisions linked to specific programs) and protect (e.g. autonomy). However, how to overcome or to mitigate the difficulties of generating shared mental models (eigenvalues) based on understanding within a network of communication, between persons of different interests, characters, professions, hierarchies (or recursion levels), sites, languages and cultures? In order to avoid the well-known filibuster-strategies in organizations (as networks), management has to execute leadership when consensus cannot be reached. In my opinion system-theory still has to resolve questions such as the relation between command (which ones?)-control-hierarchies and recursive-structures, the relation between autonomy or degrees of autonomy and the modus operandi of self-referential systems and the very job of meta-management as a second-order observer. (For instance, is management’s intervention into autopoietic systems possible or is it a kind of sabotage?)

3.4 Pathologies of viability Observing This term used by Stafford (Espejo and Harnden, 1989, p. 26) points at the experiences with danger to viability when important systemic insights or systems properties are the VSM not recognized by an organization. An empirical view on every day life confirms that this systemic disease is epidemic. However, there are no objectively given combinations of “problems and 621 solutions” or cogent indicators of viability or pathology. One observer (e.g. a general manager of a division) may use the VSM for analysis and detect some symptoms in his organization, pointing to viability (e.g. a well-equipped and highly motivated research department as part of System 4). Another observer (e.g. a shareholder or a member of the works-council) of the same organization may reach at quite different conclusions (e.g. to cut down research expenses in favor of higher dividends or wages). In many cases, second-order observation is possible only in a very limited and modest sense, because another observer’s motives for distinctions or theories used for indicating problems or solutions, are not at all transparent. This lack of mutual understanding of the first-order distinctions often leads to a paralyzing communication of speechlessness (paralysis by analysis, filibuster). At this point, we arrive at a largely unsolved problem in systems theory: the complementarity of operational closure and structural coupling. On the one hand, systems are the mutual conditions of their separate existence and on the other hand, different systems cannot determine each other’s autopoietic production and reproduction. The evolutionary link as a special relation necessary for the survival of systems is called “structural coupling”. The most important structural coupling connects consciousness-systems and social systems. Only consciousness-systems do have the sensory apparatus needed for perception. Social systems cannot hear or see, they cannot be irritated by sound or heat, but they are open for irritations (thoughts or feelings) from consciousness-systems, introduced into communication. The mechanism for those irritations is language (Luhmann, 2002, p. 270). Structural coupling exists between consciousness-systems (“individuals”) and social systems and not directly between individuals. The only reason, why social systems come into existence is the inability of people to read each other’s minds. As an example and referring to management, we can say, management should be a second-order observer of communication that is permanently irritated by consciousness-systems, which in turn are both promoted and contaminated by communications (e.g. motivation, fluctuation). An important core competence of management is the comprehension and the consideration of the delicate requirements of the structural coupling between consciousness-systems and their organization.

K 33,3/4

622

But how can these operationally closed, self-referential systems “understand” their mutual interactions despite of their idiosyncratic rationalities, experiences, knowledge and languages? Let me shortly explain the precarious conditions of “understanding” by referring to Luhmann’s concept of communication (Fuchs, 1993; Luhmann, 1984, p. 191, 1997, p. 81). There are two different notions of understanding. They are as follows. (1) Operational understanding is only the third selection of the addressee (besides information and its utterance), which generates a basic communication element and keeps communications (and thus the autopoiesis of a social system) going. (2) Interpretative understanding is the self-referential attaching of meaning of a System X to the actions of another System Z. It requires a double re-entry. . The first re-entry is a “copy” of the System X generating distinction of “system/environment” into itself (i.e. the System X). Both sides of this re-entry can be observed internally with the distinction “self-reference/outside-reference”. . A second re-entry (also constructed internally) has to be copied into the outside-reference, modeling the self-reference and outside-reference of the System Z, which actions should be understood or interpreted by System X. The observation of System Z’s self-reference, as modeled in the second re-entry, is the only source of understanding for an operationally closed self-referential system. Interpretative understanding between social systems is an iterative outcome of mutual “other-references” towards a specific degree of convergence. This short description of the very demanding conditions for “understanding” hopefully makes clear, why interpretative understanding is so improbable, why dissent is the rule and why trust is a precondition for social communication. It is exactly at this problem in interpretative understanding that the painful reductionist interfaces work almost naturally. Operationally closed self-referential systems are only able to see others (e.g. systems, persons, problems, theories or values) from within; that means through their internal theories and models of themselves and through their internal theories and models of their world outside. In short: through their interest driven forms (distinctions) and their descriptions in their “own language”. Unfortunately, these communication difficulties are not only true for everyday situations, which mostly iron out and camouflage the specific motives for distinctions.

Scientific theories (as methodologically controlled distinctions and Observing indications) only design a specific reality and provide instruments to observe experiences with this reality. In this respect they are self-referential: if used for second-order the VSM observation, they would find themselves and produce solutions for self-defined problems. A scientifically sound comparison between theories can hardly be found, 623 most probably due to the lack of a theory of comparison, which again needs a meta-theoretic fundament, finally ending in the paradox of the beginning. Quite often, theories are used to reflect a systems identity (the marked space of the re-entry); for instance, a macro-economic theory could describe the identity of the economic system. (Luhmann, 1997, p. 958ff.). But what will happen, if for instance the political system uses a conflicting theory to indicate its other-reference (e.g. Hayek vs Marx)? For a second-order observer it is extremely difficult to capitalize on the multiple views of different theories (e.g. different descriptions of problems and proposals for solutions) by taking a meta-view on their different paradigms, historical elements, ideological colors, premises and architectural features. If a deconstruction (reconstruction) of theories is not possible or simply it is not done, as it is often the case, then only normative integrators as cybern-ethical distinctions will be left: trust and tolerance. Both are paradoxical attitudes that may reduce conflicts: trust presupposes trust and tolerance must not tolerate intolerance. One thing is for sure: whatever distinctions an observer uses to compare and evaluate theories, he has to be aware of their blind spots. However, I would like to continue in a less abstract way and give examples of the main source while studying the pathologies of viability: in many disputes, among others also with system scientists, the need for second-order observation was often ignored. Different concepts of systems theory. It is natural for social complexity to generate different systems concepts. For instance, both Beer and Luhmann regard social systems as autopoietic, but the basic elements of those systems for each of them are different (people, roles, thoughts and feelings or instants of communication?). My point is that also a discussion of different systems concepts may end up in confusion, because second-order observation is left out, for ever reasons what so ever. For instance, in one of the Festschrift’s teams we could not come to an agreement on the unfolding of complexity because of different definitions of basic elements. We discussed questions as: How can people with hair and flesh, body and soul, be at the same time products and production process (that is autopoietic) of different social systems? As an advocatus diaboli, I took Luhmann’s view that differentiated social systems, for instance the economic system, operating with a specific code, does not have a social address. Other team members preferred to blame the economic for societal ills. I replied, “you are blaming yourself, because, by using its code .

K 33,3/4

624

every day via your communications you are also producing elements of this system”. (This is the concept of inclusion: everybody may link up with different social systems by adopting their specific distinctions.) Organizations do have a social address, but the economic system is not the sum of business organizations. In Germany, the Catholic and Protestant church together employ about 1.4 million people; the chemical industry in Germany has about 500,000 employees. Of which social system is the organization “church” a part? To which system can social phenomena (e.g. 6 million unemployed people or 40,000 bankruptcies per year in Germany) be imputed in the case of multiple structural couplings (e.g. economy, politics, science, education)? Who could be criticized within an organization by an external observer if something went wrong? To whom should this observer write a letter? To the “recursive network of communication” or the president of an organization (or institution?), the top of a hierarchy, which many systems people do not regard as necessary? Of what kind are the relations between an institution and its organization? All those questions are of enormous practical importance. Different Systems theories answer them very differently. Social complexity remains contingent also for systems thinking, but at least consensus about dissent should be reached without conflict. Competition between different economic sciences and ideologies. Besides cybernetic explanations there are a multitude of Social theories and ideologies with controversial paradigms. For instance, in economics many different concepts compete, representing parts of a puzzle that cannot be assembled as the picture of a coherent theory. The conflicting constructions of different “economic realities” quite often result in destructive and sometimes devastating effects. What would happen if symptoms such as poverty, hunger, millions of unemployed people and breakdowns of social security systems have to be managed as interrelated symptoms? Economic theories generate so many different and conflicting definitions of problems and solutions that “anything goes” until “rien ne vas plus” (Ortmann, 2003, p. 249) and political actions are paralyzed. But from where does the political system receive (via inner models of the “world” outside) superior knowledge for its interventions? Can the political system rely on such fuzzy criteria as “bonum commune” or “will of the people” and who has the authority for their interpretation in a recursive mode? Those questions were also discussed by one of the Festschrift teams, without agreement. Some participants assigned a dominant role to the state (as a model of the political system). But what happens with autonomy? How can in cases of conflicting diagnosis a “synoptic view” be achieved, as required by Stafford Beer for System 3 operations (Beer, 1979. p. 205) and how

can problems of interpretative understanding and diverging interests be Observing managed? experiences with Economic theories and Systems theories only rarely acknowledge each the VSM other. (Baecker, 1988, is an excellent exception). However, severe global problems intertwined by the structural couplings of differentiated global and local social systems urge for the second-order synopsis and wait for answers. 625 Every day we experience (as observers!) numerous negative phenomena as the results of mutual conflicts of interpretative understanding between the individuals, organizations and different social systems (including religions which, by definition, observe with different distinctions the principally unobservable and indicate it in a dogmatic precise manner). How shall we mitigate those phenomena as long as we are not even in a position to explain them in theory? These are a lot of challenging projects for universities where is necessary expertise in both fields. The ascension from second-order ignorance to moralizing. As I already mentioned, second-order ignorance implies blindness to own blindness. In one of his examples about “The pathology of the viable system”, Stafford describes how System 5 collapses “into Three (in the effective absence of Four) . . . thus when something goes wrong in System Three (or even One) they are likely to dive down into the problem that they understand so well – never to emerge again”. (Beer, in: Espejo and Harnden, 1989, p. 29). I call this process efficiency-paranoia. Quite often the reverse case might be encountered as well: the moralizing-paranoia. It is not the case of “demi-gods” (Beer) diving down to the mundane affairs of Systems 1 and 3. On the contrary: it is the ascension from the opaque or white spots of knowing differently or not knowing at all to the heaven of the arrogant “knowing – all manners” by moralizing. It is the tempting flight to the “oughts”, the way up of people in the Systems 1 and 3 to Systems 4 and 5. I experienced moralizing as the most frequent conflict in complex situations. Phenomena which cannot be explained because of insufficient knowledge and which do not fit in ones world view (e.g. theories of all kind) are described as unfair or unjust and combined with the demands such as “You must” (not: “I must”!). Thus, complexity is quickly reduced to precise causality; culprits are discovered and morally qualified. Moralizing constantly carries contempt for others. Therefore, communication based on moralizing arguments breaks down, destroys dialogues, trust, tolerance and thus social relations and necessary actions. Hardly ever those debates suggest good agreement with systems thinking. A second-order observer would immediately ask whether or not the difference “moral/immoral”, as it was used in a first-order observation, if applied to itself, is moral or immoral, useful or not. Without this “second thought” many

K 33,3/4

626

important questions, which are typical for complex situations, remain without answers. . Do good intentions already produce good results (“moral causality”) or can we expect bad consequences also probable? (The world is crowded with dreadful results of this “logic of the failure”.) . Do the selfish actions of individuals lead to beneficial outcomes for the society (Adam Smith)? . Which consequences are to be expected by omission, that is, by making no decision (in cases of “you must not”)? . Will the “bridge-principle” be violated? It says: that normative “ought” (System 5) imply adequate strategic (System 4) and operative (System 3 and 1) knowledge, abilities and resources. Nobody can be obliged beyond capabilities (ultra posse nemo obligatur!). Values should offer orientation for action; they should be premises for the decisions of the many without being decided themselves by a specific authority. But those values themselves quite often lead to severe conflicts. This may even be the case when the rules for debating values have been already laid down (e.g. syntegrity model) and are not themselves the subject of lets say “fair and just” discussions. Again, the “ethics-group” in the Festschrift meeting was an excellent, but harmless example. . Is there a “bonanza of eternal values” which has to be imported or imposed to a system? That would be the end of autonomy and freedom! Remember Stafford’s “absolutum obsoletum”. . If values are eigenvalues of recurrent internal communication, they reflect the specific world view, preference rules, rationality and language of that system. But how to deal with the System 5 of organizations at different levels of recursion in an organization or with different social systems (e.g. the economic and political systems)? . This question leads to the paradox of values. Instead of giving orientation for viability, values may lead to dangerous conflicts. This is the case, because in a situation of growing differentiation of systems, normative functions will more and more become incomprehensible for different systems to each other. But on the other hand, values are needed to provide an integrative orientation without command. In these situations there is a great danger of “empty consensus” (C. F. v. Weizsa¨cker). One can quickly come to an agreement on abstract terms. Everybody likes justice, peace, fairness and solidarity in societies, or high quality, motivation, good social climate and fair remunerations in companies.

However, as soon as you look for interpretative understanding and, Observing following a deductive logic based on the VSM, for recursive interpretations of experiences with those abstractions, you observe contradictions of the values, conflicts arise and the VSM overall social breakdowns are most probable. At present, I do not see any solution for the dilemma of self-referential autonomy and structural coupling (autonomy vs autodestruction)! 627 I mentioned that the “Festschrift-Syntegration” was my most surprising experience with the VSM. I was convinced to come across a “laboratory-situation of systems thinking”. However, I was more than surprised to realize that the results of the conference did not meet my expectations. I was disappointed: from (my own) illusions to delusion. I had expected that the discussions would take a completely different course compared to those numerous meetings I had attended before. This was my error! The title of Stafford’s book on syntegration is “Beyond Dispute”! It was this “beyond” that I really missed. Quite often, the discussions lack the very point of systems thinking: second-order observation, the observation of the forms of our distinctions. Only by being able to observe those forms we could have seen how a participant had arrived at his/her information, explanations, assessments, judgments or verdicts. It is not necessary to use a sophisticated procedure (syntegration) just to collect those matters we have in common. But if the concern is really to take a second-order observer’s view, then we cannot do with statements such as true or false, moral or immoral, capitalistic or socialistic, VSM or what else, or to respond outraged by answers that does not suit. Communication always requires second-order observation. Otherwise interpretative understanding via dialogue is not possible; only parallel monologues, talking at cross purposes may be achieved. (Unfortunately, quite often that is the very target of conversations, only to camouflage the motives for distinctions.) To sum up this very valuable syntegration-experience: . .

.

The team showed many typical symptoms of pathologies of viability. The syntegrity-process cannot iron out conflicts as long as the underlying cybernetic conditions relevant to the participants are not considered. Despite the participant’s differences in the explanations and evaluations of complex phenomena, the interactions over only a few days were sufficient for mutual positive feelings and a good social climate. This climate could have been an optimal situation to start with a proper communication process! But we got stuck only in another paradox of teamwork: when the teams arrive at these preconditions for communication most of the time their work is finished.

K 33,3/4

628

4. Insights and hope Stafford wanted to offer “insights and hope” with his work (Beer, 1975, p. 456). Let me explain why I am still so grateful to his contributions. 4.1 First contribution: insight For a couple of years I have been retired as a manager. Now I am also a professor emeritus. What does systems thinking and the VSM mean for a pensioner? Well, they are incorporated into my feelings, thinking and actions; they are part of my identity. As a first-order observer, I am mainly confronted by “realities” which are produced and conveyed by media. These “realities” are the symbiotic outcome of power and interests. The VSM is a very powerful model for a second-order observer, and I take the opportunity for this kind of observation whenever possible. It is only from this “paradise” of second-order observations that we can capitalize from the second thought and can start changing something, beginning with us. Sometimes of course, I also reflect on Stafford’s 30 year old theory: How viable is the VSM itself? Can it be improved by integrating recent theoretical achievements? As a practitioner, I do not feel competent enough to give answers to this question. But at least I would like to refer to some of those areas of the VSM where I think clarifications and further development could be useful. If you have a look to some quotations that I collected (not very systematically) from Stafford’s books (Appendix: Essential quotations), you would find the most important components also from recent theory. But these components are not always elaborated in his books or consistently considered in their consequences. Some examples are as follows. (1) Already in 1975, Stafford had realized the necessity of a restatement of mathematics ab ovo with concepts close to the “Laws of Form” of George Spencer Brown. Not many systems scientists have integrated Spencer Brown’s self-referential proto-logic into the architecture of their theories (most rigorously Luhmann and Baecker; Baecker, 1993a, b; 2002; Luhmann, 1997). But there is almost no further development of the calculus itself (for extensions and references, Schuhmann, 1998). This kind of trans-classical protologic of forms could be very helpful to clarify the leap from systems as ontological units to “units of a difference”, with the characteristics of operational closure, a self-referential modus operandi and the consequences of these characteristics. (2) The VSM lacks consistent definitions of its basic elements (people or roles or thoughts or communication) as well as the definitions of basic terms such as facts, information, communication, interactions within the system and with its environment, which are in line with the epistemology of second-order

cybernetics and the definition of systems. Is the VSM a model for all kinds of Observing systems (consciousness-systems, organizations, differentiated social systems)? experiences with (3) From these basic elements and their operations, the five functions of the the VSM VSM could be developed. It has to be clear that these functions are no entities that are established from the outside of the organization (a frequent misinterpretation). The functions, for instance could be explained as 629 recursively connected eigenvalues (attractors) within a network of communication dealing with a specific program (or specific tasks) and simultaneously considers normative, strategic and operative aspects that are necessary for decisions that maintain the viability of an organization. (4) Are the viable systems within viable systems autopoietic and self-referential ones? Can this recursive nature of the VSM and the unfolding of complexity be interpreted with the concept of structural coupling? Then, for instance, two neighboring units of recursion would they be mutual outsides of each other? Which conclusions have to be drawn for concepts such as identity (reflection on the form within a form), organizational learning influenced by management and learning as a result of organizational evolution and co-evolution? (5) Are the basic concepts developed in “Brain of the Firm” still in line with the results of recent brain research? I am aware that I have more questions than answers, and I am also aware that for some questions satisfactory answers are not available at all in systems sciences. However, I am still convinced that the way Stafford developed and presented management cybernetics is a unique combination of science embedded into sophisticated practical examples and case studies, written in beautiful language and enriched with expressive drawings. It deserves any effort for dissemination and therefore the need of an upgrade. 4.2 Second contribution: hope Stafford said: “I don’t know how the world should be run” (Beer, 1975, p. 456). Nobody knows! Nevertheless, we know that complexity is a problem and solution at the same time. This is why systems theory offers hope: it deals with social complexity in a rigorous way. Applying the “science of communication” to itself more often, could substantially foster the “wealth of communication” that is based on the variety of contingent distinctions. According to his motives any observer is free to select his own distinctions, as long as he consistently regards cybern-ethical conditions. But to explain social complexity is one thing, to change the social world is another. Probably one human life is too short; fundamental change will take generations. Social theories are worse off compared for instance to quantum theory, that first tried (and still tries) to explain matter, and then developed some useful

K 33,3/4

630

products. But even after 100 years, the revolutionary worldview of this very successful theory has not influenced the views of people. I think that systems theories are far behind the scientific level of quantum theory or brain theory. This is of course also a result of the only loosely coupled efforts by the different schools of thought and the relatively small amounts of funds spent in social-systems-research. (Amazing small in the light of all those social problems in our world!). Nevertheless, I doubt that we will ever have one grand social theory (“a great big theory”, Beer, 1975, p. 456). The acceptance of theories cannot be enforced. But I am convinced, whatever social theories will look like in the future. Science is only a necessary, but alone not a sufficient condition for social change. Neurobiologists tell us that the limbic system of our brain, the origin of emotions, is the unconscious censor of our consciousness, of our thinking and actions. So it is important to stress the point, that “systems-feelings”, Stafford’s concept of algedonic information (Beer, 1979, p. 406) are the necessary supplement for systems thinking! Logos and eros have to be reconciled. We enjoy a beautiful sunset in spite of knowing, that the earth is turning around the sun. The titles of Stafford’s books underline his complementary program of this togetherness: “Brain plus Heart” plus “Beyond Dispute” generate “a Platform for Change” which supports the “Design of Freedom” with “Viable Systems” as results. Stafford showed us a never ending way to find human solutions for men-made problems. From now on we have to walk alone but we must go on. References Baecker, D. (1988), Information und Risiko in der Marktwirtschaft, Suhrkamp, Frankfurt am Main. Baecker, D. (Ed.) (1993a), Kalku¨l der Form, Suhrkamp, Frankfurt am Main. Baecker, D. (1993b), Die Form des Unternehmens, Suhrkamp, Frankfurt am Main. Baecker, D. (2002), Wozu Systeme?, Kadmos, Berlin. Beer, S. (1975), Platform for Change, Wiley, Chichester. Beer, S. (1979), The Heart of Enterprise, Wiley, Chichester. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester. Beer, S. (1985), Diagnosing the System for Organisations, Wiley, Chichester. Beer, S. (1994), Beyond Dispute: The Invention of Team Syntegrity, Wiley, Chichester. Espejo, R. and Harnden, R. (Eds) (1989), The Viable System Model, Wiley, Chichester. Espejo, R. and Schwaninger, M. (1997), To be and not to be – That is the System, CD-ROM, Carl-Auer-Systeme Verlag, Heidelberg. Espejo, R., Schuhmann, W., Schwaninger, M. and Billelov, V. (1996), Organizational Transformation and Learning, Wiley, Chichester.

Foerster, H. von (1993a), Wissen und Gewissen, Frankfurt am Main, Suhrkamp. Foerster, H. von (1993b), KybernEthik, Merve, Berlin. Foerster, H. von and Bro¨cker, M. (2002), Teil der Welt, Carl-Auer-Systeme Verlag, Heidelberg. Fuchs, P. (1993), Moderne Kommunikation, Suhrkamp, Frankfurt am Main. Fuchs, P. (1995), Die Umschrift, Suhrkamp, Frankfurt am Main. Luhmann, N. (1984), Soziale Systeme, Frankfurt am Main, Suhrkamp. Luhmann, N. (1997), Die Gesellschaft der Gesellschaft, Frankfurt am Main, Suhrkamp. Luhmann, N. (2002), Einfu¨hrung in die Systemtheorie, Carl-Auer-Systeme Verlag, Heidelberg. Ortmann, G. (2003), Organisation und Welterschließung, Westdeutscher Verlag, Wiesbaden. Schuhmann, G. (1998), Mathematische Untersuchungen u¨ber die Form sozialer Kommunikation, Shaker, Aachen. Spencer Brown, G. (1997), Laws of Form, Bohmeier, Lu¨beck. Appendix. Essential quotations The Heart of Enterprise (Beer, 1979) “The whole is simply an arbitrarily defined chunk of an infinite recursion. . .” (p. 403). “. . . each level of recursion is likely to answer the identity question differently” (p. 403). “. . . viable organizations produce themselves” (p. 405). “The enterprise, that arbitrary whole, produces itself too. That is to say that its staff may come and go . . . and still it has and retains its identity” (p. 405). “In cybernetic terminology, this enterprise is called autopoietic. . . . It is in the business of preserving its own organization” (p. 405). “. . . any would-be regulator must contain a model of that which needs to be regulated, and that if this model has a given level of variety, it will handle only that equivalent variety in the outside world” (p. 235). Diagnosing the System for Organizations (Beer, 1985) “System Four is not only concerned to manage the outside-and-then, but to provide self-awareness for the System-in-focus” (p. 115). “No doubt you captured the idea of the model of its own activity’s being embodied in a model of the whole System-in focus . . . it is just this infinite regression of self-images that seems to hold the key to the characteristic of self-awareness of viable systems” (p. 116). Beyond Dispute (Beer, 1994) “Closure is a prerequisite of self-reference; two others are: recursion and iteration” (p. 210). “. . . certain stable values emerge from infinitely recursive computations. They produce themselves” (p. 215). “The brain deals with the ultrastable homeostasis of its own internal states; and the basis characteristics of neurophysiological activity to this end are logical closure, recursive iteration, and reverberation that computes the eigenvalues of its own processes” (p. 244). Platform for Change (Beer, 1975) “. . . restatement ab ovo of mathematics adequate to deal with these concepts. It would be close to Spencer Brown’s Laws of Form” (p. 415).

Observing experiences with the VSM 631

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Viability versus tribalism Trevor Hilder Cavendish Software Ltd, Trowbridge, Wilts, England

632

Keywords Cybernetics, Organizations, Social behaviour Abstract The viability of an organisation can be compromised by the tribal behaviour of its members. If this is not understood, attempts to reform it will be met with resistance, since the effort will be perceived as an attack against the norms of the organisation. The paper offers examples and evidence for this, suggests source materials for understanding the phenomenon, and practical techniques for overcoming it.

Introduction Stafford Beer dedicated his life to the application of cybernetic theory to the practical task of making human organisations work better. He was passionate about making the world a better place, and saw clearly that many of our organisations, despite the best of intentions, cause significant damage, sometimes to their own staff, sometimes to their customers, and often to the environment in which they operate. Stafford developed the viable system model as a practical guide to what it takes for an organisation to remain viable in its environment, and went on to develop team syntegrity as a method to aid the solution of the problems of the System 3 to System 4 homeostat, or, to put it in layman’s terms, how to make the decisions which will enable organisations to evolve intelligently to cope with an ever-faster changing environment. As a practitioner in the application of Information and Communications Technology (ICT), I have found Stafford’s ideas immensely valuable in helping me to design and implement information systems, which support a wide variety of working organisations. However, I have also observed that the application of his ideas is often difficult, if not positively dangerous! Having “had my fingers burned” on numerous occasions, and recognising that Stafford and his co-workers have also suffered while attempting to make the world a better place, I have hunted around for ideas which might help us to understand what the dangers of this kind of work are, and how we can overcome them. This paper is a brief introduction to some of the ideas, which I have found valuable in developing requisite variety when engaged in what Stafford referred to as cybernetic interventions.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 632-646 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523616

The argument First, I will summarise the argument I want to put forward. (1) Cybernetic interventions often do not lead to their intended outcomes. (2) A major cause of this is that human tribalism drives behaviour, which runs counter to the needs of viability.

(3) This can be overcome by: . studying the tribal groups within the organisation in-depth, and . concentrating on improving real-world tasks, instead of trying to transform the whole organisation in accordance with a grand plan. (4) Success is possible only if we can inoculate our own teams against the disease of tribalism I will now present supporting evidence for these assertions, and suggest sources of information, which I have found useful in this work. Cybernetic interventions often do not lead to their intended outcomes The evidence for this assertion is, unfortunately, not hard to find. The Chilean experience Stafford Beer wrote extensively about what happened in Chile, leading up to the death of President Salvador Allende on 11 September 1973, and the takeover of the country by General Augusto Pinochet (Beer, 1981). The experience was so devastating that Stafford felt that his life’s work had been left in ruins. The fall-out from this disaster continues. Only recently it was reported that members of Pinochet’s junta had signed a statement admitting that, in 1978, the military carried out a mass exhumation program, digging up bodies of executed prisoners and dumping them in the Pacific Ocean, to prevent their discovery (The Week, 2003). The same paper estimates that, between 1973 and 1990, more than 3,500 dissidents “disappeared” in Chile. Some of us might have experienced puzzlement when Baroness Thatcher thanked General Pinochet for “bringing democracy to Chile” (BBC, 1999). This remark illustrates the fact that events in Chile were, and continue to be, perceived in dramatically different ways by different groups of people. Allende was democratically elected to power, and his supporters considered their actions to have a mandate from the people. Unfortunately, many members of the property-owning classes in Chile, and their overseas supporters, perceived Allende’s regime to be engaged in stealing their assets. They reacted to this perceived threat by undermining and eventually destroying the regime. This difference in perception is typical of the division between different social groups which I describe in this paper as symptomatic of tribalism. Experience of whistleblowers Numerous instances show that whistleblowers lose their jobs. They might believe that they are acting in the best interests of the organisations they work for, and sometimes they are fortunate enough to obtain compensation for their efforts, but, nevertheless, they always get fired (Livingston, 1990).

Viability versus tribalism

633

K 33,3/4

634

There are so many examples of this, ranging from US tobacco companies to the European Commission, that anyone who reads the daily news is familiar with this phenomenon. This is another example of tribalism in operation. Even though the whistleblower is acting in accordance with the overtly expressed aims and values of the organisation, his actions are judged as a betrayal of the social group, and he is accordingly cast out of it. My experience With nearly 30 years of experience of ICT projects for organisations ranging from major retail groups to small businesses, I have often found that the people who care most about the integrity, which they bring to their work are passed over for promotion, while the apparent “time-servers” effortlessly rise through the ranks. In many organisations, conformity to social norms is much more highly valued than the ability to solve problems. History is littered with examples of innovators who have been punished for coming up with new ideas, which have eventually proved to be the salvation of their industries (Livingston, 1990, Chapter 12, The enemies of problem solving). Your experience I invite you to reflect on your own experience in this area. . Have you ever puzzled over the extent to which your ability to get your job done is undermined by the latest round of reorganisation, which is supposed to improve the “efficiency” of the organisation? . How many organisations have spent large amount of money on re-engineering their business processes, and ended up no better, or worse off, than they were before? Conclusion: improving organisations can be dangerous! Theoretically, our modern society applauds and encourages innovation, and is keen to apply it to the improvement of the productivity of its organisations. However, the evidence points in the opposite direction. Those who dare to apply new ideas to the improvement of our human lot can find themselves greeted with hostility, which can even threaten their wealth, health and life. This phenomenon is not confined to the application of cybernetics. It arises in any situation where attempts are made to bring about changes, which challenge the apparent vested interests of powerful groups of people. The rest of this paper will examine the psychosocial roots of this problem. Tribalism Until about 12,000 years back, all humans were hunter-gatherers, living in small closely related groups. Current research suggests that hominids have been on

this planet for well over 2 million years, although “modern” humans seem to have appeared as recently as 30,000 years back [1]. The first known civilisations, with literate elites, royal authority and large cities, only appeared between 3,200 BC (Sumer in modern-day Iraq) and 3,000 BC (first dynasty Egypt). There is, therefore, no doubt that a large part of the evolution of our minds took place under circumstances when our ancestors were living in a small, tribal groups. It should therefore be no surprise that tribalism is a major factor in human thinking and behaviour. To summarise, tribes have the following key characteristics. . Authority figures, who exemplify the values of the group. . Followers, who accept those values as “facts”. . The group is glued together by attention exchange between the members. . Tribal members are “people” – outsiders are considered to be subhuman You might feel that this is incredibly primitive, so it has nothing to do with sophisticated, modern people, like us, with our complex high-tech organisations. Unfortunately, the evidence shows that the above factors are endemic in every human community, and their influence is as powerful at present as it has ever been. Stanley Milgram’s research Stanley Milgram conducted a series of famous experiments in the 1960s at Yale University, which he published in his book Obedience to Authority in 1974 (Milgram, 1997). These experiments were ingeniously designed to discover to what extent a cross-section of members of the public would obey an authority figure, even when they believed that their obedience was causing pain and suffering to another person. Milgram was astonished to discover that 65 per cent of volunteers would inflict suffering under the authority of a “scientist” on a person they believed to be another volunteer, even though that person was banging on a wall, demanding to be released from the pain of increasingly severe electrocution. Even when the subject could clearly be heard screaming in agony, the number of volunteers who would continue to inflict pain, provided that the “scientist” took responsibility, was 30 per cent. These experiments have been conducted in several different countries, across a range of professional classes, with broadly similar results. Even the people who disobeyed found it difficult to do so, even though they would only be in contact with the “scientist” for 1 h, and he had no power to punish or reward them. This implies that the tendency to blind obedience must be even more prevalent in contexts where an authority figure has real economic, political or physical power over his subordinates. Milgram’s book focuses on the subject of obedience, but he also cites the work of Asch on conformity [2]. This shows that an individual will suppress his

Viability versus tribalism

635

K 33,3/4

636

own judgement of reality in order to conform to peer group pressure. As Milgram points out, the two mechanisms differ, in that obedience to authority is usually consciously recognised by the individual, his peers and the authority figure, whereas conformity to peer group pressure is usually unconscious to all parties. However, a moment’s thought enables us to recognise that the two mechanisms are interrelated, wherever there is a peer group. This is because all peer groups operate under the authority of a superior, even if that superior is not physically present, and even when he or she is dead, is a figure of myth or legend, or is an abstract entity. For example, all cultures have heroes who exemplify their value system (e.g. Winston Churchill for Great Britain, George Washington for the United States). These figures continue to exert a powerful influence many years after their deaths. In fact, the influence of an authority figure often reaches its peak many years after his or her demise. Probably, the greatest example of this phenomenon is the figure we know as Jesus of Nazareth [3]. In their time, such authority figures are invariably radical figures, since they are the founders of a new way of thinking. However, it is their invariable and unfortunate fate to be perceived by their later followers as authority figures who demand unthinking obedience. In this respect, it could be said that followers are the worst enemies of those whom they claim to follow. Milgram conducted experiments, which showed that an individual can more easily defy authority in concert with fellow, dissident peers. With more sinister consequences, he was also able to demonstrate that obedience to authority increased where there was a chain of cause and effect, in which each of the several persons only took a small part in a series of events leading to the oppression of a victim. It is, of course, much easier to kill people by remote control, using, for example, cruise missiles or bombs, than it is to do so face-to-face. This is particularly so where these actions can be justified under a socially acceptable rubric, such as “saving the Free World from Communism”, or “defending the Revolution against Reactionary Forces”. The level of obedience to authority increases, under these circumstances, from 65 per cent to close to 100 per cent. Tribes in the workplace Livingston (1990) has pointed out that all human work takes place in what he calls a Primary Working Group. This is a group of peers, which exhibits the following characteristics: . individuals driven by peer approval/disapproval, and . the group must demonstrate obedience to an authority figure. This is indistinguishable, psychologically, from a tribe, although it is considered impolite to say so. Of course, unlike the tribal groups typical of the

hunting-gathering life-style, these primary working groups are nested within authority hierarchies, and are replicated horizontally, usually in order to cover a wide geographical area (e.g. branches of the MacDonald’s fast-food chain). This introduces another key characteristic of the primary working groups: . members of such a group know very little about the groups above and below them in the hierarchy, and are often ignorant of other groups at the same level. Even in a company with no more than 60 personnel, I have been surprised to discover just how little understanding there is at board level of the day-to-day activities of operational staff, and vice versa. Even more remarkable to me is the lack of interest that members of the board of directors have in this subject, even when I have submitted a detailed report on it. One company of this size which I have worked with appointed a new financial director. Several months after he had started work there, he had still not set foot in a number of areas of the building where activities key to the day-to-day business of the company were being conducted. He had also not looked at the computer software which we had written to track the company’s key activities, even though he was in charge of ICT. This surprising fact suggests that obedience to the next level authority above, and the approbation of peers close at hand, are such key drivers that most other concerns pale into insignificance, and are generally of very little interest. This state of affairs can be very dangerous indeed, as several recent examples of corporate collapse should teach us. Corporate dysfunction The precipitate fall in stock market valuations in the year 2000 was the trigger for a series of spectacular collapses of large US corporations, of which Enron and WorldCom were the most conspicuous examples. In both these cases, the following characteristics were revealed (Jeter, 2003; McLean and Elkind, 2003). . The board of directors had become a self-serving elite. . They had surrounded themselves with flunkies who reinforced their delusions about themselves. . They neither knew nor cared about delivering value to customers. These are the characteristics of a tribal elite which has lost touch with reality, because it has systematically cut itself off from the information sources which could have maintained the viability of the organisation. Its sole preoccupation had become to keep the share price high, by fair means or foul. Its concern was to appear all-powerful in the eyes of its peer group of fellow senior executives and stock market analysts. There is nothing new about this. The same pattern can be observed in the evolution of every dominant elite throughout the history of human civilisation.

Viability versus tribalism

637

K 33,3/4

638

If we want to be able to understand recent history, we need to examine patterns, which were successfully identified centuries back. Lessons of history The recorded history of every civilisation, from one perspective, presents a rather depressing cycle of the rise and fall of a series of ruling dynasties. This is partly accounted for by the fact that most historical records were commissioned as works of propaganda by whichever dynasty was currently in the ascendant, but it also reflects objective truth. To my knowledge, the first person to attempt a systematic understanding of the laws which govern this phenomenon was Ibn Khaldun, a Sufi thinker who was born in Tunis, North Africa, on 27 May 1332 AD and died on 17 March 1406 AD. Ibn Khaldun As is typical of the Sufi tradition, Ibn Khaldun’s study of history was grounded in his experience of the practical politics of 14th century North Africa and Spain. He came from an aristocratic family, which had enjoyed great prominence for several centuries in Moorish Spain, and had moved to North-West Africa just before the fall of Seville to the Christians in 1248 AD. He entered politics at the age of 20, and led a very active life as an adviser and ambassador to various ruling dynasties. His fortunes rose and fell with those of the dynasties he served, as a result of which he spent nearly 2 years in prison, as well as finding time for the pilgrimage to Mecca, writing a universal history of the world, and lecturing to large numbers of students about his findings on that subject, presented within the framework of the world’s first sociological theory. His most famous work is the Muqaddimah (Ibn Khaldun, 1978), which consists of the Preface and Book 1 of his universal history (Kitab al-‘Ibar). This is the work which inspired Arnold Toynbee to dedicate his life to writing his own Study of History (Toynbee and Somervell, 1987), probably the 20th century’s finest attempt at a universal history of civilisation. Ibn Khaldun explains the rise and fall of dynasties in the following terms. . Every social group is animated by a factor which causes its members to co-operate with each other. In Arabic, he calls this ‘asabiyah, which is translated into English variously as “solidarity”, “group feeling” or “group consciousness”. . Some social groups have this factor more than others. Those which have more of this ‘asabiyah are able to dominate those with less, as a result of which their leading figures are able to find a dynasty. . Ibn Khaldun, in accordance with the assumptions of his time and place, associates a dynasty with a political state system. In fact, he uses the

.

.

.

identical Arabic term dawlah to refer to both. When a dynasty rises, its authority creates a state, which leads to the growth of urban life. Urban life enables the dynasty to live in luxury, but a life of luxury causes the degeneration of those subject to it. The first generation of leaders are tough, ambitious and charismatic, since these are the qualities associated with the powerful ‘asabiyah which enabled them to find the dynasty. The second generation, being brought up in luxury, begin to take for granted their elevated status, despite the efforts of their parents to keep them in touch with their roots. The third generation are even more degenerate, surrounding themselves with flatterers and sycophants, and entirely losing touch with the roots which enabled their grandparents to attain power. The third generation commit injustices in an effort to maintain their position and neglect the maintenance of their authority, since they assume that urban civilisation is theirs by automatic right. As a result, disunity breaks out, and they are overthrown by either an outside force or an oppressed sub-group of their own dynasty. The new dynasty undergoes the same fate, and another arises.

There is plenty of empirical evidence for the correctness of the above analysis. It is even reflected in folklore, such as the Lancashire saying, Clogs to clogs in three generations. In the world familiar to Ibn Khaldun, the source of new dynasties was always amongst the Bedouin, because the hardness of their lives gave them the necessary qualities – natural selection in the tough desert environment favoured those with strong social bonds, since those groups who fell out amongst themselves would not survive in such conditions. The first generation of each dynasty would always try to encourage their children to spend time amongst their desert relatives, and to acquire the skills of hunting, horse-riding, and hardy living which had led to their own success. The ability to maintain this connection would be conducive to the continued success of the dynasty. The loss of this connection, and addiction to the trappings of wealth and luxury, was the signal for the collapse of their ‘asabiyah and the rise of a new challenger for power. It is worth noting that royal dynasties all over the world are, in fact, the descendents of pirates, whether these are sea-born, as is the case with the Normans [4] who conquered England in 1066 AD, horsemen from the steppes, such as the descendents of the Mongol, Genghis Khan, or Bedouin of the desert such as the Hashemites of Arabia, who became the aristocracy of the Caliphate after the rise of Islam. It is also noteworthy that all royal families tend to be interested in hunting, or games derived from it, such as polo [5]. There is a sense in which the continued success of any dynasty depends on its ability to resist being nice to its children. Some people have suggested that

Viability versus tribalism

639

K 33,3/4

640

this is the logic which underlies the English aristocrat’s traditional tendency to send his male children off to boarding school at a very early age, where the poor young fellows are subjected to a sustained campaign of brutality. If this is the case, making these schools “caring” and open to girls as well as boys is a disaster in the making! The Muqaddimah is fascinating reading, and surprisingly “modern” in much of its content. Ibn Khaldun was well aware of the importance of commerce to the sustenance of urban life. The backbone of commercial life in the Islamic world was the network of overland trade routes which stretched from Spain, through North Africa, then across the Red Sea to Arabia, and from there by sea to India, or northwards to Syria, then eastwards towards Central Asia and on to China. Ibn Khaldun assumed that ‘asabiyah would lead any social group which possessed it in abundance to the quest for royal authority, and that the only authority higher than this stemmed from the prophetic tradition, passing from Abraham (common ancestor of Jews and Arabs) to Moses, then to Jesus and, finally, to Muhammad and the line of the Caliphs. The tendency to royal authority was natural and universal throughout humanity, whereas the latter authority stemmed from supernatural revelation and was therefore rare. He assumed that the secret of commercial success lay in obtaining, and holding on to, a monopolistic licence to trade granted by royal authority. He therefore emphasised the importance of sycophancy and “not getting above your station” to anyone who aspired to commercial prosperity. This attitude strikes me as the most “old-fashioned” aspect of his work. Although he was surprisingly well-informed about the revival of classical learning in Italy and other parts of western Europe, he entirely missed the importance of the constraints on royal power which had already been developing in England for over a century before his birth. If he could see the modern world, I suspect that the biggest surprise to him would be the rise of the power of monetary capital, and the manner in which we constrain the power of ‘asabiyah within the framework of democratically elected political parties, with their short-term grip on power, and their accountability to the people, rather than to supernatural power. However, it is not hard to see that Ibn Khaldun’s analysis of the rise and fall of dynasties works perfectly if applied to the history of a modern political party such as the Conservative Party in the UK under the leadership of Margaret Thatcher, or the history of a commercial venture such as IBM or Enron. I am also sure that a comparative investigation of Stafford’s System 5 of the viable system model and Ibn Khaldun’s concept of ‘asabiyah would be very fruitful.

The work of Idries Shah As mentioned earlier, Ibn Khaldun was a thinker within the Sufi tradition. My attention was drawn to his work through my contact with a modern teacher in the same tradition, Idries Shah. Shah published over 30 books between 1964 and 1998, making Sufi materials available to our western culture [6]. These works exemplify a very sophisticated knowledge of human social dynamics, which I have found very useful in understanding the complexities of organisations I have worked with. Here is a brief example, directly relevant to the subject in hand: Goat-Leader (Shah, 1983, p. 31) Some goats at pasture saw a lion in the distance. A few of them were alarmed, and ran to the leader of the herd for his help and interpretation. The lion came nearer, looked at the goats and roared. “No need to worry,” said the goat-leader, “and I can prove it. See, what an ugly colour his fur is! And, as for his bleat – you can tell by that that he’ll never amount to anything.” As you will observe, this simple story captures, in a brief and memorable form, the behaviour of a typical human leader at a certain stage of evolution of his social group. Practicalities I will finish this paper by looking at the practicalities of applying an understanding of the tribalism inherent in everyday life to the business of improving organisations. There are a number of aspects of this which need to be examined: (1) identifying the tribal affiliations of the participants in the organisation, (2) considering these when proposing changes to the organisation, and (3) “inoculating” your own team against tribalism. Taking the example of the Chilean experience, Stafford’s account (Beer, 1981, Part 4) shows that, although the reformers working with President Allende worked heroically against the forces undermining their efforts, and recognised by the summer of 1973 that the government would not survive the year, the visceral power of the “tribal” response to perceived threat, when it came, was beyond their worst imaginings. This reflected the tribal affiliations of the two opposing camps: . To one group, Karl Marx’s Labour Theory of Value suggested that “property is theft”. To that group, the middle class claim to ownership of

Viability versus tribalism

641

K 33,3/4 .

642

assets is based on expropriation of the value created by the sweat of the workers, so the workers are entitled to take those assets back from them. To the middle class group, property rights are enshrined in law, and obedience to the law is paramount. The other group is therefore engaged in outright theft, and therefore is acting criminally. As criminals, they deserve to be punished.

These two groups simply did not appreciate the depth of each other’s views, and neither side found effective means to bridge the gulf of understanding between them. In the clash of tribes, the latter group were able to deploy greater power, leading to the disaster of 11 September 1973. Identifying the tribal affiliations of the participants in the organisation The work of Bill Livingston shows that the tribal response to attempts to solve problems is a major factor in the failure of complex projects with a strong technical element. It leads to what Livingston refers to as the universal scenario [7] of project failure. The general assumption that these projects fail because of difficulties with the technology is not born out by any studies of real-world projects. The problematic area is invariably in the social sphere. There is only one way to solve problems caused by tribal affiliation. This is to carry out extensive anthropological research into the organisation’s tribal groupings. Of course, this is not the kind of research that is likely to be welcomed within the organisation, precisely because it is “bad manners” and is likely to be perceived as disloyal to it. In other words, it is not compatible with accepting the values of the tribe! However, it is essential, even though it may have to be carried out by your team without mentioning the matter to anyone within the organisation itself. The rule here is don’t mention this to anyone in the organisation, unless they ask you about the subject. Even if they do ask, only tell them what they want to hear. Because this research requires you to study the realities on the ground, you must get permission from the management to work alongside the operational staff, so that you can gather your evidence. Bill Livingston refers to this as gaining permission to visit the boiler-room. He is here using the analogy of the workings of a steam-powered ship. The captain on the bridge may have his ideas about what drives the ship forward, but it is unlikely that he knows much about the stokers, slaving away down below. However, if they did not stoke the boiler, the ship would not go anywhere. To successfully gain the trust of the “stokers”, you need to work alongside them, and identify with their difficulties. You need to make friends with them and show that you are keen to make life better for them. They will then tell you everything you need to know to really understand the organisation. If this is not possible, either because you cannot obtain permission to visit or because your visit arouses hostility, the likelihood is that the tribal groups

within the organisation have “hidden agendas” which are going to make your job impossible. You then have several options: (1) if possible, report this to your client group, and attempt to obtain permission to attack the real problem, not the theoretical one; (2) go along with the charade, and tell your client group what it wants to hear; (3) quit. In terms of the viable system model, you cannot understand an organisation unless you have a firm grasp of the System 1, operational units. You cannot get this without going to that level, because the people in the levels above don’t know much about it. Once you understand the organisation at that level, it is relatively easy to understand the higher levels. Taking tribal affiliations into account when proposing changes to the organisation Having completed your anthropological research, you should be able to see clearly what can be done to improve things. You will then need to present this to the client group in terms which make sense to them, using the language which their tribal affiliations demand. If possible, you should try to create a virtuous circle, where change gets driven from within because each tribal group perceives it to be in its own interest. The best way to do this, if possible, is to improve the productivity of the System 1 operational units which are under the most stress. If this can be done, neighbouring operational units will then spontaneously demand more of the same, the productivity of the whole organisation will improve, and higher level units will also be able to perceive the benefits. I call this approach Business Process Re-engineering by Stealth. It works a lot better than trying to impose change from “above”, which always encounters resistance from those “below”. This resistance is caused by the creation of a tribal boundary, where the “below tribe” judge their interests to be in opposition to those of the “above tribe”. This phenomenon is excellently surveyed by Stafford, using the terminology of autopoiesis (see Healthy & Pathological Autopoiesis, Beer, 1981 pp. 337-42). The key to success is to start where the pressure is worst, improve things, then move to other areas. Inoculating your team This kind of work is impossible if your own team is embedded within the tribal affiliation network of the client organisation. If it is, you will not be able to “see the wood for the trees”. Bill Livingston has successfully identified the appropriate social organisation for such a team, which he calls the skunkworks [8]. It has the following characteristics.

Viability versus tribalism

643

K 33,3/4

. . . .

644

.

It is small and multi-skilled. It has no fixed hierarchy. There is a very high information bandwidth between members. It operates outside the standard cultural framework of the organisation. It is optimised for rapid learning.

Forming such a team is itself a major challenge, since it cannot be fitted into the normal pattern of tribal affiliation of any organisation and it requires unusually skilled members. Experience suggests that such teams usually form around a leader who has learned this working method by apprenticeship with an existing skunkworks. Skunkworks organisations therefore tend to spread out in a complex network from their point of origin, as the members of one such team leave and found new ones. This phenomenon is probably the reason for the clustering of economically successful organisations, as analysed in-depth by Michael Porter [9]. Bill Livingston, who is now retired, spent most of his working life inside skunkworks teams within US corporations, where these “centres of excellence” were the backbone of the engineering capability which contributed so much to the rise of the USA as a world power. The evidence, unfortunately, suggests that the political power of finance departments within those corporations since the late 1970s has led to the closing down of many of them, since they do not “perform well” by the metrics popular amongst financially trained managers. Because this organisational structure is poorly understood outside engineering circles, skunkworks organisations are quite difficult to sustain over a long period of time. The only large organisation of which I am aware, where this method of working is used consistently throughout the whole structure, is Semco in Brazil, where the CEO, Ricardo Semler, has operated this method for over 20 years and grown the company to 3,500 staff [10]. Although he has written two best-selling books about this (Semler, 2001, 2003), few people appear to be interested in emulating his working methods. The reason for this is probably that they undermine the tribalism in the workplace which most people are comfortable with, and which this paper describes. Implications of cheap ICT The tendency towards obedience to authority, and conformity to peer group norms, leads people to act in a manner which suppresses their intelligence. People who do not “fit in” often fail to do well in environments where these traits are encouraged. I have found that these “members of the awkward squad” often thrive in the skunkworks environment. The development of inexpensive computers, combined with cheap telecommunications, has delivered a new infrastructure, the Internet, and made it available to all of us. The implications of this revolution have only just started to make themselves felt.

Basically, computers are ideal “rule-following morons”. Once programmed with the appropriate software, they will obediently follow a given set of rules at a pace which no human can match. It is no surprise that they have had their greatest impact in industries which are highly rule-bound, such as insurance and banking. As the cost of ICT plummets, it becomes cost-effective to use them in other industries, where the rules are more complex. The modern corporation, as pioneered by Alfred Pritchard Sloan at General Motors in the 1920s, tends to encourage obedience and conformity, but, if this reduces people to “rule following morons”, it merely puts them in competition with cheap computers. This hardly seems like a good approach to wealth creation! Stafford Beer understood this issue a very long time back. He covered the issues in Designing Freedom [11], his CBC Massey Lecture, presented on Canadian radio in 1973, in a manner which has never been bettered. He understood that the proper use of ICT could enable a new kind of viable, adaptive organisation to be created, but that its misuse could instead lead to the suppression of human freedom. I hope that this short paper offers some pointers which will enable admirers of Stafford’s work to avoid some of the pitfalls, and work towards creating organisations which can help to fulfil his vision. I can think of no more fitting tribute to his life’s work. Notes 1. For a study of what can be gleaned from palaeontology about the evolution of the human mind, see Mithen (1999). 2. Chapter 9 of Milgram (1997), citing Asch (1951). 3. He has exerted massive influence during the last 2,000 years, despite the fact that there is very little evidence for his existence. See Eisenman (2002) for a survey of the historical evidence for the existence of Jesus and his brother James. Even the name, Jesus of Nazareth, is a mistranslation, since Nazareth did not exist during his supposed lifetime. 4. The name Norman is a corruption of Norseman. The Normans were the descendents of the sea-going Norsemen (i.e. Vikings) who had conquered the land which is now known as Normandy. 5. Polo is derived from the Afghan game Buzkashi, which originated as a military training game amongst the Mongols, but is still played in Afghanistan. 6. For information about these books, contact the Octagon Press, London at: http://www. clearlight.com/octagon/ 7. Livingston (1990, p. 248). 8. The term skunkworks originates with the team based at Lockheed’s Burbank, California plant who developed the U-2 spy plane, the SR-71 Blackbird (the fastest aircraft ever built) and F-117 Stealth fighter. For a history of this organisation and its achievements, see Rich and Janos (1994). 9. See http://www.isc.hbs.edu/econ-clusters.htm for Michael Porter’s works on this subject. 10. For an account of this, see Semler (2003). 11. Re-published as Beer (1994) as part of the Stafford Beer Classic Library.

Viability versus tribalism

645

K 33,3/4

646

References Asch, S.E. (1951), “Effects of group pressure upon the modification and distortion of judgments”, in Guetzkow, H. (Ed.), Groups, Leadership and Men, Carnegie Press, Pittsburgh. BBC (1999), BBC News: Friday, 26 March 1999, Published at 19:49 GMT, headline UK. Thatcher Stands by Pinochet. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester (Stafford Beer Classic Library). Beer, S. (1994), Designing Freedom, Wiley, Chichester. Eisenman, R. (2002), James The Brother of Jesus, Watkins Publishing, London. Ibn Khaldun (1978), The Muqaddimah, Translated from the Arabic by Franz Rosenthal, abridged and edited by Dawood, N.J., Routledge and Kegan Paul, London and Henley. Jeter, L.W. (2003), Disconnected: Deceit and Betrayal at Worldcom, Wiley, New York, NY. Livingston, W.L. (1990), Friends in High Places, FES Publishing Ltd, New York, NY. McLean, B. and Elkind, P. (2003), The Smartest Guys in the Room: The Amazing Rise and Scandalous Fall of Enron, Viking. Milgram, S. (1997), Obedience to Authority, Pinter and Martin Ltd, London. Mithen, S. (1999), The Prehistory of the Mind: The Cognitive Origins of Art, Religion and Science, Thames and Hudson, London. Rich, B.R. and Janos, L. (1994), Skunk Works, Little, Brown and Company, Boston, New York, Toronto, London. Semler, R. (2001), Maverick!: The Success Story Behind the World’s Most Unusual Workplace, Random House Business Books, New York. Semler, R. (2003), The Seven-Day Weekend, Century. Shah, I. (1983), Reflections, Fifth Impression, Octagon Press, London, p. 31. The Week (2003), No. 417, 12 July 2003, headline Santiago, Dennis Publishing, London, p. 8. Toynbee, A. and Somervell, D.C. (Eds) (1987), A Study of History: Volume I: Abridgement of Volumes I-VI, Volume II: Abridgement of Volumes VII-X, Oxford University Press Inc., USA.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Observations on the development of cybernetic ideas in Colombia

Development of cybernetic ideas

647

A tribute to Stafford Beer German Bula Universidad de los Andes, Bogota´, Colombia Keywords Cybernetics, Transformational leadership Abstract This paper is an account of management and organisational interventions in multiple institutions and enterprises in Colombia during the past 10 years, mainly influenced by Stafford Beer’s ideas and work. It offers comments about the use of these ideas in three projects; particularly it focuses on an intervention in the National Controller’s Office. These are accounts of failure and success. However, assessments of success and failure are tempered by difficulties in appreciating the complexity of social processes and our inability to see cause-effect connections. The paper offers insights about concrete aspects of these interventions and the footprints they have left in the country.

Introduction I find myself unable to believe in the idea of digging at corruption as if it were a dragon to be flagged. It is a product of the system, not an entity. The purpose of the system is what it does; if the system is corrupt, it is because it has been designed to be corrupt; ergo, redesign it!! (Stafford Beer, Cartagena’s forum, 1996)

Since the early 1980s, several academic and consulting interventions in Colombia have used Stafford Beer’s proposals, distinctions, and tools braided with conceptual contributions from other sources, such as critical hermeneutics, second-order cybernetics, the analytical theory of language, constructivism, system dynamics, systemics in general, structuralism and post-structuralism, Marx’s critical thought, and in particular Habermas’s “reconstruction”, etc. These interventions have taken place in public companies such as SENA (Servicio Nacional de Aprendizaje) – National Learning Service, the Contralorı´a General de la Repu´blica – The General Controller’s Office – and some regional controller’s offices, the Fiscalı´a General de la Nacio´n – The General Attorney’s Office, the District’s Veedurı´a – “ombudsman” office of Bogota´, the ministry of Education, educational programmes at the Universidad de Los Andes, and dozens of punctual interventions, syntegration workshops, conferences, seminars, etc. that had involved NGOs, companies and institutions of all kinds. Through the General Controller’s Office and the Veedurı´a, the

Kybernetes Vol. 33 No. 3/4, 2004 pp. 647-658 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523625

K 33,3/4

648

intervention has reached more than a hundred national and local public entities. To a greater or lesser degree, some of the dynamics generated by these interventions are already visible since they have reached their conclusion. Others are still in progress such as teaching programmes at the Universidad de Los Andes and work with the district’s Veedurı´a and 30 public entities. All these are additional reasons to thank the honour of the invitation to this celebration[1], of friendship, gratitude, admiration, and commemoration. I also assign generative value to this celebration; it offers the possibilities of reaching an interested and active listening in the medium- and long-term. This paper is based on an observation of the current state of affairs in Colombia. Work has taken place in the public and private sectors, but this observation is focused on public policy and the State rather than on particular companies or organizations. To ask about the effectiveness of a process is to ask about an evaluation which is linked to an expectation (ex-ante). I do not forget that a certain spirit of Procusto can accompany this kind of reflection, which in any case, I judge useful for the immediate and future action in my Country. I believe, as Roberto Zarama and some other friends and consultants do, that the best tribute that can be made to Stafford Beer is improving the effectiveness of our societies and the organizations they constitute. Indeed, we are stimulating efforts to enrich our society’s capacity to use Stafford’s contributions as well as those of other systemic and humanistic schools. Central affairs of the present world, such as peace and equity and the Aristotelian “eudemony”, that were Stafford’s meta-purposes, are also our own meta-purposes. In this paper, I take some of the points made by Stafford while discussing success in a cybernetic intervention (Beer, 1989). Tyrians and Trojans blame the Colombian State as responsible, by action or omission, excess, deviation or defect, of the diseases that the nation suffers. In that sense, one cannot say it is a “successful” enterprise, judging by the fact that Colombia is recognized, internally and externally, as a country whose viability can be questioned. Stafford’s views were that “There are no problems and solutions, except as abstractions of reality” and “it is central to the viable system that it should be in constant state of evolutionary flux” (Beer, 1989). Drug dealing, paramilitarism, guerrilla, corruption, impunity, poverty, vulnerability to natural disasters, dependency and identity crisis are the nine main cracks in Colombia’s identity. The perception of their importance varies depending on the perspective of the observer. In fact, there are those who argue that in comparative terms, institutions in Colombia are stronger than those of other Latin American countries, which would imply that a debate about them would have to be located in the metasystem, that is, in the “world system”. In fact, it seems sterile to try to solve or dissolve problems of Colombia, a country highly plagued by international drug trafficking, without working on the metasystem: “we do not confront closed systems, and we do not have the algorithm” (Beer, 1989). In spite of this, stubbornly, we think that a more

effective State, and more effective companies and social organizations, with learning capacity and equipped with a clear mission centered in the people, are necessary for the task of solving or dissolving the aforementioned cracks. In the following sections, I reflect upon three projects aimed at improving effectiveness. Brief commentary on three interventions The SENA project: an intervention that, though applauded, was ineffectual Having already learnt something about cybernetics, I was called to advise the National Learning Services, SENA, a public entity in charge of technical and professional higher education in Colombia, to produce written and audio-visual materials for the management training of small and medium business executives. In SENA we had the support of enthusiastic instructors, and although we know that our training material was used, it is evident that it did not generate a visible dynamics. I believe that if it were decided today to continue with that task it would imply a fundamental rethinking, giving a central role to workshops and the learning of conceptual distinctions. The production of training material by itself did not give origin to a dynamic process. It would have been better to try a radically different approach; instead of producing training materials for institutional use, it might have been more powerful to get instructors and businessmen intensely involved actively in a few important distinctions through workshops and other exercises. The experience in the CGR Commitment and expectations of high level management and the challenging nature of the organization. In 1995, I was requested to advise the General Controller’s Office (Contralorı´a General de la Republica – CGR). I shared this idea with Raul Espejo, who accepted to be the Project Scientific Director. The history of the organization placed it as one of the most corrupt and inefficient of the Colombian State, and citizens judged it as a true cancer that not only failed to control, but also traditionally had increased administrative corruption in complicity with the fraction of Congress that supported the General Controller of the time. In 1995, the General Controller gave us carte blanche to improve the effectiveness of the service and the Vice-Controller gave his support as the Director of the Project. Both of them had extensive and intense meetings with Raul, some of the consultants and myself, and at the beginning they were committed to the project. In any case, it turned out a powerful experience to work in the very organization that was in charge of “control”. The topic of achieving the commitment of high-level management is as old as the history of management consultancy, and in the case of the CGR, I can only state that it was adequate. However, the CGR is an institution to control State activities and unfortunately the project that did not have the commitment

Development of cybernetic ideas

649

K 33,3/4

650

at the “owners” of the regulated system, that is, of the Presidency, The Congress, Ministries, Planning Department, etc. When our intervention in the CGR was well advanced, we were able to talk to the President about the project. Unfortunately, nothing came out of this meeting. In fact, we hope that one of these days a President of Colombia will take some time to review what science can contribute, and has already contributed, to reform the State. Stafford pointed out a similar case in which he had to settle for initiating his work in a specific dependency of a company and suggested: “The absolute need to determine the total recursive system before engaging in such detailed work” (Beer, 1989). For us this intervention in CGR was not, however, a waste of time, in spite of the limitation of working in a regulatory organization without the full commitment of the management of the whole system. Consultants often find that upper management expects immediate, if not magical, results. In the CGR, the traditional failures of the organization and the pressures on the General Controller made it more urgent to obtain results in the short-term. This contradiction was never solved; it was undertaken by a combination of theoretical presentations that we felt were well received, interminable discussions on short- and long-term actions, and the overload of the directors that actually generated room for us to make progress. Elsewhere, I have narrated an extract of the precise conversation that illustrates the difference in perspective: question from the General Controller, “The process that you plan is aimed to improve the efficiency of the CGR?” answer of Raul Espejo, “we do not seek for the Controller’s office to do what it does now in a better way. Our proposal is to question what it does” (Bula, 1998). From consultant to administrator of the assisted company. As politics would have it, the Vice-Controller (and, as said earlier, Director of our project) resigned. He did it at the end of the very same day in which the official presentation of the project was made in front of an audience of about a thousand people in front of whom he himself had talked next to the General Controller, and other personalities. The first issue to work on was the surprise of Raul Espejo who could not believe that something like that had happened. At the end, he convinced himself that the attitude of the Vice-Controller had been the healthiest. The Vice-Controller preferred not to ruin the presentation ceremony and waited for it to pass. The General Controller then proceeded to offer me the newly available position. After a few discussions with my colleagues and other close people, I accepted knowing that it was a technical mission, and that my condition of being an opponent of the government had to be recognized for any political purposes; in fact, my predecessor in the position because of the perceived need for political checks and balances in an institution like the CGR, had also been an opponent to the government. It became obvious that my contribution to the project had to reduce as I needed to face multiple additional tasks related with my new position.

The value of having a central message, of a team, and of starting out with properly selected prototypes. A key moment in the advance of the project was the enthusiastic agreement of the General Controller to develop a “New Discourse for Control”. Raul Espejo coined the phrase “control is much more than detecting faults in the institutions; control is to obtain desired results through promoting its proper operation”. ´ ngela Espinosa, Roberto Zarama, Alfonso Reyes, and Ernesto Lleras A formed the team of national external consultants. Some students from the Universidad de Los Andes (who are at present lecturers with masters or doctorate degrees) were engaged as assistants and we also structured an internal team who received special training. We chose a couple of public entities where to start work, in order to learn and then work with more entities, learn again, increase the number of involved entities and so forth. We were to perform “second order audits” whose intention was not the traditional one, to make “findings”[2] of illicit activities, but to make organizational diagnoses of structural faults that could be the origin of control failures. By the time of making this choice of entities, the “Project of change and institutional strengthening” was already known in the organization and was producing antibodies. It was seen either with curiosity, or scepticism, or outright hostility. The CGR was at the time divided into sectors, each in-charge of a senior director, covering the totality of the State in the national level. It seemed suitable to involve several directors including organizations of different sectors in the prototypes to search a bigger internal political support; but doing this implied working with organizations that did not have recursive relations among them. It was necessary to partially sacrifice the power of the prototypes. However, their results allowed increasing coverage. Now, we could work widely with the so-called “New Discourse of Control”, and perform 30 Systemic Audits, where systems, not institutions, would be observed. What aspects of this new auditing could be reported, which in fact had already caught the interest of some civil servants? In the Agrarian Bank (La Caja Agraria) – state owned bank supporting agriculture – it was diagnosed, among other faults, one related to its identity: it was suffering from a clear schizophrenia. The bank was on the one hand a bank, and on the other a social services institution. It was in-charge of advancing social tasks such as caring for reinserted people in communities, from the armed conflict, victims of earthquakes and floods, populations of farmers that by different reasons confronted difficult situations, etc. If questioned about its inefficiency as a bank it responded by pointing to its heavy social burden, and that prevented it from being compared to private financial organizations. If questioned about its ineffectiveness on the social sphere, it argued the inadequacy of its structure and profile to deal with such challenges. The natural result was the incapacity to learn and an erratic walk in

Development of cybernetic ideas

651

K 33,3/4

652

which for decades it dragged its identity crisis, its non-viability, and the need to be refinanced from time to time by the Central Government. Another structural fault was the over centralization of credit approval, which in fact was the source of corruption. Such a finding was certainly perceived as old news, but the point was that traditional audits never questioned these types of issues. In the Civil Registry (Registradurı´a Nacional del Estado Civil) – in charge of citizens’ identification and the running of political elections – it was found, among other structural faults, that the regulatory role of the Systems Office was not understood. This office produced services to the public that were not noticed by the organization (one of the archetypes of structural problems). This very fault appeared later in many entities and it was easily linked to operational deficiencies that were familiar to civil servants. The fact that these faults were easily visible, and made sense to CGR functionaries, gave the project the space it needed. Meanwhile, however, in the CGR we faced our own schizophrenia. As we worked to achieve structural change, we were required to participate in the fight against corruption. We had to find a way to do it coherently with our discourse, pointing out, for example, that the corruption we were reporting in the Civil Registry was linked to the lack of transparency in the process. The “law of the glass of scotch” – any scotch served in the kitchen has hand served ice in it – was coined in the occasion with great figurative power. Creating capacity to obtain requisite variety. The transition to 30 systemic audits (SAs) demanded a drastic extension of capacity, understood the size of the trained civil servants team, to take care of that type of intervention. Here, the training plan became crucial; it was structured around three types of workshops: the epistemological-ontological, the methodological and the poietic, that benefited from the introduction of the VIPLAN Learning System software[3] as tutorial expert system. The idea was to emphasize the value of the internal control of organizational processes, beyond the traditional emphasis in external controls of procedures and legality. Auditors and an important number of employees of the audited organization were observing the organization from the same side of the table. The preparation of a set of tools by Raul Espejo, including guidelines for systemic audits and structural archetypes to facilitate the observation of problems of identity, cohesion, citizenship and performance in the audited organisations, took the project to its most interesting stage. We even advanced in producing a systemic model of the State, unfolded in 18 sub-sectors, with the aim of restructuring the CGR as a good model of the regulated system, but by then it became evident that we lacked indispensable co-requisites. Project’s shortcomings. I have already mentioned the fruitless conversation with the President to validate the project for the State as a whole beyond the CGR. Without the support of the Head of the State, it became impossible to build up a constructive relation with the National Planning Office. This office

had been working in the modernization of the State, through a project called “SINERGIA” and felt in competition with our cybernetic intervention. Our relations with the Finance Ministry were neither fortunate. In this case, we failed to construct a valid proposal to make consistent our regulatory models with their budgetary control. Explanations are of little importance. The Ministry decided to implement a World Bank’s system; SIIF – Integrated System of Financial Information – and conversations with them did not get us anywhere. The World Bank’s project was not going well, but our participation was weak, something that affected our so far well earned image. Finally, it became impossible to build-up adequate connectivity between the State’s control, budgetary and investment institutions, something necessary to connect second-order auditing to resources bargaining and investment decisions at the national level. This failure diminished the global impact of the project. Progress and stagnation in the educational sector: a project in waiting. New errors and limitations After the 1998 elections, I was appointed in the Ministry of Education. In that Ministry, we promoted projects in primary, elementary and higher education. In primary education, we proposed to structure the educational system as a recursive organization with high capacity of learning. However, an incomplete and not sufficiently apt team, restricting conditions of a multilateral bank’s loan, inadequate appreciation of the complexity of relations with educational trade unions, and other reasons that are not the subject of this text, resulted in the project not making significant advances. The challenge is still there, for Colombia and Latin America, where the classroom holds an enormous possibility for creating a new type of citizen – critical, supportive, creative, enterprising, and democratic. Regarding higher education, we proposed a social mobilization around its autonomy, purpose, identity, systemic character, and relevant advances were obtained. We made institutional modifications, which tended to equip the system with suitable regulators. We worked in an ongoing design of regulators that were “a good model of the system to regulate”. The work has been continued with its ups and downs, and there have been – regrettably – voices asking for a “simplification” of the regulators – that is, asking for a bureaucratic centralisation. However, the reforms arose from an ample social mobilization, which aimed at implementing the often preached autonomy of higher education, with all its implications, starting from the fact that the system was self-conscious and would observe itself as such. What advances have been registered? . . . the success and the failure, those two imposters (Rudyard Kipling, in the poem “If”)

Development of cybernetic ideas

653

K 33,3/4

654

It is not possible to know about the longer-term impact of each of the interventions. Life has surprises stored for us and, the seeds of change that we planted may end up producing dynamics on their own, even if they are small and localized. However, the advances produced in second-order auditing, in the academic sector and higher education are undeniable. One of life’s surprises was the decree 2309 of 2002 that blesses for the first time in Colombia second-order audits. What was the process that led some of our workshop participants to consecrate this novel idea in a concrete new norm? I believe that the Colombian higher education found a conceptual space to reconstruct its autonomy. An oximoron coined in the course of the discussions turned out to be of great communicational power to entail that autonomy is not something you get by decree, a kind of legal entity or mere political vindication. We presented the statement “I order you to be autonomous” as a contradiction in terms, to indicate that autonomy is constructed by recognizing the system boundaries, reflecting itself as a system, and reconstructing its functions to create, regulate, and implement its total mission. As an example, we have made others see how while the more advanced universities had an intelligence function, caring for their long-term development, the so-called higher education system lacked the articulation to provide it with capacity to deal with future challenges and opportunities. This clarity allowed a thorough corrective action. New distinctions based on cybernetics were introduced in the language of audit, as well as in that of organizational structures. These distinctions came not only from extensive teaching programmes and workshops, but also from working with them during the projects. Concepts and ideas such as the ones provided below have become a commonplace for a good number of actors in the educational system in the civil service. (1) Extrinsic and intrinsic control. Automatic control (as a basis of the internal control). (2) Control of control (as a basis of the audit of second-order). (3) Irrelevance of most of the reports demanded from entities (as the foundation that justifies the distinction “balance of the control”, fashionable, in the international audit). (4) Archaeology, mining, unnecessary filling of shelves and sensationalism, as metaphors of the old style for audit reports. (5) Corruption as an emerging property of systems with structural pathologies, not as a “thing”. (6) Anti-corruption and “antis” in general as reductions that end up propelling the phenomena that they aim to eliminate.

(7) Primary activities and regulating functions. Regulating organizations, organizations that develop primary activities, and mixed organizations (as a basis of Systemic Audits). (8) Difference between statements and judgments. The need to corroborate judgments. Auditing as a tool to align distinctions and recognize the discourse of the other. (9) Audits that generate commitments as opposed to policing audits. Audits whose reports are conversed with affected organizations versus simple unilateral reports. (10) Efficiency, efficacy, and effectiveness. The lack of pertinence of sensationalist indicators in relation to the intention of the system. (11) Informational versus operational domain. (12) The construction of confidence as a central task. (13) Accountability (availability for interpellation), as an element in the construction of trusting relationships. Other distinctions remained in the hands of academics and of an advanced minority. . The law of requisite variety was stressed many times (only variety absorbs variety), but we felt that it has still not passed from the informational domain, the same is true for the concepts of residual variety and the dilemma of control. . The same can be said about the balance between autonomy and cohesion, the pathologies derived from separating thinking and doing, and of the effective design of social control mechanisms. In any case, we know that our trainees were massively transformed by distinctions such as emotion as a space for action, complaint as diverse from claim, to entail distinctions, the eye of the observer, “know” in a certain dominion is to know how, the cognitive blind point, etc. Workshops to appreciate communication structures (Bavelas’ structures), beer game, construction of trust, chromatic perception, distinctions between the different beats in the vallenato music etc., are object of pleasing memories whenever and wherever we accidentally meet with any of the hundreds of participants. The “law of the scotch’s glass” is fully used by those who push processes towards transparency.

What are we doing, what do we intend to do, what do we expect? Stafford indicated that a value of the cybernetic intervention was “to inject the language of viability” (Beer, 1989). While referring to the CGR, Raul Espejo said, “without changes in the State, I think it will be difficult to consolidate the changes in the CGR”, and also that “an enormous effort to

Development of cybernetic ideas

655

K 33,3/4

656

plant a new language of control is required” (Espejo , 1998). Elsewhere, we have said that the long-term effects of the project in the CGR are impossible to anticipate. The new administration since 1998 continued with the main thrust of the project and accepted the need for improved self-control of organizational systems to counter illicit practices” ( Espejo et al., 2001). I have indicated that we cannot track the developments of all our interventions. In the case of higher education, not only did we plant abundant seeds for a system with capacity to redesign itself, but also during each academic semester at the Universidad de Los Andes courses recurrently plant these seeds again. Currently, vigorous training is taking place with employees of the Veedurı´a Distrital of Bogota´ (Ombudsman for the city of Bogota), and in many places we see activities taking place that are congruent with our goals, from systemic family therapy to complexity networks, from the approach of resolution to that of dissolution of conflicts used at present by hundreds of activists, from traditional teaching practices to constructivist pedagogical experiences in which teachers of basic education see their ethos achieved, from the old lessons on peace, coexistence and ecology of knowledgeable Amazonian caciques and the Mamos of the Sierra Nevada to the indigenous and peasant opposition to violence and their demands of respect towards their communities. Marylin Fergusson wrote: “All this is a process. Solid world is a process, a subatomic particle dance. A personality is a set of processes. Fear is a process. A habit is a process. A tumor is process” (Fergusson, 1980). In the same work, talking about the new way to fight that millions of citizens around the world have assumed, she indicates: “since the process is the same objective, it cannot fail. All collective fight in favor of social transformation becomes an experience of social transformation”. “It is the same in any process. Independently of the clear vision that one can have at the beginning of a process, the meaning of a process of change arises as it is taking place” (Espejo, 1998). Stafford called without hesitation to his interventions “the continuing process of organizational reform” (Beer, 1989). The previous reflections should not serve as an excuse to evade the critics and the rigorous evaluation of our work and its weaknesses. If we do not do it well, we will imitate Procusto soon and we will cut the corpse to make it fit the coffin. I have already said what we do; we work and think towards the improvement of the ways inhabitants of the planet relate among themselves and toward their surroundings. We mainly do it in Colombia, and we use systemic thinking. We are also conscious to overcome many obstacles and deficiencies surrounding our work. Why do we sometimes seem exotic academics more capable of showing elegant diagrams than applying them on the field? Why do we show the scientific basis of our work instead of just applying our thinking demonstrating

common sense and wisdom on the way? What are we lacking, so that our work can be put in friendlier terms? We dream of a better future and we will work for it every day. We do not have a crystal ball, and we do not know what will happen. However, we know that here and now, thousand of citizens are talking about contributing to change. caminante no hay camino; se hace camino al andar (Antonio Machado, in the poem Cantares)

How Stafford’s ideas were planted and how they are being reaped currently in Colombia is a subject that might never be known exactly. What I can tell is the initial story; to go from there, but recording the plight of his ideas in Colombia is an impossible task because of their inherent dynamics. The beginning took place on board of a small plane, accompanied by two of his disciples, one senior and one junior who worked on her doctoral thesis under his supervision. During the journey, I commented frankly that I was going to the SMEs symposium to which we were all going, not only as a past-president of the association, but as an interested but a sceptical listener of a planned conference on managerial matters, specifically on business processes reengineering. I explained how, by then, from past experience, I doubted that those matters were anything, but administrative fads, more useful to the business companies of their authors than to the attainment of organizational excellence. Professor Raul Espejo, a disciple of Stafford Beer, who indeed was the programmed lecturer to speak about re-engineering the following day, was my counterpart. After an interchange maintained during the flight my interest in his conference increased. When his presentation was finished we still had some time to continue talking on organizational and social matters. Briefly, that was the start of a productive friendship with Raul. It led me to Stafford and Heinz von Foerster, and later to work with the indirect contributions of Humberto Maturana and Edgar Morin, as well as with the direct contributions of Markus Schwaninger, Allenna Leonard, Barnett Pearce, Joe Truss and that handful of ´ ngela Espinosa, and Colombian consultants, Alfonso Reyes, Roberto Zarama, A Ernesto Lleras, who still continue leading the systems field in the country. Professor Gerard de Zeeuw, other European academics and many Colombian students, in the same school, add to the list of colleagues of whose friendship I honour myself. With Stafford, in celebration, let us join together saying the words of Violeta Parra: Thank you, to life that has given us so much. Notes 1. This paper was a contribution to the Celebration of Stafford Beer’s Life and Work, at the London School of Economics on 3 March 2003. 2. The team “hallazgo” (“finding”) is characteristic of auditing reports and is used from the emotion of an inspection. In this work, the same term was used with the purpose of forcing a

Development of cybernetic ideas

657

K 33,3/4

658

new interpretation, which takes away the connotation of something illicit or hidden. The “findings” of the new type of auditing were of structural character. 3. Viplan Learning System is the tutorial created by Professor Raul Espejo to learn to use the Viable System Model. It is a Toolbook based software that was developed by Syncho Ltd, Aston Science Park, Aston Triangle, Birmingham B7 4BJ, UK. References Beer, S. (1989), “The evolution of a management cybernetic process”, in Espejo, R. and Harnden, R. (Eds), The Viable System Model: Interpretations and Applications of Stafford Beer’s VSM, Wiley, Chichester. Bula, G. (1998), “Esquizofrenia. En La construccio´n de un nuevo discurso del control, Reflexiones sobre los elementos del proceso de cambio y fortalecimiento institucional 1995-1998”, Contralorı´a General de la Repu´blica, Bogota´. Espejo, R. (1998), “La auditoria en la construccio´n de un Estado efectivo. En La construccio´n de un nuevo discurso del control. Reflexiones sobre el proceso de cambio y fortalecimiento institucional 1995-1998”, Contralorı´a General de la Repu´blica, Bogota´. Espejo, R., Bula, G. and Zarama, R. (2001), “Auditing as the dissolution of corruption”, Systemic Practice and Action Research, Vol. 14 No. 2. Fergusson, M. (1980), La Conspiracio´n de Acuario, Editorial Kairo´s, Barcelona.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Heinz and Stafford

Heinz and Stafford

A tribute to Heinz von Foerster and Stafford Beer Rebeca Donoso

659

Colegio Nueva Granada, Bogota´, Colombia Keywords Cybernetics, Education, Poetry Abstract In June 1996, Stafford Beer and Heinz von Foerster were invited to Colombia to contribute to a number of events, among others to a symposium about second-order auditing organized by the National Controller’s Office of Colombia in Cartagena de Indias. As part of the occasion, they visited Mompox, a colonial town in the delta of the Magdalena river. During that trip, the author met Stafford and from there on she developed a special relationship with him, similar to that she already had with Heinz. The sad death of both great men of cybernetics within a short period of time was the origin of this tribute.

Tribute Two of the most brilliant minds of the 20th century are gone, within less than 2 months difference between them, and I am sad. Stafford Beer died in a hospital in Toronto on 23 August and Heinz von Foerster died in his home, in Pescadero, CA, on 2 October. At the risk of sounding superstitious, I imagine them making the decision to leave together as was actually the case, on 23 April 1616, when William Shakespeare and Miguel de Cervantes also decided to die. I am sad for the world that has lost these two great men, but there is a selfish sadness in me too because I shall never see them again. They were my mentors and also my friends. Beyond the fact of death as such, it seems to me that the worst part of death is this solitary absence in which we, the living, have to continue our existence. For some strange and stubborn reason, I remember them together, on a trip to Mompox, Colombia, 6 years back. This is why today I write about both of them. Neither of the two could ever imagine how much and how deeply they would influence my life as of that day. Mompox is a small and beautiful, but oppressively warm, colonial city, south of Barranquilla, on the Magdalena River. To be there is like going back in time, to the 16th century. Nothing seems to have changed since then. Access by land or water is virtually impossible because of the jungle and also for security reasons, so the best way to go is on a small aircraft that announces its arrival by honking its way into a rather primitive airstrip that appears at the last This paper was originally written in Spanish for a 2-day conference held at Universidad de Los Andes (Bogota´, Colombia) on 7 and 8 December 2002, in honor of Stafford Beer and Heinz von Foerster. It was translated into English in February 2003 and slightly modified by the author for the purpose of this contribution to Kybernetes.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 659-667 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523634

K 33,3/4

660

minute amidst the thick foliage. Then, after a few bumps, the plane comes to an abrupt stop. On that day, in June 1996, Stafford and Heinz began a conversation by the shade of a huge rubber tree while the rest of us listened awestricken and waited in silence. They seemed so happy to be together, totally oblivious of these tropical surroundings, yet oddly too, a part of them. In this conversation, they crossed the Atlantic Ocean many times as they remembered the different occasions in which they had met at different professional meetings in America or Europe. Heinz was wearing brown corduroy pants and a dark, Oxford, long-sleeved shirt. Stafford had a cotton plaid short-sleeved shirt, but he was wearing heavy woolen socks and leather sandals. I remember their attire because of the heat. It was surprising that with those clothes, seemingly so inappropriate to the climatic conditions, they could be so happy to have met each other once again. Moreover, when the rest of us (all significantly younger than them) could barely walk through the streets of Mompox at midday, at least Heinz hopped from one sidewalk to the next with incredible resilience (at the time he was 85 years old!), celebrating this encounter and, tacitly giving us a lesson on how to live. At lunchtime, in an old colonial hotel, they even danced outside and tasted all the food that was offered to them without inhibitions or distrust, something that many others in the group were unable to do. Those Latin American women, who, like me, have been educated in a strict Victorian manner, know very well what their place is when they travel with their husbands. I was not there because of my professional “importance” (I was then, and continue to be today, barely an elementary school teacher), but because I was traveling with my husband who was part of an international meeting on auditing practices to which both Stafford and Heinz had been invited as keynote speakers. Because of this same Victorian upbringing, I have the tendency to lurk in the background and to sit as far back as possible so that my presence will be as unobtrusive as possible. On social occasions, I always prefer to observe what is going on rather than being an active participant, because I usually feel too intimidated to speak. Such was the case, in the evening of that same day, when we had already returned to Cartagena, where we were all staying, from our Mompox adventure. This is what happened. Stafford Beer, with his rather slow and playful voice, began to recite some of Dylan Thomas’s poems by heart during our dinner. Without intending to, when he paused, I continued the recitation. He gave me a puzzled look. He did not expect Dylan Thomas to be present in any shape or form in this Hispanic, tropical context. On that day, I had become so absorbed by his musical verses that, without intending to, I impulsively finished what Stafford had begun.

That is how we met. I have always loved DylanThomas’s poetry, but from that day, it took on a special meaning. I had met Heinz a few years before while training to become a systemic therapist. He had asked me to translate his Paris paper [1]. Although I was only one of many students in that program, I had noticed that his hotel room was not as comfortable as he would have liked it to be so I asked him if he needed anything. His eyes lit up like those of a child caught in mischief and, very softly so that no one else would overhear, he asked me for a desk lamp and an additional blanket for his bed because he was cold. He did not want to disturb his hosts for fear that they could not afford better accommodations for him and he asked me to bring the lamp and the blanket from my house if that was not a problem for me. That became our “secret”. Perhaps what I most admired in these two men was their sensitivity and their complete absence of arrogance. They could behave like normal human beings among other more commonplace human beings and they never felt the need to “show” their importance. I loved them for this because they made me feel comfortable and accepted in their presence. Stafford Beer had worked for Salvador Allende’s Government in Chile in the early 1970s. He helped his government to set up a communications and management system and being myself a Chilean, I was aware of that work, but I never imagined that a few decades later I would meet this man whose picture appeared in newspapers and magazines giving personal interviews, and who was as distant to me as Britney Spears or President Bush. Our country then was far from being the “American tiger” that many people think that it is today and that is frequently set forth as an example to follow. We were poor, disorganized, provincial, and much more insular than at present. But because we had democratically elected the first socialist president, the whole world was observing us. Stafford Beer who came from the most sophisticated academic and entrepreneurial environment of the so-called first world had no problem in putting his intellect and experience to serve this impoverished third world country. After having met him, I have no doubt in my mind that he did this with an unselfish and spontaneous desire to help. Among his poems [2] I found one entitled A Chilean Spring Later, presumably written in September 1974, in which Stafford Beer pours out his feelings with respect to this aborted experience of democratic socialism that subsequently had such a great impact in his life. In a very precise and succinct manner, the poem expresses the feelings of pain and perplexity that Stafford felt a year after Salvador Allende had died. These feelings were shared by many Chileans in those days. The poem is extraordinarily lucid if we consider the fact that the poet lived in the country so briefly. His vision offers sensory details such as luminosity, the particular sound of the ocean, or the refreshing feeling of the Mapocho River near the mountains, that would not have been better expressed by a local

Heinz and Stafford

661

K 33,3/4

662

poet. But in those verses we can also perceive Stafford Beer’s outstanding emotional sensitivity and insight into the Chilean character of that time. I have always found that his conclusion, though not particularly tactful in Chilean terms, is very amusing. He says that the 1973 military coup lacked requisite variety. I discussed this concept with him many times not so much with respect to Chile, but with respect to education and schools. He made me see that the world needed creative means to absorb all the existing variety instead of mindlessly reducing it and, moreover, he convinced me that it was a viable possibility. This became a significant challenge for me in my daily school work. In my opinion, it is important to have a systemic point of view in education. For example, I think about the implications of Heinz von Foerster’s statement that “the hard sciences are successful because they deal with soft problems, but that the so called soft sciences continue to struggle because they have to deal with much harder issues”. For example, Pablo Casals, the Spanish cello player, was mortified because, “What do we teach our children in school? We teach them that two and two make four and that Paris is the capital of France. When will we also teach them what they are? You are a marvel. You are unique. In the entire world there has never been another child exactly like you. In the millions of years that have passed there has never been another child like you. And look at your body – what a wonder it is! Your legs, your arms, your cunning fingers, the way you move! You may become a Shakespeare, Michelangelo, or Beethoven. You have the capacity for anything” [3]. But most schools are still far from that ideal. In my, perhaps, limited experience teaching in Chile and Colombia, very rarely have I seen teachers thinking, even for a short moment, that they too are a part of the school system or of the classroom system as a totality and, therefore, directly responsible for what may or may not occur there, like any other element of the system: students, curriculum, values, behavior, etc. It seems more comfortable to dichotomize their vision and look at the school and its students on one end and themselves at the opposite end, separated from each other and from the system. Of course, if teachers are to have a share in the success of a particular action, it is easier for them to accept their inclusion in the system, but if it means accepting responsibility for a failure it seems better to be left out. In such a case it is better to look for responsibility and for answers to the possible causes of such a failure in the child’s family, or in the school’s administrative policies, or even in the student’s lack of ability rather than to look within themselves and experience that uncomfortable sensation of having had the chance to make a positive difference, but having failed to act resolutely and with commitment as they should have. From that moment onwards, I have been obsessed with the idea of creating a cybernetic school. I imagine it to be round, like the Guggenheim Museum in

New York, where those long and straight eye-vigilance corridors cannot exist and where administrators can no longer “spy” on teachers and students, but where the architecture itself favors the concepts of trust and self control, because there are no vantage points from where the first can control the actions of the latter. I do not see any other way in which we can generate self-discipline and trust if it is necessary for some to supervise others in order for them to behave correctly. It seems to me that, if in the future we want to live in a society that functions well due to self motivation and not for fear of punishment, we have to learn, from a very early age, to be responsible for our own actions and for ourselves. We must avoid doing to others what we do not like done to us and we must think about the common good before we think about that which is best for us. Heinz and Stafford also shared the notions that any kind of knowledge is constructed and that reality is in the eye of the observer. This has significant implications in the way we teach. For example, the explanation that Heinz gave of “the blind spot” at some point in one of his workshops helped me to accept many of my own blind spots at family level as well as in my work. I had never thought about my very stubborn Western tradition of thought bias, or in my learning style, among many other “blind spots” that still persist. Because what is so upsetting about our blind spots is that we cannot see them! Heinz von Foerster and Stafford Beer taught me to look at these things many years back. I learnt to observe myself within the system and to see how I could hinder or help to improve certain situations. Likewise, I have been faithful to their precept of “knowing as little as possible” because the adventure of looking for knowledge or of constructing it with my 9 or 10 year old students is much more exciting than standing in front of them to regurgitate what I already know. This is why I believe that I am a better teacher now than before. Ross Ashby (whose name is “engraved” in my ears as Ross Eshby, because that is how Heinz referred to him) looms like a ghost when I am in school having to individualize instruction for my 22 students. None of the strategies that worked before for me and other students are any good now. If I treat my students like adults, they rebel. If I treat them like children, they are offended. If I ask them to study, they get upset and if I do not give them enough work they get bored. I imagine that, because it is so difficult to individualize instruction, most schools end up reducing variety as much as they can, but from the point of view of the Law of Requisite Variety, this is something reprehensible. It is no easy task to set up a system where the variety of encoders and decoders (if we understand that a classroom is a communication system) is greater than the variety of interferences. But because Heinz and Stafford were so intent in making me understand this, I now feel that it is an ethical imperative for me to try to keep as much of the variety that my students bring with them intact.

Heinz and Stafford

663

K 33,3/4

664

Figure 1. The Duchess of Malfi

Freedom is necessary for cultural and personal diversity and, in turn, both of these are necessary (or required) to safeguard both desire and ability to survive. Given this fact, I have no other alternative but to work much harder to provide each one of my little students with the necessary space to develop as he or she needs. Another interesting poem that appears in Stafford Beer’s Transit is The Duchess of Malfi. The design of the poem – a sun or a daisy – attracts the reader (Figure 1). I imagine that the title refers to the tragedy by the same name written by John Webster in 1613-1614. There are two reasons to believe this. First, its seems to me that Stafford Beer was also obsessed with death as a topic and as a form of liberation. This is a recurrent theme in many poems. At the same time, T.S. Eliot, the contemporary poet and playwright, said that “John Webster was much possessed by death and saw the skull beneath the skin”. Secondly, the theme in Webster’s tragedy seems to be that of being trapped or confined. The soul, like a caged bird, is trapped by the body and can only be freed by death. The Duchess of Malfi, on the other hand, is trapped by her brothers’ malice.

Stafford Beer probably felt this very poignantly because an ailing body significantly limited his mind and extraordinary ideas. I remember him on a trip to the Kennedy Space Center when he wanted to climb aboard a rocket. He was like a child in front of a new toy. Everything that he saw there seemed fascinating to him and his eyes shone with enthusiasm. It is not difficult to imagine that he would have volunteered to fly on a space mission had he been able to do it. Space and all that it had to offer to humanity captivated him, and, unlike many of us, it could be apprehended by his brilliant intellect. The Duchess of Malfi’s situation is less dramatic in appearance, although it is equally painful in the end. She has just been widowed, but wants to remarry in secret with her butler who is beneath her station. She yearns for a placid family life. But her brothers oppose this marriage and indicate that her decision constitutes a double crime. It is a double crime because she wishes to remarry and because she has selected a commoner. When we read Stafford’s poem we are forced to see and feel recursion (that he studied in-depth with his “feedback loops”). It is a circle and the same verses are repeated over and over again, although the meaning may change depending on where the reader chooses to begin reading. This turns this poem into a sort of metaphysical experience as would often be the case when we engaged in conversation with Stafford, regardless of the simplicity of the topic. Towards the end of November 2002, my husband, our best friend, and I were talking about the privilege that the three of us had had of meeting Stafford Beer and Heinz von Foerster. Both of them generously opened the doors to their respective universes and through them we were able to meet other extraordinary characters such as Ludwig Wittgenstein, Warren McCulloch, Gordon Pask, Ross Ashby, or Margaret Mead. These people ceased to exist in books for us and became flesh and bones, some profoundly loved and admired, with their virtues and vices, merely because, at some point, they had belonged to Heinz and Stafford’s own world. For instance, I remember Margaret Mead as if I myself had seen her entering the Hermitage Museum in Russia with a large walking stick. But I have never been in Russia nor have I ever met Margaret Mead. Even though in those days no one could enter a Soviet government building with a walking stick because it was considered a potential weapon, Mrs Mead walked resolutely in front of the armed guards whose gaze continued to follow her small figure without really knowing what to do because there was no walking stick in sight. It transpired that the said walking stick had been hidden in Heinz’s pants. All these anecdotes, lived by Heinz von Foerster and then shared with his students and friends made it possible for many of us to come into contact with many “giants” of the 20th century.

Heinz and Stafford

665

K 33,3/4

666

Warren McCulloch, whose extended arm pointing we have all seen in pictures, called Heinz von Foerster on the phone at 2 a.m. in the morning one day to tell him that he absolutely had to hire Gotthard Gunther because he was a man who asked the right questions [4]. Stafford Beer wrote a sonnet for Warren McCulloch’s 65th birthday. It goes more or less like this: “Others may insist that science is something serious. But we shall take our laughter from the universal fear and have fun”. This is the way that Warren McCulloch will live in my memory, although I never met him. He was intense, he enjoyed life, and he loved his friends. This makes him lovable to me too. I also suspect that the famous picture refers to Margaret Mead’s anecdote with the aborigines of the South Pacific and to that historic moment when she could finally penetrate their language and realize that the word she heard constantly meant nothing more than “to point”. From that moment onwards, the gesture of pointing at something became the metaphor for the knowledge of reality. The Guardian’s obituary dated 4 September 2002, describes Stafford Beer with two very good adjectives: flamboyant and charismatic. This same publication emphasizes his work on complex social systems and, in particular, it highlights his idea that societies, governments, and businesses have to constitute interdisciplinary teams to be able to solve their problems. Further on they talk about his more recent work on participatory methods that allow many different people to come to a consensus on key issues that affect them. It is not so easy to make such a complex person as Stafford Beer fit into the limited space of an obituary. Personally, I shall remember him as the poet and as the loyal and loving friend that he always was with all of us. In literary terms, he is to me like Oscar Wilde’s “The Happy Prince” who gave away his own fortune to enable other less fortunate people to survive. As they also say in his obituary, Stafford was larger than life, “tall, broad, brimful of energy, and, in later years, bearded like an Old Testament prophet. His enthusiasm for life could be over-powering and quite non-Anglo-Saxon” [5]. The text goes on to say that he had both admirers and critics. I belong to the first group. I think that his critics were unable to see the real man behind his unconventional demeanor. He was loving, loyal, courageous, and deeply human. I loved, respected, and admired him for it. As to Heinz, I share Francisco Varela’s comment about him: “Without his influence and his presence for the past 30 years (in my case they were only nine, but very full years), my life would have lacked a deep, joyous, and nourishing dimension. I call him Heinz the Great”. These were two very exceptional men whose brilliance lit up many lives. I hope that they will rest in peace and that those of us who are still alive will have the courage and strength to continue with their work and the intelligence not to forget what they did for us and for the world.

Notes 1. von Foerster, H. (1992), “Ethics and second-order cybernetics”, Cybernetics and Human Knowing: A Journal of Second Order Cybernetics and Cyber-Semiotics, Vol. 1 No. 1. 2. Beer, S. (1983), Transit, Mitchell Communications, Charlottetown, Canada, pp. 96-98. 3. In “Daily Thoughts for School Administrators”, edited by James Halloran, Salem, VA, q 1979. 4. This anecdote appears in Metaphysics of an Experimental Epistemologist, 1995. This is a paper commemorating Warren McCulloch. Available at: www.vordenker.de/metaphysics/ metaphysics.htm 5. Martin, D. and Rosenhead, J. (2002), The Guardian. EducationGuardian.co.uk

Heinz and Stafford

667

File name{KYB}Issues/33_3-4/Part 4.3d 09:51 – 1/4/04 – Total page(s) 3

PART IV Illustration by Stafford Beer (IV)

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

The footprint of complexity: the embodiment of social systems

The footprint of complexity

671

Raul Espejo Syncho Ltd, Birmingham, UK Keywords Cybernetics, Complexity theory, Communications, Social interaction Abstract This paper deals with organisational complexity, seen from the perspective of its unfolding from global to local concerns. Historically, this unfolding has produced rigid social systems, where those in power positions have forced unfair constraints over the majorities at the local level, and often excluded them. There is a need to move towards flexible, fair, social systems, inclusive in character. This transformation requires an increasing appreciation of communication problems in society and the embodiment of effective social systems. This transformation is presented as a problem-solving paradigm which requires social systems with capacity to create and produce their own meanings, with capacity to manage necessary structural couplings among existing social systems, thus making this management a heuristic to produce necessary social differentiation to overcome communication failures among existing self-producing, operationally closed, social systems. A key construct used in this paper to practically produce this management is the viable system model, developed by Stafford Beer.

1. Introduction As social beings, we produce the social systems we belong to through communications with others. Unfortunately, our social participation in these social systems can have undesirable consequences. At a macro-sociological level, they may produce bureaucracies, oppressive regimes and the like. At a micro-level, they may produce unfair and inflexible enterprises and institutions. The challenge is to learn to intervene in societies in ways that enable and increase flexibility; socially, we need to learn to overcome rigidity and produce fluid systems. Societies are complex, chaotic systems. However, collectives of one kind or another, under pressure, may evolve into social systems with an organisation and capacity to maintain stability far from equilibrium. In nature, organisation emerges from self-organising processes in the form of organisms that co-evolve with multiple others, as complex adaptive systems. On the other hand, human organisation emerges not only from self-organising processes but also from human interventions. As one would expect, people and collectives want to steer social processes according to their purposes and values, however if they do not understand the organisational processes involved – their nature and complexity – the chances are that any steering will produce an unfair and dysfunctional organisation, assuming that an organisation emerges at all. It is important to understand the complexity entailed in the constitution of social systems. It is also important to learn how to manage organisational

Kybernetes Vol. 33 No. 3/4, 2004 pp. 671-700 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523643

K 33,3/4

672

processes and provide closure to our communications and interactions so that desirable social systems are produced. We need to learn to produce desirable systemicity. A first step is to account for the complexity of social systems. This entails far more than accounting for interactions, which at the surface appear as the clearest manifestation of our social behaviour. Often we design or simply participate in conversations with no appreciation of their articulation with the huge complex social context that makes them meaningful. This context emerges from recurrent communications over time, independent of particular individuals, which produce information and social codes, and account for the meanings we use and experience in social life. A purpose of this accounting is to learn how to manage it. In practice, over time, we have learned that there are some strategies that are more advantageous than others to manage this complexity. Strategies have emerged from the co-evolution of social and technological processes. We increasingly, but by no means easily, accept that it makes sense to distribute power, decision-making and implementation responsibilities. It also increasingly makes sense to consider and respect the views of others affected by our actions, or to be inclusive rather than exclusive, or to deal with complete, rather than fragmented, processes and so forth. These are learned strategies that enable our capabilities, and support the development of our potentials, and that most likely are currently increasing our chances to operate in flexible, rather than rigid, social systems. From the perspective of complex adaptive systems, these are all evolutionary strategies that increase the chances for viability. This paper offers Beer’s viable system model (VSM) as one of the most interesting and practical frameworks to support social intervention towards the production of fluid social systems. This model, if understood as a heuristic for desirable interactions and communications, offers a sophisticated and powerful understanding of these evolutionary strategies to manage complexity. The paper ends with an illustration of how to use the VSM in social construction processes. 2. Complex adaptive systems, communications and the viable system model Complexity is a hallmark of modern societies. We all experience this complexity in our daily lives; chaotic agency, rich connectivity and diversity of participants and resources are common. From this interactive chaos some form of order may emerge. Simple interactive rules among participants constitute self-organising processes that produce complex adaptive systems. Learning about this kind of systems has been the focus of much research over the past one or two decades (Mitchell Waldrop, 1993). Complex adaptive systems emerge from self-organising processes among interacting participants,

experiencing stretching demands from the environment in which they realize their relations. These systems develop an organisation that requires external energy and resources to maintain themselves over time and therefore are open dynamic systems, in competition with many others sharing the same ecology, constantly searching for these sources of energy and resources. These are systems that must develop stability far from positions of equilibrium; reaching and accepting to remain in positions of equilibrium would make them nonadaptive, even if they were complex. The complexity of their organisation depends on the complexity of their elementary components, the richness of their connectivity and their functional differentiation. At the same time that an organisation is open to energy and other resources, it is operationally closed, that is, the connectivity of its components produces a closed network, which is, and can be, differentiated from its background, medium or environment. This is what makes them units with their own identity. The stronger the environmental stretching, the more this closed network of relations will have to develop inner response capacity to maintain its unity; the organisation will need further differentiation and thus will need to become more complex. This is the background from where I want to start discussing the issue of social organisations. I assume that enterprises and social institutions are complex adaptive systems, but what are the human costs of achieving this capacity for adaptation and viability? Do they achieve viability at the expense of their individual participants? What is the quality of their organising processes? How can we improve on these processes in order to constitute humane, fair and just, organisations? These are key questions I want to explore in this paper. For this purpose we start with a general discussion of communications and continue with Beer’s VSM. What are communications? Though we can construe the purposes of communications in varied forms, such as sharing mental models, bridging the realities of autonomous agents, creating new insights about shared experiences, enhancing the understanding of phenomena and so forth, we can only experience communications when people – as they share information, express their intentions through utterances and produce some understanding in the receivers – coordinate their actions ( Maturana and Varela, 1992). We know whether there is understanding or not as people recurrently succeed or fail to coordinate their actions. This paper is built on the communication theories of Habermas (1984) and Luhmann (1995). Luhmann considers communications as the elementary components producing social systems. These communications (as opposed to communication acts) are complex units of utterance, information and understanding independent of particular individuals. Therefore, for Luhmann social systems are disembodied social meanings produced by information eigenvalues [1] ( Varela, 1979). The ideas of

The footprint of complexity

673

K 33,3/4

674

this paper build upon this notion of social meanings, but are different to Luhmann’s in the sense that I understand social systems are more than disembodied communications; in my view social systems are the outcome of people and resources in interaction. Luhmann’s theory accounts for the self-organising forces in society where functional differentiation takes place. My proposition accounts for the interplay of self-organisation and design, which also depends on requisite organisations to produce desirable meanings. Organisations are media to create and transmit meanings. Also, in this paper I make use of Habermas’s ideas about communicative action. Habermas seems to be concerned with the quality of people’s interactions or conversations rather than with disembodied social communications. His views emerge from a history of inflexible social systems, where those in power positions through strategic actions have used people instrumentally. His concern is people’s emancipation through communicative actions. This paper argues for the need to complement Habermas’s theory of communicative action with the notions of complexity and requisite variety. The balance between strategic and communicative actions is based on the context of the interactions. There is a social metabolism that needs to be kept in progress and this often makes it difficult, if not impossible, to maintain pure communicative action. What is the VSM? It is fundamentally a model about a system’s strategies to manage complexity to achieve long-term viability. Its origin is in the study of nature’s evolutionary strategies that have produced human beings who are able to maintain viability over time. Human beings have been extremely successful in this endeavour. We have evolved from the most basic elementary components into extremely sophisticated complex adaptive systems, with response capacity to cope with all kinds of unexpected environmental disturbances. What kind of organisation is this that allows us to maintain viability? Stafford Beer argues that the human nervous system offers an answer to this question; this system shows us how multiple interconnected checks and balances and distributed decision-making, problem-solving and implementation processes can produce a cohesive system with capacity to deal with a large environmental complexity (Beer, 1981). I argue in this paper that this model gives us a methodological answer to the problem of embodiment of social systems and offers a pragmatic approach towards people’s emancipation. 3. Social systems, communications and complexity 3.1 Society and the individual: from rigid to flexible social systems Mrs Thatcher’s 1980s remark that “there is no such thing as society” is well known. Presumably this was her attempt to establish the supremacy of the individual over the collective. The spectrum of Orwell’s 1984 was too close and all too powerful bureaucracies were feared as social evils. Bauman (2000) gives a powerful account of the origin of these fears:

That heavy/solid/condensated/systemic modernity of the “critical theory” era was endemically pregnant with the tendency towards totalitarianism. The totalitarian society of all – embracing, compulsory and enforced homogeneity loomed constantly and threateningly on the horizon – as its ultimate destination, as a never-fully-defused time-bomb or never fully-exorcised spectre (Bauman, 2000, p. 25).

In particular, during the second part of the 20th century, individuals in society started to assert their rights in response to oppressive regimes and enterprises. These regimes and enterprises had emerged to make manageable social complexity; they were necessary to reduce the complexity of these individuals by standardising their actions and making them more predictable. The ruling classes had to exercise their power to maintain social complexity within their span of control: That modernity was a sworn enemy of contingency, variety, ambiguity, waywardness and idiosyncrasy, having declared on all such “anomalies” a holy war of attrition: and it was individual freedom and autonomy that were commonly expected to be the prime casualties of that crusade (Bauman, 2000, p. 25).

Expressions of diversity were threatening and had to be contained; creativity was a dangerous attribute leading to uncontrollable variety. This was particularly the case when creativity came from people in the shop floor, in the emergent manufacturing industries of the late 19th and first part of the 20th century. Autonomy at that level was synonymous to anarchy and disrespect for the authority. Organisation of social activities was a zero-sum game where those in power positions had to win their game vis-a`-vis the rest to maintain order in society. Among the principal icons of that modernity were the Fordist factory, which reduced human activities to simple routine . . .; Bureaucracy, . . . in which identities and social bonds were deposited on entry in the cloakroom together with hat, umbrellas and overcoats, so that solely the command and the statute book could drive, uncontested, the actions of the insiders as long as they stayed inside; Panopticon, with its watch-towers and inmates never allowed to count on their surveillants’ momentary lapses of vigilance; Big Brother, who never dozes off, always keen, quick and expeditious in rewarding the faithful and punishing the infidels; and – finally – the konzlager. . . the site . . . where all those presumed not to be or found not to be malleable enough were doomed to perish of exhaustion or sent to gas chambers and crematoria. (Bauman, 2000, p. 25; my italics)

As the idea of a social system became connected to these excesses, the seeds for individualism were sown. However, these extremes cannot blind us from our daily experience that we are social beings that realise our potentials in interaction and communication with others. It is in these interactions and communications that we create and produce social meanings that go beyond us as individuals. Indeed, collectives may develop identities of their own. To a large degree, these identities are the outcome of self-organisation processes responsible for the emergence of social systems, but contrary to what is the case with other social collectives in nature we can purposefully interfere with these

The footprint of complexity

675

K 33,3/4

676

self-organising processes to construct enterprises and institutions of our own choice. There are multiple advantages in doing this; we can speed up social learning and accomplish complex tasks in much shorter times. However, we often achieve all this, as suggested in the quotes above, at the expense of the individuals who are grinded by the wheels of relentless, powerful institutions. The recognition of all this and the need for individual emancipation in these oppressive conditions was the trigger for much of the critical thinking produced in the latter part of the 20th century. But at present, we need a new kind of critical thinking, one that supports individual emancipation operating within desirable social systems; these are social systems constituted by fair interactions and communications that enable people to realise their potentials. These are liberating, robust, flexible and adaptable social systems that are accountable to all their stakeholders as well as for their ecosystems. Proposing this kind of emancipation is easier to say than to do and implies much effort and learning but it is unavoidable in social evolution. Not to achieve a fair balance between individual and collective interests, in a highly interconnected world, is a recipe for social instability and major breakdowns. A better understanding of how complexity is managed in successful complex adaptive systems may well give us useful pointers for improving the processes leading to the constitution of fair social systems. In fact all the evidence, as implied by Bauman, suggests that we have been constructing social system without much understanding of the underlying processes producing them. Those wielding resources, in position of power, submit the most to their intentions, purposes and values and hope for the best, as workers make things happen. In these situations, people have to find ways to coordinate their actions and perform as required within the imposed constraints by those wielding power. As they carry out their tasks, a good deal of self-organisation and self-regulation may take place, however often at a high cost to them. The challenge is to learn how to combine effectively organisational design with self-organising processes. Rigid, unwieldy, solid social systems, with constraining structures, need to be replaced by fluid and flexible ones, able to accommodate the natural diversity of people. This transformation from rigid to flexible systems has important social consequences. If the ethical/political discourse of collectivists was that of a “just society”, and of individualists was that of individual “human rights” (Bauman, 2000, p. 29), we need at present a participatory discourse focused more precisely on the constitution of fair societies that recognise individuals with rights and responsibilities, that is of societies and organisations that recognise the difference between individuals and citizens. The “citizen” is a person inclined to seek her or his own welfare through the well-being of the city – while the individual tends to be lukewarm, sceptical or wary about “common cause”, “common good”, “good society” or “just society” (Bauman, 2000, p. 36).

Indeed, citizenship is a requirement that applies to societies and enterprises alike. Of course, individuals can always step out of society, however they do this at a cost both to them and society. There is a need to redress apathy and defection for the good of all. In other words, there is a need to strengthen the systemicity of society. This is a systemicity that maintains social cohesion at the same time of enabling its components’ autonomy. The duality cohesion-autonomy is a requirement to make social and individual learning more responsive and effective. The clarification of this duality in terms of managing complexity is perhaps a key contribution of the VSM which gives us a heuristic for transforming rigid and unwieldy social systems into flexible and malleable social systems. Indeed, it is a heuristic for emancipation. The challenge is to clarify the epistemological scope of its contribution and the methodological requirements to make it practically relevant. These two aspects are the concern of the rest of this paper. 3.2 Understanding the role of communications in social systems Though rigid, inflexible, hierarchical and exploitative social systems are common at present, it can be argued that they are less so than in the past; social learning is in progress. Indeed these are not inherent characteristics of social systems but the outcome of contingent, largely misconceived, efforts to maintain cohesion in situations running out of hand. Systems emerge one way or the other from the interactions and communications of members of collectives; they produce meanings that belong to the collectives and not to the individuals. The problem is how to avoid their oppressive versions. Rigidity and inflexibility may emerge from unguided interventions in organisational processes that we do not fully understand. There is much to understand about “human social systems”. Under which conditions can we accept systemicity and well-defined boundaries for institutions? Is it not that people and groups regularly defect, whether physically or in their commitment, and therefore that in operational terms necessary connectivity does not evolve making the boundaries of any social system far less clear than those implied by institutional boundaries? As stated by Anthony Giddens in the introduction to his book The Constitution of Society: We have to be very careful indeed with the concept of “social system” and the associated notion of “society”. They sound innocent terms, and they are probably indispensable if used with the appropriate measures of caution. “Society” has a useful double meaning, which I have relied upon – signifying a bounded system, and social association in general. An emphasis upon regionalization helps to remind us that the degree of “systemness” in social systems is very variable and that societies rarely have easily specifiable boundaries – until, at least, we enter the modern world of nation-states. (Giddens, 1984, p. xxvi).

This quotation makes apparent not only the very important issue of the systemicity of social situations but also the need to have a flexible

The footprint of complexity

677

K 33,3/4

678

epistemological tool to work out the boundaries for social systems. It is desirable to specify boundaries for social system not only at the macro-level but also at all possible levels of association producing emergent meanings. This is necessary to clarify relations of belonging and cohesion and thus avoid undesirable constraints. The inclination is to look for physical boundaries, like the city. A city has a concreteness that is appealing and operationally practical. Indeed I used it in Section 2 to contrast the non-systemic behaviours of people driven by individualism to the apparently more systemic behaviour of citizens. In fact it would have been useful to have a more general metaphor, one able to reflect the notion of operational belonging, of active constitution, beyond the specific case of people constituted as citizens. Beyond the idea of boundaries for a city we naturally think of boundaries for enterprises, institutions, social bodies and even for teams as instances of social systems with more or less well defined boundaries. However, Giddens only appears to feel comfortable with the notion of a boundary when reaching the concreteness of the nation-state. This is an issue that needs methodological clarification. Perhaps our confusion starts with the definition of the elementary components of a social system. It is tempting to say that an individual is this elementary component. In other words, if we define a social system as “a network of interrelated individuals whose interactions produce a whole or totality that constitutes them as its components” it is tempting to infer that the elementary components of this social system are the individuals themselves. Two clarifications are proposed. (1) It is natural for individuals in a collective to “come and go” and therefore we cannot assume specific individuals producing the social system, so really we are talking about individuals who fulfil particular “roles” in the relations of production of the social system; these individual roles, and not the individuals in their entirety, produce the social systems and are constituted by it. (2) Another clarification of this definition is that individuals’ interactions cannot account for the full complexity of the social system. This is a major issue that has fundamental consequences for organisational design. Most coordination of actions among the members of a collective is specified by their social context rather than by their direct interactions. Interactions account only for a very small part of the complexity of their relations, while contextual communications account for the large majority (Conant, 1979). In Conant’s terms, most of the communication among participants in a situation takes place without channel capacity among them. In other words, assuming recurrent interactions among the members of a collective, to the point where they

produce a social system, their interactions account only for a small part of the complexity of their relations; their contextual communications account for most of this complexity.

The footprint of complexity

3.3 Luhmann and disembodied social systems Luhmann (1995) captured, albeit only partially, the second point above when he defined social systems In his framework, social systems are produced by communications and not by communication acts.

679

Although differently inspired, Luhmann’s 1984 sociology can with hindsight be considered as another attempt to bridge the gap between action theory and social systems theory. During the 1960s and 1970s, leading scholars among symbolic interactionists and Parsonian systems theoreticians had recognized the “incommensurability” of their respective perspectives Grathoff 1978. Luhmann suggested resolving this debate by considering the core concept of symbolic interactionism, that is, the interactive construction of social meaning, as the unit of operation of social systems. Whereas Parsons 1937, had considered action as the unit of the system’s operation, Luhmann’s social systems theory thus provides a mirror image of Parsons’s so-called “structural functionalism.” The analysis of social structure should not be based on (the aggregate of) action, but on the interactions between actions (Leydesdorff, 2000).

This is an important proposition. Social systems emerge from the information content of the recursive operation of communication acts and not from these acts. A social system is independent of particular individuals, who can be any as long as their utterances and understandings contribute to eigenvalues of meaningful social information. A social system emerges from this information, which in effect is the synthesis of myriad of communications. It is apparent that in Luhmann’s terms a communication is the unity of information, utterance and understanding. Luhmann defined communication explicitly as the unity of information, message, and understanding. By taking the communication of an information and its reception as a single – albeit complex – unity, . . . the concept of “meaning” could be made constitutive of his idea of a social system (Leydesdorff, 2000).

The key of all this construction is the production of a collective meaning, beyond individuals. Hence, when we talk about the “education system”, we are not talking about the particular members of educational institutions producing educational acts (as well as many other types of acts), but about the socially shared meanings – eigenvalues – emerging from educational (communication) acts produced wherever takes place what the collective means by education (e.g. in schools, homes, libraries, museums, parks . . .). It is this meaningful information or more accurately the infinitely large set of communications producing a social meaning that produces a social system. In this sense, social systems are disembodied, they are meanings produced by communications, that is, by intentional utterances carrying information that are recurrently understood by social actors. If there is no one to understand the

K 33,3/4

680

utterance, whether as intended or not, it would not be a social act. But, reiterating, what produces the social system is communications and not communication acts. Assuming that this understanding of Luhmann’s work is correct, I can infer that social systems in his view mean domains that allow us to communicate without channel capacity; are the domains of self-referential, meaningful information, that we share in our daily drift. But, and very significantly, though shared information may account for the largest source of complexity that we experience in social life, it does not account for the small, but very significant, part of the social complexity that we experience in everyday ongoing conversations or interactions. Interestingly enough, interactions’ relatively small complexity may appear to us in psychological terms (the consciousness system in Luhmann’s terms) as the most significant source of complexity. This is reflected in the great relevance that we give to dialogues and conversations in all kinds of situations where we are concerned with fair participative processes. This point is discussed later in the paper. The above discussion should help us to recognise not only the difference between communications and interactions, but also that in Luhmann’s terms, the basic components of a social system are communications. However in this work, I consider communication in generic terms, including both communications with channel capacity, that is interactions, and communications without channel capacity, that is the sharing of stable contextual information. Once interactions are accepted as constitutive of social systems, their components cannot be disembodied communications alone but also embodied roles in interaction. The problem with Luhmann’s framework is that, in general, since communication acts producing a particular social meaning may be distributed in all parts of society it may become very difficult to trace all these sources and account for them in an embodied social system and therefore methodologically it seems more sensible to disembody it and focus attention on meanings and the codes producing them. This disembodiment offers several observational advantages. It is currently possible to observe communications that produce a social system that constitutes these communications as its elementary components or in other words it is possible to observe a social system that produces the communications that produce the system in an ongoing recursive process that produces a particular social meaning; this is the case of an autopoietic system (Luhmann, 1995). This is a system that produces itself (Maturana and Varela, 1992) that is observable and measurable and this has significant value for social studies; for instance it can allow us to measure the information content underpinning the interactions of a collective. But, communications among autopoietic systems is a major challenge indeed. It seems useful to visualise further the communication challenges within and among social systems. As discussed earlier, when we interact with

each other we produce shared linguistic codes. These codes are complex expressions of intentions and expectations. A particular message not only reduces uncertainty by selecting one among a set of possible messages but it is also a vehicle to express our intentions, purposes, values and expectations. This is the illocutionary force of a message (Searle, 1979). As these expressions recur, these codes become independent of particular individuals; communications become eigenvalues (stable encoded social information) that produce particular social systems. The important point here is that the structural deepening of these codes is responsible for social complexification. This implies, as explored later in this paper, possible unfoldings of social complexity, in the form of either new integrated social systems emerging from several existing systems or new socially differentiated systems constituted within a social system and so forth. But most significantly, this ongoing process producing multiple social meanings is mediated by a wide range of multi-cultural projects, institutions and the like, which act as channels of communications between the participating people. Hence, when individuals contributing to particular social systems (e.g. education, transportation, etc.) interact with each other, they not only transmit ad hoc information but also communicate their socially accumulated information codes, in the form of intentions and expectations. The participants’ decoding of the others’ codes is indeed a significant communication challenge, which may be sorted out over time by self-organising processes, assuming that they will find the need to remain in interaction. New social systems indeed emerge from these processes. However, these processes if left to a chaotic co-evolution may take too long, not to mention may be too costly in human and economic terms. Organisational design should help to make these communication processes more effective. 3.4 Social systems design: the clarification of meanings through (embodied) relations Luhmann’s solution in order to account for the complexity of social processes appears to concede that there is little we can do in terms of organisational design; how can we create and design an institution that avoids oppressive, hierarchical, structures? How can we intervene to avoid many undesirable, if not inhuman, social projects (cf Bauman above)? Answers to these questions require embodied social systems that allow us to intervene in social processes, as I explain later. In order to understand this issue of embodiment it is necessary to make the distinction between informational and operational domains. It is useful to distinguish between meaning creation and meaning production. Creating a meaning is in the informational domain, producing a meaning is in the operational domain (Espejo, 2002). Creating a meaning relates to an espoused purpose independent (to different degrees) of what is happening “on the ground”, in the ongoing communications of the people affected by this

The footprint of complexity

681

K 33,3/4

682

meaning. On the other hand, producing a meaning is the emergent and evolving meaning of the myriad communications “on the ground”, as experienced and appreciated by the relevant stakeholders. It is apparent that Luhmann’s social systems are emergent meanings constituted by evolving communications. These are produced meanings, or eigenvalues, in the operational domain of a sometimes highly dispersed collective. But, what is particularly human is that for the same collective created meanings may not correspond with produced meanings; we are self-conscious individuals with creativity and imagination. This gap between created and produced meanings may be the trigger for policy and political projects that catalyse a collective’s purposeful behaviour. In other words, the question arises of whether it is possible to produce desirable meanings that so far, perhaps, have only been articulated by philosophers, politicians and leaders, but have little to do with emerging meanings that sociologists may observe in particular information footprints. This is the challenge of linking the informational and operational domains. For as long as we fail making the structural links between these two domains the eventual production of a meaning will be the outcome of self-organisation alone (disembodied social systems) and not of both self-organisation and purposeful organisational design (embodied social systems). Therefore, producing desirable meanings is the potential pay-off of working out the embodiment of social systems. It is interesting to observe that the evolution of an embodied social system is the outcome of two different forms of social coupling. It is the outcome of the structural coupling of interacting participants, that is, of the history of their recurrent interactions leading to structural congruence (Maturana and Varela, 1992), and also of the operational coupling among non-interacting people who share the codes of the social system. In fact, the latter coupling is the most significant when we think about coordination of actions within a system. Quite naturally, as implied earlier communications based on operational couplings are less likely to take place among differentiated social systems. However, we may expect a degree of overlap between social systems and therefore that some form of operational coupling takes place anyway among the people embodying them. But the purposeful development of these overlaps requires concrete structural (coupling) mechanisms such as ad hoc, purposefully created institutions integrating different social systems. The creation and production of these mechanisms is the scope for both social development and organisational design. A most significant example of this point is the role played by the institutions of the European Union in bridging cultural and social gaps between nations in the continent. A key aspect in the creation, rather than production, of desirable meanings is interactions, whether as debates, dialogues, or any other form of conversation. It is this fundamental need for the creation and clarification of social meanings that triggers the need to understand how to make our social interactions

effective. This is an important aspect of communications in general, and emancipation in particular, which has been explored in-depth by Habermas (1984) in his work about communicative action. According to this work, the purpose of communication (in the sense of interaction) is clarification, which is recognised as communicative action and not as manipulation or strategic action. The speaker makes three validity claims that he should be prepared to redeem at any stage of the interaction: (1) that what he is saying is the truth, (2) that he is sincere/authentic, and (3) that what he is saying is legitimate (ethical). The problem is that however desirable this might be, speakers, particularly in decision-making situations where they interact with many people at the same time, often are not able to afford communicative action. Forced to make a decision, even with the best of the intentions, a speaker is likely to manipulate the others, particularly if he is in a position of power. Communicative action requires speakers with a good deal of available complexity to engage in interactions (i.e. with at least as many possible states as relevant states to the conversation are generated by the other participants), something that structurally most of the time is unavailable to him. In fact, the quality of his interactions in general depends on the quality of this structural underpinning. It is crucial to design conversations in such a way that the ideal of communicative action is approached, and for this requisite structures are necessary. The design of conversations will also depend on whether the participants share or not a common social system. If those in interactions are operationally coupled through a common social system (e.g. are members of the same profession) the chances are that they will require far less channel capacity in their interactions than those who are not operationally coupled and therefore who have to overcome far more communication hurdles in order to clarify what they mean. This is a major communication issue; the creation of a new policy issues implies the creation and production of a social system whose boundaries are defined by the social codes emerging from the intersections of multiple social systems as communicated by the intentions, purposes, values and expectations of all stakeholders in recurrent interactions. This is a problem of incursion among social systems (Leydesdorff, 2000) or of orthogonal interactions among social systems. The complexity of these incursions is often overlooked in the design of dialogues or conversations in general (cf. Riscom project – Espejo, 2001). This lack of “requisite variety” (Ashby, 1964) in conversations poses significant ethical problems, which correspond to what some philosophers have labelled “discourse ethics” (Apel, 2001).

The footprint of complexity

683

K 33,3/4

684

4. Requisite organisations: the embodiment of social systems 4.1 Requisite organisations for problem-solving Thus when faced to environmental complexity collectives often need to work out organisational responses that allow them to articulate these environmental challenges, that is, to create new meanings, and to produce them (i.e. produce a new social system). The challenge for the collective is to create an organisation to respond to the multiple agents challenging the production of its purposes (i.e. meanings). This is the ingenuity gap between a collective and the multiple environmental agents stretching this collective (Holmer-Dixon, 2001). It is in this endeavour that we may create new institutions and institutional responses. If the solution implied is more of the same it would hardly be a significant challenge requiring ingenuity; single loop learning would be enough (Argyris and Schon, 1978). If the solution requires creating and producing a new meaning we would face the complexity of a new social system with a requisite organisation, capable of connecting all resources, distributed or not, creating and producing this new meaning. The problem is that seldom these responses are systemic in the sense proposed here; they seldom articulate, or more correctly integrate, (all) the implicated resources, thus often leaving these tasks to lengthy, painful and costly self-organising processes. Indeed the institutional responses may become expensive trial and error problem-solving efforts, which over time may produce a more or less effective social system that deals with the environmental complexity that triggered the problem-solving effort in the first place. Social learning will always be necessary; in fact most of the time we do not even know what the problem situation is, let alone how to solve it. This indetermination poses the question: How do we know which is the body (requisite organisation) for the meaning it has yet to be created, let alone produced?

But realising this indetermination is not a good reason to leave all action to trial and error alone. Nature has produced a most remarkable system, with capacity to create and produce ingenious solutions to environmental demands. We humans have achieved, through our nervous system, the most remarkable capacity to link effectively our informational and operational domains. Our nervous system is a most powerful recursive learning system. Learning how this system manages complexity should give a powerful heuristic for social problem-solving. This is what the VSM offers. An institutional response should be more than a communication’s bridge between several or many disembodied social systems as suggested by Luhmann’s theory; it should be an embodied social system with capacity to create and produce its own meanings. This solution demands solving significant communication challenges; first it requires providing the context to bridge the differentiated codes of multiple social systems intersecting with the one yet to be created and produced (indeed we may expect professions and practices of different kinds involved in a particular problem-solving situation)

and second providing the capacity to hold conversations with requisite variety for a situation yet to be articulated (indeed we may not know what is the meaning we want to produce). These are characteristic circular situations that depend on bootstrapping operations to move ahead. This bootstrapping suggests the need for a methodological solution. But, let us first clarify the idea of a requisite organisation. From the perspective of organisational cybernetics, the creation and production of a social meaning (e.g. an education social system) poses the problem of a “requisite organisation”, able to both integrate existing background meanings (these are produced rather than espoused meanings) and support communicative action among those participating as stakeholders (with their espoused meanings). Now, we deal back with the problem of the embodiment of social systems. What is the requisite organisation able to produce a desirable meaning? For instance, if the concern is education, it is critical to understand not only the interactions involved in the creation of the policy but also the communications entailed in the production of the meanings implied by that policy. The concern is connecting meaning creation with meaning production, that is, connecting the informational and operational domains of the collective bootstrapped by the meaning yet to be formed and produced. Linking the informational and operational domains provides an epistemological grounding to the VSM. Once we are focused on the creation and production of a meaning we are discussing about an embodied social system. From Luhmann’s work, we take that social systems are autopoietic systems, that is, self-producing systems. The fact that we can observe stable meanings that go beyond particular individuals suggests that the systemicity of social activities is a tenable proposition. Indeed, for instance in education, we can observe not only eigenvalues/behaviours (e.g. the footprints of education in myriad of educational aids, tools, instruments, journals and publications of all kinds) but also (disembodied) relationships which encode (as a result of recursive operators applied to our interactions) people’s intentions and expectations as they engage in communication acts; content and identity allow us to make the educational distinction in the huge background of social activities; this is a disembodied education social system that we experience in every day life. This is a system with identity and content, but not with structure. It is this structure that is required to link the informational and operational domains, the one that the VSM helps us to recognise in social systems. It offers a structural model of how interactive complexity should be distributed and managed in a social situation as the involved collective strives to give viability to a policy issue in a complex environment. The point is that unless this structure works effectively the cost may be too high for the social system to emerge, and produce itself over time with adequate energy and resources, thus making the policy viable.

The footprint of complexity

685

K 33,3/4

686

Regardless of whether we are talking about interactions among cells or people, their collective tasks, vis-a`-vis agents in their environment, emerge from their inner connectivity, that is, from the capacity of their communications to produce those tasks. In evolutionary terms it is apparent the extraordinary complexification of these tasks and how these strategies for managing complexity have had to increase in sophistication. The human nervous system is indeed the pinnacle of this complexification. However, as discussed earlier, there is something special about being human. Our language not only contains information (e.g. a message where to find food) but also meaning (e.g. for what purpose). Humans have managed to decouple their operational domain from their informational domain, something that other mammals apparently have not achieved, and in the process have become purposeful rather than just purposive. Only for humans we can say that if the channel capacity of their collective nervous system (whatever that means) is insufficient for the purpose they ascribe to their collective task, their performance will suffer, that is, their coordination of actions will be inadequate to produce the task. The other side of this proposition is that humans may consciously develop the “bodyhood” required for their self-defined purposes (Espejo, 2003). 4.2 Evolutionary lessons to manage complexity Here, understanding the human nervous system as nature’s evolutionary response to this decoupling of the informational and operational domains is of great significance. Perhaps the three constructs that emerge as crucial to understand this management of complexity and related embodiment of viability in a structure are organisational closure ( Maturana and Varela, 1992), ultrastability and structural recursion ( Beer, 1979). The first is the mechanism related to self-awareness and the closing of interaction loops among the components of a network. This is a mechanism underpinning structural determination and redundancy of potential command. The behaviour of a complex adaptive system is determined by its own structure, that is, by its internal connectivity, and not by the disturbances it receives from its medium. This insight can be appreciated from the following quote with reference to our central nervous system (CNS): Consequently, the synaptic gap can be seen as the “micro-environment” of a sensitive tip, the spine, and with this interpretation in mind we may compare the sensitivity of the CNS to changes of the internal environment (the sum-total of all micro-environments) to those of the external environment (all sensory receptors). Since there are only a hundred million sensory receptors, and about ten-thousand billion synapses in our nervous system, we are 100.000 times more receptive to changes in our internal than in our external environment. ( Von Foerster, 1984, p. 300).

Additionally, decisions in the CNS are made wherever the necessary connectivity to trigger a response happens to emerge and not in a predefined

place (Beer, 1979). In other words, in the CNS there is redundancy of potential command; there is a huge structural flexibility in response capacity. The second is the mechanism that allows the body to maintain critical variable (like for instance body temperature) within critical limits even under severe environmental variations. This mechanism emerges in the body as an outcome of a wide range of checks and balances that allow its multiple components to adjust their points of stability as changes in neighbouring components occur. The third is the mechanism of structural recursion, which distributes sensitivity and response capacity to environmental disturbances throughout the body, thus increasing its chances of coping with large external complexity. Most significantly, in addition to the conceptualisation of these mechanisms, Beer offers variety, or the number of possible states of a situation, as a metrics to study the performance and quality of their operations. This metrics together with Ashby’s Law of Requisite Variety (Ashby, 1964) offers a powerful approach to study the management of complexity. Ashby’s law states that in a particular situation, a regulator to succeed in its regulatory task needs to have at its disposal at least as many responses as disturbances are likely to take the regulated situation out of control (e.g. in the case of the human body, unless its regulatory system produces as many responses as disturbances threaten to take its performance out of physiological limits). These communication requirements for viability, since they are driven by complexity management requirements and not by the specific nature of the system, also apply to human social systems as long as their systemicity is guaranteed. But, is it not the case that individuals may not align, if so they wish, their interests with those articulated by politicians and managers (i.e. those who take collective responsibilities)? This may be so, but as long as a new meaning emerges there must be communications and roles producing it. There must be a human autopoietic social systems, however unclear its embodiment might be. Indeed, there is much to learn about this embodiment and we will need to develop powerful observational tools in order to test empirically this proposition, however it seems to me to be logically tenable. Social systems are viable wholes with their own identity, which require maintenance in time, that is, which require resources and energy. What is still under defined is their embodiment, however, this is what the VSM offers us as a heuristic. In evolutionary terms, it is necessary to understand better how organisational closure, ultrastability and recursion can evolve humanely in social systems, respecting the autonomy of individuals and opening the space for cohesive collectives, with rights and responsibilities. In fact, the big issue for social systems is not systemicity that as I have explained before is entailed in the emergence of collective meanings, and in any case, is often forced upon people by those in positions of authority, but that of ethics. It is interesting

The footprint of complexity

687

K 33,3/4

688

to note that nature, through its evolutionary learning of how to manage complexity, also has something to teach us in the ethical domain. When their interactions do not have requisite capacity for producing ascribed purposes, those in power positions can, and often do, manipulate those on the other side. This is a daily experience for a large number of people as they constitute rigid, inflexible systems. The challenge is to learn how to give requisite capacity to these interactions. 5. The viable system model Systemicity in societies may emerge from the co-evolution of relations of production between members of a collective and agents in the environment who put pressure on them. This stretching may trigger relations of constitution of an organisation among the members of the collective, as they engage in increasingly demanding processes of meaning creation and production, eventually emerging from the collective an organisation with identity, content and structure. However, this transformation of a collective into an organisation is not always successful and often we encounter collectives that fail in producing social systems, and which either require distributed resources beyond them to achieve the closure of a social system or simply fail to achieve the necessary connectivity. There is nothing to suggest that collectives will naturally succeed in evolving into organisations. Indeed, achieving organisational closure is a tall order in any case. Connectivity among the members of a collective implies a structure. What is this structure? It is a stable form of interaction that allows people to operate together as a whole. Structure is therefore a particular configuration of relations, that is, of embodied relationships, underpinning the creation and production of a social meaning. Since an embodied social system emerges mainly from the self-organisation of distributed resources, it is unlikely that its boundaries will coincide with those of the institutions or bodies that society formally creates to bridge a problem-solving gap [2]. It is the structure of the social system and not of the formal institutions or bodies that is of interest to us in studying requisite organisations. But a social system can be more than self-organised, it can be self-constructed [3] since its components can be self-aware of the social processes in progress and can influence consciously their production, through for instance, organisational design. However, without functional capacity, in the form of resources and energy, self-construction of a requisite organisation will not take place. The VSM offers a heuristic to produce the requisite organisation for a social system. Its basic assumption is that the different and perhaps highly distributed resources focused on a shared problematic situation, will end up relating to each other if they succeed in bootstrapping the production of a desirable meaning. This implies producing an autonomous system, with the

capacity to create, regulate and produce that meaning. These resources can be distributed individuals, members of small collectives, large companies, regulatory bodies, government institutions or whatever. This assumption seems to be reasonable, since it implies the expectation that relevant resources will find means to relate to each other, in such a way that they solve a socially relevant problem, reducing fragmentation and dependence on unnecessary weakly related resources. Achieving this autonomous social system is often the outcome of lengthy, chaotic learning processes, which need to be made less painful and costly; this is the meaning of designing flexible/liquid social systems. Autonomy in this context means accepting responsibility for one’s affairs within the framework of being a system that is part of one or more larger systems. Indeed, meanings emerge in a context; they are the outcome of processes of functional differentiation. Those systems producing a meaning usually contribute, wittingly or not, to produce wider networks of meaning creation and production through their interactions with other autonomous systems and they are produced by the interactions of more specialised networks of meaning creation and production. These contributions are likely to be messy processes that require competent observers to be unravelled. However, structural differentiation occurs wherever we succeed in producing [4] a stable (far from equilibrium) meaning. These are chaotic and complex processes of synthesis and branching that can fold and unfold in multiple and varied forms; these are processes of unfolding complexity that trigger the development of new relations and different distribution of resources. We can hypothesise the emergence of different systems and observe on the ground whether they fulfil the conditions of viability, that is, of being embodied social systems. Most of these hypothesised social systems will run short of the requirements for viability as autonomous units; this is the space for organisational design. If we use the example of local transportation as a system-in-focus, the hypothesis implies that there are resources in a given locality to create, regulate and produce its own transportation policies. A larger, embedding system could be the system for regional transportation with capacity to create, regulate and produce its own meanings, responsible for the wider policies within which several local transportation systems produce their own policies and the regional policy. Equally, within these local systems we may expect to find self-organising teams creating and producing specific products/services for their communities (e.g. country roads, bus services, traffic management, etc.) and together producing the local policy. These are primary activities relevant to the hypothesized organisation for this policy issue. This devolving strategy assists collectives in coping with chunks of their environmental complexity. Observing this “Unfolding of Complexity”, or cascading structure, in recursive levels consisting of autonomous units within autonomous units (Figure 1), is

The footprint of complexity

689

K 33,3/4

690

Figure 1. Unfolding of complexity

a way to check the coherence of the organisation’s created meanings (as languaged by participants) and the evolving body producing them. Together they should produce the requisite organisation, which has coherent informational and operational domains. This hypothesis could be easily falsified if it is observed, for instance, that the local transportation policy is defined at the regional level and therefore that there is no such thing as the local transportation social system. Lack of coherence may suggest either revising the hypothesised unfolding of complexity or redefining the system’s espoused purposes (i.e. created meanings for the organisation) or developing an alternative structure (i.e. body). In Figure 1, we can see in operation the bootstrapping of purpose and identity through structure. The hypothesis is that each autonomous unit is a viable system, in the sense that they can sustain themselves in time in spite of unexpected environmental disturbances; they are ultrastable. For this purpose, they need energy and resources, which in societies are provided by and through institutions and bodies of all kinds. People are constituted as roles in the embodied system. As discussed earlier, these roles are the elementary components of the social system. It is individuals and groups constituted as roles that transform disembodied relationships (encoding social meanings) into embodied relations, disembodied social systems into systems with identity, content and structure. These individuals/groups can be any as long as they produce the relations that produce the social system. By definition, an autonomous, viable system is one that has roles (people/groups) with the capacity to create, regulate and produce

its meaning. These are the systemic roles producing a social system. This idea is further expanded below. From the VSM, we know that two organisational processes produce an autonomous unit (circles in Figures 1 and 2); these are the adaptation and cohesion processes. These processes emerge from relating the resources producing five systemic functions; Policy, Intelligence, Cohesion, Coordination and Implementation, which together create, regulate and produce its products/services (Espejo et al., 1999). Anyone recurrently contributing through his interactions to the creation and production of a meaning, however distant or close might be, is constituted as a role or roles in one or more of these five systemic functions. In the end, all human resources can be mapped onto these functions. This is the great generative power of the VSM. The adaptation process relates policy, intelligence and cohesion resources. Policy creates the policy issue as an outcome of managing the interactions of intelligence and cohesion resources. The intelligence function is concerned with the “outside-and-then”, that is, with the long-term, considering the organisation’s problematic environment. The cohesion function is concerned with the “inside-and-now”, that is, with the organisation’s current achievements in its operational environment.

The footprint of complexity

691

Figure 2. VSM as the embodiment of social system

K 33,3/4

692

The cohesion process is focused on balancing local autonomy and organisational cohesion. For this purpose, the cohesion function manages its interactions with implementation with the support of coordination resources. Some degree of non-negotiable corporate intervention [5] needs to go hand in hand with resources bargaining to enable primary activities to create and produce their own autonomous tasks. Interactions between those in primary activities, with the local knowledge of their tasks, and those in the cohesion function responsible for the cohesion of the system are crucial and cannot be based either on excessive intervention or naı¨ve trust about the competence and sincerity of each other. Sporadic, but ongoing, mutual audits are necessary to build up responsible trust (cf. orthogonal interactions below). Additionally, the coordination function is concerned with local interactions among primary activities, balancing their activities and efforts; it is fundamentally an anti-oscillatory function. For this it needs to operate within the framework of shared operational standards defined as rules by the cohesion function. The lower the operational diversity and variability in those aspects that are not central to primary activities’ individual purposes and values (i.e. self-constructed meanings), the easier will be their mutual coordination and therefore less corporate intervention will be required. This design leaves free from intervention, and open for negotiation, those aspects that are central to their autonomy; this is the mechanism to align meanings. The implementation function produces the organisation’s meanings. The need for alignment of meanings is perhaps a key difference between organisms and organisations, while an organism’s components do not have meanings of their own, an organisation’s components do have meanings of their own. An organisation is the outcome of several if not many (autonomous and diverse) social systems, which interact through the structure of that organisation. Intersecting social systems provides the communication background to all the above structure of interactions. Their interactions are orthogonal in the sense that people have to cross the boundaries of diverse meanings in order to build up the necessary structural couplings for recurrent interactions in time. Orthogonal interactions are necessary to produce new meanings, that is, to produce a new social system; this is in contradistinction to people’s interactions when they are operationally coupled by sharing the same social system. Therefore, key aspects contributing to effective organisational processes are first, the quality of incursion, that is, the quality of the structural couplings built up between the so far non-intersecting social systems (e.g. in the case of nuclear waste management, between, among others, the scientific, technical, entrepreneurial, governmental and non-governmental (NGO) social systems) and second, the quality of stakeholders’ operational coupling through the new social system, that is the quality of the structure enabling stakeholders’ communications beyond their interactions. Should these aspects fail or be weak

no amount of interactions (e.g. dialogues and conversations) are likely to overcome fragmentation. The same organisational processes relating the five systemic functions recur in all embedding and embedded primary activities (see isomorphic interactions in Figure 2), as requirements for their viability. In summary, the VSM is a powerful construct to study the embodiment of a social system assuming that the people involved have the bridges to cross each others’ boundaries as influenced by the relevant social systems they have embodied historically. The power of the idea of requisite organisation is that it allows us to visualise a two-way bootstrapping of meaning creation and meaning production and therefore, it makes apparent, necessary communications and interactions in order to produce a new meaning. Illustrating a requisite organisation is the purpose of Section 6. 6. Applying the idea of an embodied social system The next issue addressed in this paper is the application of the idea of a requisite organisation in a concrete social situation; how can we make use of it? Its most comprehensive application so far has been in nuclear waste management in several European countries (The RISCOM Project) (Espejo, 2001, 2003; Wene and Espejo, 1999). The Viplan methodology (Espejo, 1993; Espejo et al., 1999) has been used in this application. The initial assumption is that countries with a nuclear industry need to deal with the issue of nuclear waste management. This is a meaning that needs to be created and produced in the context of an already existing nuclear industry. In fact, most countries in this position still do not have a well-defined and viable solution for highly radioactive, long-lived, waste. To different degrees, these countries have created structures to deal with the issue but in the process have triggered opposition from environmental groups and local communities, who strongly oppose the meanings they appreciate in nuclear waste management, to the point that the long-term viability of the nuclear industry may hinge on changing these appreciations. This is a policy issue that needs producing in each country a requisite organisation for a meaning that is still undefined, let alone produced. However, there are already multiple nuclear waste management activities; nuclear waste produced in nuclear plants is being disposed of and to a larger extent stored in an interim form. Therefore, whether it is desirable or not, there are structures already in place producing a tacit meaning for nuclear waste management. These structures are modelled for the French case in Figures 3 and 4 (Espejo and Hoverstadt, 2002). Figure 3 shows the hypothesised unfolding of complexity for nuclear waste management and Figure 4 shows the institutional resources focused on creating and producing the nuclear waste management system in France. Figure 3 makes apparent that local resources within institutions/enterprises such as EDF, CEA, COGEMA and ANDRA (L&I short-lived RW) produce the

The footprint of complexity

693

K 33,3/4

694

Figure 3. FNWMS unfolding of complexity

Figure 4. Embodiment of FNWMS

French Nuclear Waste Management System (FNWMS), and that the first three are embedded within institutions that together with the FNWMS produce the French Nuclear System; this dual embedding of Interim Waste Management create conflicting relations of belonging that may blur the identity of the FNWMS. Figure 4 shows the ascription of roles to institutional resources and allows us to see a number of resource configurations. For instance, the same institution may constitute more than one primary activity (e.g. ANDRA appears contributing as part of the intelligence function to the production of the global nuclear waste management system and also, as part of the implementation function, to Low and Intermediate Short Lived Radioactive Waste Management System (L&I short-lived RW). Also IPSN appears as constituting two different systemic functions, namely the cohesion and intelligence functions of the FNWMS. These are instances of distributed resources coming together, adequately or not, in the embodiment of one social system. In situations like the one illustrated in this example, where it is hypothesised that the system is produced by distributed institutions, producing their own meanings, operational coupling among them is likely to be weak and therefore coherence to produce a common desirable meaning will not be easy to achieve. In this case, unless adequate attention is paid to orthogonal interactions among all the stakeholders, the chances are that some of the resources will have unchecked influence on policy processes, while others will feel out of touch and powerless to influence them. This lack of adequate systemicity (that is, the absence of stretching mechanisms leading to stable, far from equilibrium, relations) becomes particularly apparent in democratic decision processes, where the views of the affected silent majorities may not be taken sufficiently into account at the peril of the policy makers, who eventually may find that they have lost legitimacy to carry out their preferred policies. Diagnosing relations among resources, as they participate in producing organisational processes, is necessary to refine our appreciation of the grounding of purposes into identity through structure. This diagnosis may clarify that a hypothesised primary activity does not have resources to create its own meanings and therefore that it is not an instance of structural differentiation, but merely an amplification of a meaning that has already been created at a wider level (e.g. interim waste management in a particular nuclear establishment may just be an amplification of energy production in that establishment, a characteristic of a hierarchical relation rather than of functional differentiation). How resources are distributed, where is the discretion in their use, throughout the requisite organisation depends to a large degree on self-organisation but also hopefully on design. This design should consider desirable long-term unfoldings of complexity and also distribution of resources, based among other factors on the nature of the policy issue and an appreciation of the knowledge available. For instance, nuclear waste management is the case

The footprint of complexity

695

K 33,3/4

696

of national policy that has a critical impact at the local level (the locality where a repository is finally built). This verticality suggests that there should be several levels of meaningful debate for this policy issue, which may eventually influence the way its complexity is unfolded. However in practice, the distribution of resources and the development of relations are less the outcome of design and more of accommodation, which often is far from ideal. Understanding the actual distribution of resources makes apparent the messy and often chaotic organisation of collectives and also gives directions for improvement. This example of a requisite organisation to produce a social meaning for nuclear waste management shows the heuristic value of the VSM. A wide range of resources such as waste stores in nuclear plants, experts in national research and regulatory bodies, local and national NGOs, politicians in local and national government, are mapped either as constituting one or more systemic functions at one or more levels of recursion or as relevant environmental agents at different structural levels. The quality of their relations can then be tested empirically and recommendations can be made to increase the chances of having a fluid and desirable emergent social system. Of course in the end, this wide range of resources may not develop the requisite connectivity to achieve structural and functional differentiation. The implications of this failure are hazy meanings, fragmented structures, lack of transparency in decision processes and increasing difficulties in giving legitimacy to policies. This is a common situation, which is recognised in controversies of all kinds and the failure of democratic processes. In spite of all these difficulties the idea of a requisite organisation offers powerful guidelines to increase the quality of people’s participation in decision and implementation processes. 7. Conclusion: critical reflection The systemicity of human activities is reflected in the social meanings we produce in our communications. A focus on the eigenvalues/behaviours of utterances, information and understandings allows us to recognise disembodied social systems, which is Luhmann’s insightful way to avoid the methodological difficulties of dealing with often distributed resources that cannot be accounted for in the production of a meaning. However, when the purpose is intervening in social processes to create flexible, fluid, desirable social systems, solving the methodological problem of their embodiment becomes paramount. This embodiment implies a social system that develops an organisation capable of creating and producing its own meanings (i.e. of functional differentiation). In general, this is the outcome of processes of incursion, or structural coupling, among differentiated social systems, which require orthogonal interactions, in order to create these new meanings, and structural

recursion in order to produce them. Methodologically this requires a model to map onto a few logically interconnected systemic functions the diverse and disperse social resources focused on a particular policy issue. These functions are necessary and sufficient for the creation, regulation and production of a social meaning, and that is precisely what the VSM offers; policy, intelligence, cohesion, coordination and implementation are necessary and sufficient for this purpose. The VSM’s capacity to map hugely complex social situations is given by its recursive nature, which allows that mapping to be made in multiple embedded systems, from the global to the local, all sharing the same need for aligned viability. In practice, because socially we do not know how to harness all these resources in the direction of an aligned viability for a desirable purpose we are not in control of the meanings we produce; they just emerge as an outcome of poorly understood and often misguided interventions on self-organising processes. There is a huge need for social learning. The purpose of this paper has been to reflect upon this learning. How do we know which is the body (requisite organisation) for the meaning it has yet to be created, let alone produced?

I have argued in this paper that for design purposes, reflecting upon this embodiment has to be focused on ingenious social responses to particular problem situations, rather than on already existing social systems. The challenge has been to reflect on the institutional responses for the creation and production of specific policy issues. The concern has been in particular desirable systems at present. Nuclear waste management was the issue of concern in this paper. In historic terms, a major concern of those in positions of power has been to control an ever-increasing social complexity. Different forms of control have influenced the evolution of social systems. The driving forces have been the mutual interplay between the resources at hand for this control and the ideologies necessary to justify their use. The outcome has been the exclusion in part or totally of people from social processes, leaving in the hands of minorities the creation and production of meanings, simply using the majorities as amplifiers of their interests. Bauman (2000) gives powerful examples of these forms of exclusion. As discussed earlier, it was the recognition of all this and the need for individual emancipation in these oppressive conditions that triggered much of the critical thinking produced in the latter part of the 20th century. The driving force at present is social inclusion, and much has improved in social terms in the past decades; however, still we are far from the production of truly inclusive social systems. Good intentions are not enough to produce the required transformation. The argument of this paper has been that emancipation requires a deeper understanding about the embodiment of social systems and in particular about the organisational processes leading to the production of desirable social systems.

The footprint of complexity

697

K 33,3/4

698

The VSM gives us a heuristic for transforming rigid and unwieldy social systems into flexible and malleable social systems. Indeed, it is a heuristic for structural coupling and as a consequence for emancipation. Yet the complexity of social communications and interactions is such that any claim of this kind has to be tempered by the pragmatic restrictions of the systems in place and the capabilities of the available resources. Producing required connectivity, implementing necessary levels of meaningful debate in order to connect the local and the global, respecting the autonomy of all necessary structural levels to perform a task, aligning purposes, achieving the mutual stretching of stakeholders through effective orthogonal interactions are all desirable activities but are extremely difficult to achieve. In fact they are constantly testing our ingenuity and often stretching the available resources beyond their capabilities. This may imply that the systemicity we achieve today is as exclusive as it has ever been; only today it is perhaps more sophisticated and subtle and also people’s awareness of their rights and responsibilities is higher. However, the way towards true emancipation is still incipient. Effective democracy and participation demand much creativity and are expensive in resources. These issues of inclusion and exclusion make apparent the relevance of having tools to define and recognise the boundaries of social systems. In the example of nuclear waste management, it is important to establish who are responsible for the meanings that society experience for that policy. Are these meanings the outcomes of aligned resources focused on nuclear waste management or the outcomes of fragmented meanings produced by isolated resources. Do the global meanings – purposes and values – as espoused by politicians and experts, have anything to do with people’s local experience about nuclear waste management in particular communities? The ideas of functional and structural differentiation, developed here, offer useful epistemological and methodological insights about the discussion of boundaries; we can hypothesise social systems and discuss their boundaries from the local to the global. In evolutionary terms, it is necessary to understand better how organisational closure, ultrastability and recursion can evolve humanely in social systems, respecting the autonomy of individuals and opening the space for cohesive collectives, with rights and responsibilities. In fact, I would argue that the big issue for social systems is not their systemicity that is entailed in the emergence of collective meanings, and in any case, is often forced upon people by those in positions of authority, but that of ethics. Notes 1. These are behaviours/points of stability emerging from recurrent processes in organisationally closed systems. In Varela’s terms the self-referential and recursive nature of a network of processes, characteristic of the autonomy of natural systems, is captured by

2.

3. 4.

5.

the invariant behaviour proper to the way the component processes are interconnected ( Varela, 1979, p. 170). This may be different for enterprises emerging as self-organising entities in free markets; this is not only a particular case of the more general discussion of policy issues explored in this paper but also a point that needs further debate. They can also be constructed if purposes for them are ascribed from outside the collective. Producing a meaning requires, producing some form of product or service, however the essence of an organisation is not producing particular products or services, but producing a meaning through particular products or services. Hence, the same organisation may produce many different products or services over time; as long as it maintains its identity it is the same organisation. For instance, defining safety standards for road construction go beyond the competence of individual plants and therefore, may be issued as “corporate intervention” without negotiation.

References Apel, K-O. (2001), The Response of Discourse Ethics, Peeters, Leuven, Belgium. Argyris, C. and Scho¨n, D.A. (1978), Organizational Learning: A Theory of Action Perspective, Addison-Wesley, Reading, MA. Ashby, W.R. (1964), An Introduction to Cybernetics, Methuen, London. Bauman, Z. (2000), Liquid Modernity, Polity Press, Cambridge. Beer, S. (1979), The Heart of Enterprise, Wiley, Chichester. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester. Conant, R.C. (1979), “Communication without a channel”, International Journal of General Systems, Vol. 5, pp. 93-8. Espejo, R. (1993), “Management of complexity in problem solving”, in Espejo, R. and Schwaninger, M. (Eds), Organisational Fitness: Corporate Effectiveness Through Management Cybernetics, Campus Verlag, Frankfurt. Espejo, R. (2001), “The wisdom of the people: a framework for transparency”, Proceedings of Values in Decisions on Risk, Second VALDOR Conference, Stockholm, Sweden. Espejo, R. (2002), “Self-construction and restricted conversations”, Systems Research and Behavioural Science, Vol. 19 No. 6, pp. 517-29. Espejo, R. (2003), “The Riscom model: boundaries and contextualisation”, Proceedings of Values in Decisions on Risk, Third VALDOR Conference, Stockholm, Sweden. Espejo, R. and Hoverstadt, P. (2002), “Structure for transparency in nuclear waste”, Management Report as the System of Waste Management in France, Electricite de France (EVF), HE75/02/035/A, July 2002. Espejo, R., Bowling, D. and Hoverstadt, P. (1999), “The viable system model and the Viplan Software”, Kybernetes, Vol. 28 Nos. 6/7, pp. 661-78. Giddens, A. (1984), The Constitution of Society, Polity Press, Cambridge. Habermas, J. (1984), Theory of Communicative Action: Reason and the Rationalization of Society, Tomas MacCarthy (Trans.) Vol. 1, Polity Press, Cambridge. Holmer-Dixon, T. (2001), The Ingenuity Gap, Vintage, London. Leydesdorff, L. (2000), “Luhmann, Habermas and the theory of communications”, Systems Research and Behavioural Science, Vol. 17 No. 3, pp. 273-88.

The footprint of complexity

699

K 33,3/4

700

Luhmann, N. (1995), Social Systems, Stanford University Press, CA. Maturana, H. and Varela, F. (1992), The Tree of Knowledge, Shambhala, Boston and London. Mitchell Waldrop, M. (1993), Complexity: The Emerging Science at the Edge of Order and Chaos, Virking, London. Searle, J. (1979), Expression and Meaning; Studies in the Theory of Speech Acts, Cambridge University Press, Cambridge. Varela, F. (1979), Principles of Biological Autonomy, North Holland, New York and Oxford. Von Foerster, H. (1984), Observing Systems, (2nd ed.), Intersystems Publications, Seaside, CA. Wene, C-O. and Espejo, R. (1999), “A meaning for transparency in decision processes”, Proceeding of Conference on Values in Decisions on Risk, First VALDOR Conference, Stockholm. Further reading Beer, S. (1985), Diagnosing the System for Organizations, Wiley, Chichester.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

A Latino American Requiem for Stafford Beer Roberto Zarama, Jose´ Bermeo, Nelson Lammoglia and Jairo Villamil

A Latino American Requiem 701

Department of Industrial Engineering, Universidad de los Andes, Bogota´, Colombia Keywords Cybernetics, Systems theory Abstract This tribute to Stafford Beer is written from the particular situation of a Latin American country. Within this context, we believe that it could be convenient to dissolve the systems movement to assume an ethical-political position that transforms the relations of dependence in the world-system from the point of view of economic development. A contribution to this task is the formation of amplifiers of regulatory capacity that has been developed in Colombia. In this paper, we present some preliminary advances of autoethospoiesis and some of its future perspectives of research and action. One of our main objectives is to apply this functor with recurrent, recursive, and incursive operators to the world-system model.

Introduction The authors of this paper have found themselves celebrating Stafford Beer’s life because, in a manner of speaking, they were set adrift (Maturana and Varela, 1984) and, as a result of this, encountered some extraordinary people in their lives. With little merit to participate in this celebration, other than this, the first thing that they remembered was a conversation with Stafford. We recalled Stafford saying to Roberto that Mozart had been able to compose his music because he had stumbled upon a recursive structuration for his work. Based on this memory, we feel that Stafford Beer’s life could be recognized and inspired in W.A. Mozart’s Requiem, K. 626. To fulfil our original purpose, we faced two major obstacles. The first one was the fact that we had to accept our lack of expertise in music and the second one was to acknowledge the fact that Stafford had painted and described his own Requiem (Beer, 1993). We decided not to overcome these two obstacles. We resolved to invite others to take on this task collectively. Our proposal was to assume a periphery class position (Marx and Engels, 1975) in order to reconstruct this Requiem. The reader who is interested in this invitation should be one who will be willing to embody this position. This invitation is tetra-dimensional in totality. In the first dimension, we propose, as in Mozart’s Requiem, a metaphor to celebrate Stafford’s life. In the second dimension, we seek to find the possibility of burying a kind of systems movement. In the third dimension, we intend to show a type of knowledge that can contribute to extend Stafford’s work in time. In the fourth dimension, we shall give some advance notice on research that is being done and we think that

Kybernetes Vol. 33 No. 3/4, 2004 pp. 701-716 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523652

K 33,3/4

702

can connect with Stafford’s own line of work. As our closing statement, we want to propose some thoughts on future perspectives. A Requiem as a metaphor to celebrate Stafford Beer’s life Our invitation intends to acknowledge the fact that Stafford went beyond the call of duty in his work and then proceed to celebrate his life by means of a new analysis of Mozart’s Requiem. A Requiem Aeternam for Stafford implies that we are left with many commitments to fulfil ourselves on his behalf. This proposal consists in taking up those commitments that emerge in the light of Stafford’s work and to pursue them in order to make it possible for those people living in the periphery of the world to produce a well-being as a result of their capability (Sen, 2000) to take care of their dignity. Thus, we penetrate the metaphor from a specific point so that we can extend our invitations to all those who want to follow us in this endeavour. We believe that we have to take up our commitment much in the same way as Franz Su¨ssmayr took up his with respect to his teacher’s Requiem. The task at hand is not to search for Stafford’s legacy. Stafford’s legacy will be constituted by that which his followers choose to produce. It will be a low profile legacy and not something to make any of us famous because this is something reserved only for our masters. This legacy is for those who do not see themselves as the sole heirs of their masters or of their thinking. The legacy implies the acceptance of criticism and the absence of ad baculum fallacies with the proposed results. Mozart’s Requiem is a good starting point to define the context of our own commitments. We refer to a context such as a Kyrie for the situation in which most peripheral societies find themselves. We sing a Kyrie for an education that produces domination instead of liberation and we also sing a Kyrie for those of us who play a part, no matter how insignificant, in this system that produces and reproduces this centre-periphery relationship. We believe that this Kyrie metaphor can be the starting point to build a context where it will be possible for the culpabliss error (Beer, 1996) to emerge. Stafford invented this word from the Latin word for “guilt” and from the acronym for “Blind in Systems Sensitivity” (Beer, 1996). We believe that our task is to give root to the culpabliss error thus dissolving the condition of periphery or structure of dependency. These tools require certain processes to make distinctions, to give them root and to embody them as we have previously modelled (Reyes and Zarama, 1998). We should also remember that the Dies irae is the beginning of the Sequenz. In this beginning, we listen to a warning about the day in which the world will be turned to ashes. We believe that such a statement is an opportunity from the point of an embodied culpabliss error. We also believe that if we apply this warning to the structure of today’s world, from the point of view of Stafford’s work, we are in a position to build hope. We can generate these possibilities if

we dissolve the present world-system structure that is suffering from a “systemic crisis characterized by regulatory collapse” (Beer, 1992). We understand that this crisis can put us in a state of confusion, but, in the same way in which Mozart’s Confutatis closes the Sequenz, we too can, metaphorically speaking, go from hell to heaven. If we fulfil our proposed task then we shall be in a position to sing a hymn of praise such as the Sanctus or the Benedictus in Mozart’s work. This hymn could be our contribution to Stafford’s everlasting peace as in Mozart’s Agnus dei. If we manage to fulfil the political commitments that we still have ahead of us, we will have certainly contributed in a recursive fashion to give operational closure to Stafford’s assured place in the glory of humankind as is the case with Mozart’s Communio at the end of his last musical work. To open up the possibility of creating a Requiem for the systems movement In the context that we have just described, we would like to invite you to turn Stafford’s sad death into an opportunity. Under these sad circumstances, the opportunity could be an embodied metaphor (Lakoff and Johnson, 1999) to bury a certain type of systems movement and to generate the necessary conditions to preserve the most valuable aspects of the systems discourse in the conversations of future generations. This could become the greatest hymn of praise for Stafford and it could be extended to include Heinz von Foerster’s life who passed away a few months after Stafford. Likewise, in this way, we could also pay homage to Norbert Wiener, Arturo Rosenblueth, Warren McCulloch, Ross Ashby and many more. The motivation to bury the systems movement To understand the process that has undergone the systems movement we have used the concept of “charismatic structure” (Weber, 1964). In our case, the structure was born within the group that, after World War II, brought about a complete renovation of the way of thinking about doing science. Stafford was not included in that first “cybernetic group” (Heims, 1991). He would become part of a second generation of cybernetic thinkers, referred to as “The Three English Musketeers”, after Namur (von Foerster, 1997). The slow disappearance of the charismatic founding group has encouraged many people to come together under the initial ideals. This gave rise to an inner structure that channelled authority, institutionalised the movement, and organized federations, meetings, and magazines that would encourage the exchange of personal opinions among its members. The external expression included symbols and discourses that gave this group a common identity when they were together. In this process of rationalization, some modes of subjection (Foucault, 1990), that were seeking recognition for an autonomous discipline, came into being.

A Latino American Requiem 703

K 33,3/4

704

They turned the system into an object of knowledge and went as far as to define accepted methodologies for this purpose. This pretension caused the disappearance of “a form of cross-disciplinary thought which made it possible for members of many disciplines to communicate with each other in a language which all could understand” (Mead, 1968). We propose the metaphor of a funeral of the systems movement in order to effect a transformation of the world and, more specifically, of the world of systems. We propose this because we feel that the concept of cybernetics has been grossly misunderstood to mean “just a magic word of our present time” (Beer, 1992). We believe that cybernetics is not a tool to act as if we could possess the abracadabra powers of a magician. We think that the repetition of certain words, such as the use of the word Cartesian, is not the right way to achieve our purposes. We must take the trouble to make the necessary distinctions that those words evoke in our minds. We should make all the necessary efforts to avoid being like the character in the The Canterbury Tales who liked to use phrases in Latin that he had learned by heart, but could not understand because “like a parrot he was really dense, he’d learned the words but could not grasp the sense” (Chaucer, 2003). Talking about this, Stafford once said that we should leave a space for ignorant people. We could say that “like all poetical natures he (Stafford) loved ignorant people. He knew that in the soul of one who is ignorant there is always room for a great idea” (Wilde, 1986). We believe that this situation emerges because “the potency of the new idea may be dissipated by . . . entropy” (Beer, 1992). One such case is that of the debate over methodologies. It is possible that, in academic circles of the centre, this debate will have some validity, but we must say that, from the periphery, this is seen as a useless pursuit. Moreover, the effort to generate a multi methodology could yield as a result a state of multiphrenia (Gergen, 1991). A funeral to make the culpabliss error possible A hypothesis of this paper is that the funeral of the systems movement will help the energies to build synergies (Fuller, 1975) so that the culpabliss may emerge as a daily distinction in the present world-system. We believe that there could be two ways of understanding the blindness to which Stafford is referring. One way would be to reduce the sphere of influence of the methodologies to the point where systems become mere objects of knowledge. In this perspective, the research question refers to the means whereby we can get to know a system better. This point of view is anchored in theories that belong to the domain of epistemology. Here, the political problems are usually circumscribed to disputes about the best methodology. The second way of understanding the blindness is in the sphere of influence of the constitutive relationships within social systems. The research question that underlies this point of view refers to the capacity to transform the

relationships that constitute and maintain a given social system. This second approach interprets the interests that emerge in a peripheral country. This point of view focuses on the praxis. It is a point of view in which we could say that “starting from the interpreted sensitivity, not as knowing, but as proximity – (it is sought) to find contact and sensitivity in language” (Levinas, 1989). This point of view springs from the ethical-political domain. Our invitation consists in burying a movement that debates within the first point of view so that the resources can be invested in dissolving the system produced by the second point of view. In this position, the epistemological problem keeps its validity. Its validity springs from its usefulness in dissolving ethical-political problems within systems. In this sense, we believe that “when knowledge (. . .) works in such ways that makes possible the production of the ethos, then it is useful” (Foucault, 2002). In particular, we refer to systems whose purpose is opposed to the goal that has been ascribed to them. In this respect, we should not forget the aphorism that says that “the purpose of the system is what it does” (Beer, 1985). To illustrate this, we have two examples: the welfare state that castrates the possibilities of a large majority and the diffusion of the discourse of development that increases poverty in the periphery. From this second vantage point, we recognize the advances in the formalization of systems thinking at the level of interpretation of situations. Nonetheless, we have not seen the same advances in the transformation of the world. This is why we believe that the systems movement has produced more effect in the domain of abstract systems theory than on the policies that generate and maintain these systems. It has been unable to produce the rules or channel the resources for their reproduction to effect the necessary transformations in the world-system. However, “it is misleading to speak of ‘rules of transformation’ because all rules are inherently transformational” (Giddens, 1984). In this sense, we believe that we can apply the systems movement to Marx’s statement on philosophers when he said that “philosophers have only interpreted the world in various ways; the point is to change it” (Marx, 1969). Now, our invitation is to observe “globalisation and its discontents” (Stiglitz, 2002) from the point of view of culpabliss error. The idea is to dissolve a model that produces and reproduces relationships of dependence that structure present world-system (Wallerstein, 1974). Moreover, we believe that we have a responsibility to show how this world-system has generated “an order that effectively suspends history and thereby fixes the existing state of affairs for eternity (. . . an order that) not only regulates human interactions but also seeks directly to rule over human nature (. . . that is) the paradigmatic form of bio power” (Hardt and Negri, 2000).

A Latino American Requiem 705

K 33,3/4

706

An offer of knowledge that has been constructed to continue Stafford Beer’s work We refer to some of the activities within the realm of informal education that we did for the Contralorı´a General de la Repu´blica (Colombia’s National Audit Office) and further discussed by German Bula in this issue of Kybernetes. We want to highlight a key difference between this work – summarized in Systemic Practice and Action Research (Espejo and Reyes, 2001) – and the project that was developed by Stafford in Chile (Beer, 1981; Espejo, 1980, 1990). The Chilean project was characterized by investment in resources to construct attenuators of social complexity, whereas the Colombian project was characterized by investment in the formation of amplifiers of regulatory capacity. In the Colombian project, we in-serviced 900 auditors in 24 months with 3,497 h of workshops. Elsewhere, we have commented on the learning model that supports the process of advanced learning (Zarama, 2001). Here, we want to emphasize that, with the results of the aforementioned project, a process of formal education was designed. This product was a postgraduate degree in Systems of Management Control offered at the Universidad de los Andes. The same research group has also formed many amplifiers through research conducted by students who are studying for a Master’s degree or a PhD. Based on the knowledge acquired in the processes just described, we propose to create an alliance of research groups that will be willing to adopt a position of peripheral class and whose purpose – in and from the action – will be to generate amplifiers who will work in the transformation of those relationships of dependency that structure the world-system at present. Report on the state of some research projects that we are working on at present The future of the alliance that we are proposing to form will depend on the conscience of peripheral class and not on the triumph or failure of each isolated battle (Luka´cs, 1985). The purpose of this alliance is to influence the transformation of the world-system from the point of view of a recursive model (Beer, 1979). We shall now report on the advances of some of our investigations that we are pursuing in this direction. An approximation to autoethospoiesis processes When speaking of culpabliss, we made reference to two ways of understanding the blindness that Beer mentioned. Let us remember that the epistemological problem maintains its validity “when knowledge (. . .) works in such way that it makes possible the production of the ethos, then it is useful” (Foucault, 2002, the emphasis is ours). This takes us to consider that it is fundamental to establish how the way of knowing and the form of how we know the systems is going to affect those systems. With this we want to count with processes that

affect the way in which people act. This means to have-in-hand forms to produce the ethos of the subject. In this sense this investigation is directed at constructing the capabilities (Sen, 2000) to bring-forth a world. Such construction does not follow a “deductive or inductive, but a heuristic” (Luhman, 1996) method. This means that we follow an abductive logic (Pierce, 1931-1958). By abduction we refer to a Bayesian rule of conditional probability. This is a type of forced-choice diagnosis. In this way, the autoethospoiesis is a process that helps to observe a social system. In this process, we articulate – try to go beyond – two traditions. First is the process of interpretation originating in the hermeneutic tradition, and in second place are those processes in which actions emerge with meaning. These second processes require the coordination of the coordination of actions. This coordination of actions generates organizations. These processes originated in a point of view from a structured observation of organizations. For the generation of the heuristic, we articulate two concepts. On one side, the concept of world developed by Heidegger. On the other side, the concepts of organizational identity constructed by Espejo. In this sense, we understand the world as being configured by the necessary equipment to make possible certain actions, by the purposes that give direction to those actions, and the emergent identity for the subjects who produce these actions (Heidegger, 1962). The organizational identity is understood as the result of a triunion (Spencer-Brown, 1994) of relationships of cohesion, citizenship, and performance (Espejo, 1999). Next, we will make explicit the results of the articulation of the two mentioned concepts: worldhood and organizational identity. The material configuration of a world is realized in an organization. The shape of an organization manifests itself in a body that constructs itself. The bodily self construction is a process of continuous reiteration of the organization’s identity. This recurrence is a creative process developed by the regulatory processes of the organization (Conant and Ashby, 1979). The model for the regulator emerges from a process of structuration in which resources, rules, and languages are articulated (Giddens, 1984). In this body, it is possible to have “the action of some over the action of others” (Foucault, 1994). This is why we say that the body is a sort of space (or clearing) that limits those it holds and unfolds those objects that are susceptible of manifesting themselves and those actions that are susceptible of being undertaken (Heidegger, 1962). The body of an organization is a space that governs in a “non-urging” manner (Heidegger, 2000). In this space, the collective boundary is established. These processes produce at once the cohesion (Espejo, 1999) of the social body and distinguishes its subjects as autonomous beings. We understand that the autonomous person is the minimum specifiable unit within a social body. In this manner, the body is the space through which a population of free subjects

A Latino American Requiem 707

K 33,3/4

708

makes decisions. For this reason, we say that an organization is recognized as a polis. We propose that the basic operation to bring-forth the body of an organization be the poiesis (Heidegger, 1977). In our understanding, poiesis is a process that produces, makes possible, causes, or makes apparent a system. Here, we consider this operation as a process that produces, reproduces, and deduces the world that constitutes the body of a social organization. The body is produced from the articulation of the four causes that were conceived to play within the limits of the poiesis (Aristotle, 1998). We do not understand cause as a mechanism to produce an effect. The fourfold causality “are the ways, all belonging at once to each other, of being responsible for something else” (Heidegger, 1977). This relation constitutes the accountability of an organization. These processes produce at once the performance (Espejo, 1999) of the social body and distinguishes its subjects as responsible persons. The poiesis operates from the knowledge of knowledge. The model in this process has a circular shape to draw distinctions in language, embody distinctions in the body of subjects and root distinctions in the body of an organization. We believe that Kant’s questions to produce the body of an organization are helpful, namely: What can we know? What can we do? What can we expect? We can reformulate these questions like this: Which are the limits of knowledge of knowledge? Which are the possibilities (or limits) to transform the constitutive relationships within a social system? Which are the limits of the process that produces a given social system? For us, the answers come from recognizing the relationships constructed by the modes of subjection that configure the organization. The boundary of knowledge is basically produced in the articulation of two types of modes of subjection. In the first place, we are speaking of a relationship developed by the modes of subjection to the biology that we share as species. Secondly, we refer to a relationship constructed by the modes of subjection to the multiple polis in which the citizens, in the conversations that constitute them, ascribe a sense to their activities. In this way we, the subjects, emerge from the physical conditions that come from the biochemical system that configures us and from the citizenships (Espejo, 1999) of the various polis. We have referred to two modes of subjection. For the first type, we shall borrow the concept of autopoiesis as a characterization of a given biology (Maturana and Varela, 1995). For the second type, we shall borrow the ethospoiesis as a characterization of a given polis (Foucault, 2002). With this, we can observe a union of biological and political systems that we are subjected to through relationships. Interpreting the complexity of these two characterizations we have begun to construct an invention that we have entitled autoethospoiesis.

Towards the modelling of autoethospoietic organizations It is necessary to note that the body of a social organization is neither a biological system nor a conglomerate of individuals. Moreover, subject is not the same as individual. In this sense, we shall explain how to represent an autoethospoietic organization. We understand this process as a process to present again. To present again is a process to model an organization. For this process of representation or process for modelling, we shall show the three operators that, at this point, configure autoethospoiesis. For the time being it suffices to say that the citizenship in different polis implies being able to recognize the difference and not accepting exclusion. In this way, the function autoethospoiesis frees the subjectivities of the forced domination of totalitarian institutions. These institutions tend to have rigid ethical codes. From this function, the subject emerges as a dimension of the personal and organizational autonomy. This development strays away from individualism and collectivism. This means that the body of a person emerges from a lived body (Merleau-Ponty, 1975) in the daily experience of a variety of distinctions that it is capable of absorbing. We have taken this measurement of variety from the Law of Requisite Variety (Ashby, 1964). Even more, given that we propose that the social systems are autoethospoietic, we require a way to model them. We understand a model as a function. This function is the one that maps a system to the model. For an autoethospoietic organization this has to be a self construction process of continuous reiteration of the organization’s identity. This recurrence is a creative process developed by the regulatory processes of the organization (Conant and Ashby, 1979). We have not limited ourselves to stating the functor, but have also designed a game as an active mechanism of learning autoethospoiesis. The game is inspired in the invention of “team syntegrity” (Beer, 1994). We have not remained within the realm of the original protocol. The game that we have proposed reinterprets syntegration from the point of view of the theorem of the four colours and the patterns that generate spaces for action through connections with the synergetic geometry (Fuller, 1975). In this context, we shall now describe the present state of articulation of the three operators that constitute the autoethospoiesis. The first one is the recurrent operator that operates observing the body of the subjects as relatively autonomous systems. We are speaking here of the body of biological subjects or product of the autopoiesis function. In particular, we are referring to this function when it operates to transform molecules – such as peptides and receptors (Pert, 1997) – in that which is reported in language as dispositions. We understand dispositions as eigen behaviours (von Foerster, 1984), or as a type of “fixed pattern of action” (Llina´s, 2003). We are referring to patterns that emerge in the relationships of citizenship (Espejo, 1999) that we establish as subjects in the various systems in which – in

A Latino American Requiem 709

K 33,3/4

710

interaction – we recognize ourselves as subjects. The dispositions “represent the pre-motor platform that encourages or stops the majority of our actions” (Llina´s, 2003). This operator configures that which we call pathos. Pathos is the capacity of subjects – that manifests itself in the body of persons – of feeling, of being affected, and of affecting. This capacity develops conversations that configure the coordination of the coordination of actions that, in turn, is the daily manifestation of the body of a social organization. This is the level in which “things, people, and selves . . . show as meaningful” (Spinosa et al., 1997). In this sense, conversations emerge from dispositions that effect cohesion (Espejo, 1999) within an organization. Following the origin of the term, we call sym-bolic those dispositions that effect cohesion in a dialectical relationship with the dia`-bolic dispositions that, in turn, are the ones that produce fragmentation (Boff, 2000). Metaphorically speaking, we can say that the sym-bolic dispositions resemble a centripetal force whereas the dia`-bolic dispositions are like a centrifugal force. Thus, as is in the case of a dynamic balance, we require both forces. The sym-bolic dispositions are meaningless without the dia`-bolic and vice versa. From the point of view of this metaphor, in the methodology these dispositions/conversations are represented in the vertices of a graph. The balance is reached in transitional states that show dissipative structures (Prigogine and Stengers, 1984) that keep the cohesion of the system. The reconstruction of the sym-bolic dispositions is a structuration “to preserve the identity” (Beer, 1994). For this, the organization makes a problem of its self (Heidegger, 1962) in which the organization reveals itself in the moment in which a relationship with something that has been thrown is established. This tool looks at going beyond structuralism and hermeneutics (Dreyfus and Rabinow, 1982) to design the mechanisms of effect/affect(ion) that constitute the ethos. This capacity for cohesion allows us to recognize the social interrelationships that configure identities. The second operator is the recursive operator that operates by observing the body of the organization. This operator produces the boundary of the social systems. This boundary is constituted by the ethos that the organization reproduces. The body of the organization is the result of the ethospoiesis function. This function transforms truth (for that system) into ethos (Foucault, 1994). This operator configures that which we call logos. Originally, the word logos meant “to bring together”. According to this, logos has to do with “the unification that is produced in speaking: telling and narrating” (Brugger, 2000). We understand logos as the capacity of a recursive organizational structuration to build a rationality that will produce cohesion as within a systemic totality and that will be equipped with feedback mechanisms. This structuration

provides a balance between autonomy and cohesion. The structuration brings together subjects; it does not conglomerate individuals. Recursion is configured by means of two operations. The first operation determines the relationships of cohesion of those organizations that are relatively autonomous. The second operation establishes the relationship of citizenship of organizations that are contained within larger organizations. These operations constitute the equipment that is reproduced by the knowledge that emerges from the organization. The equipment is the tool box that configures the power as a differential capability for action within an organization. From the point of view of our metaphor in the methodology model, the equipment is represented in the folds and in its possible recursive structures. The third operator is the incursive operator that operates by observing how the modes of subjection are produced. We think that the modes of subjection appear in “an inter-subjective field, not in spite of my body or of my historical situation, but, rather, because this body and this situation exist through them” (Merleau-Ponty, 1975). Within this inter-subjective field, the power connections, that connect bodies of persons with bodies of organizations, will emerge. The incursive operator observes the transformation of power into rules. In this situation, we observe the bond of the ethos in and with the polis. Ethos and polis should always be understood in the relationship that exists through the nomos (Bernstein, 1993). The basic expressions of nomos are the constitutive and regulative rules (Searle, 1995) of truth games (Foucault, 1994). In this sense, the incursive operator answers the question about the subject of the subject. To understand this operator, we should remember that ethos is a circular articulation with two meanings: ethos, written with “eta” and ethos written with “epsilon”. When ethos is written with “eta” it refers to our dwelling (Heidegger, 2000). When we talk about dwell, we refer as much to an animal shelter as to a home for human beings. Our home is where we grow roots, where we build our certainties, and where we feel comfortable because we are surrounded by our worldhood. This worldhood relates pathos with logos. Thus, this first meaning of ethos is reconstructed by observing the limits of the recursive operator. When ethos is written with epsilon, it refers to customs and habits. Habits are related to behaviour. In this realm, we have the coordination of actions that configure the daily life of an organization. This second meaning is reconstructed by observing the limits of the recurrent operator. With the incursive operator, we challenge the limits of the ethos that is understood as a sponge of variety that gives operational closure to the organization (Beer, 1985, 1994). This debate demonstrates that the limits of the ethos can be observed within the limits of the polis. In this debate, we can also point out the level of cohesion of the polis. Ultimately, this means that with this debate, we can find out if relationships of citizenship are present within the organizations.

A Latino American Requiem 711

K 33,3/4

712

The question indicates whether the organizations emerge as relatively autonomous systems that are contained in relatively autonomous organizations. This operator also shows the level of development of a structure of care (Heidegger, 1962) – that configures listening – in which a relative autonomy is developed through the relationships of organizational citizenship. This means that this operator shows if there is a relationship of citizenship that obliges. To be obliged means that “as people we are necessarily ligated” (Cortina, 1998). Here we are dealing with the modes of subjection. In this sense we question “the way in which the individual establishes his relation to the rule and recognizes himself as obliged to put it into practice” (Foucault, 1990). This operator develops the structure as a social poietic learning process. This process appears, in the social reproduction of knowledge, as a circular process to undertake, understand, and learn the generative praxis of effective action. With this operator we intend to observe that “only in the understanding of praxis can a suitable new approach to human settlements – the communities, villages, and cities – be propounded” (Beer, 1983). With this operator the subject of the subject configures the praxis of the game that is played. In this game the purpose, that allows the subject to ascribe meaning and build a space of trust within the organization, emerges. The praxis is developed as an activity with meaning for an organization. This means that the praxis configures the recurrent action in which the meaning, that gives cohesion to an organization, emerges. In this way, the organization emerges as a relatively self-contained and autonomous system. Here we connect with the first operator because the praxis configures the pre-motor platform in which the dispositions emerge. Given the fact that the incursive operator challenges the rules of the game – understood as the limits for the action – in the metaphor of the methodology, the meta game is represented in the folds of the graph.

Development as a purpose of the world-system observed from the point of view of autoethospoiesis We have a case study in which the function of autoethospoiesis operates on the concept of development of the world-system. We shall start with the incursive operator. For this we have reconstructed the rules that Braudel has de-duced (Braudel, 1992). In these rules, we recognize the purpose of the game. The purpose of the game is the construction of meaning for our action. We express this by saying: “I doubt that there has been even one government anywhere in the last thirty years that has not declared as its objective that it pursues economic growth” (Wallerstein, 1991). Then we take these nations as economic and political subjects and we apply the recurrent and recursive operators to them.

Currently, we just have a preliminary result. We are reconstructing the result as a fractal in which we identify topological relationships such as the ones we described with respect to closeness to the centre and periphery. At present, we are trying to refine the model to see if we can identify a topology for a model. In this process, we can say that, for the time being, we have a metaphor – the yo-yo methodology (Beer, 1966) – in which the centre of the periphery is closer to the centre of the centre than the periphery of the centre is to the centre of the centre. For the purposes of this invitation, this would mean that if these relationships are maintained, then the universities of the periphery are in the centre of the periphery and, therefore, are closer to the universities of the centre of the centre, than the universities of the periphery of the centre. The advantage of the universities of the centre of the periphery consists in the possibility that they can get closer to the periphery of the periphery. This closeness can help to adopt the position of class that is required to dissolve a system that has been set up on dispositions of dependence. In that sense, the work of Albert-La´szlo´ Baraba´si indicates the possibility of studying these networks as free-scale network models that allow us to understand the centres of the present world-system as hubs of the network of global interaction (Baraba´si, 2003).

Tools to model autoethospoiesis At present, we are seeking to produce a greater articulation in these works. We believe that the answer can be found within the field of “A New Kind of Science” (Wolfram, 2002). We have used cellular automata, models of agents, fractals, some aspects of chaos theory such as attractors, and paradoxical games. We are looking into the concept of structural coupling as a possibility of producing coordination from the point of view of chance. Something that has been very helpful is a case in which we have modelled the resources of the country as a fable in which dogs represent the expenses in management, wolves show the losses as a result of corruption, and sheep represent investment in the production of public goods. The model shows the importance of constructing trust. In this sense, we are working on a hypothesis that consists in observing games in which we have the historical possibility of losing (probability lower than 1/2). We hope to observe, in coupling two of these games, a probability of winning that is greater than 1/2. With this we are working on cases such as the merger of organizations or the integration of countries. We do not think that the universal consensus with respect to integration should lead us to totalitarian positions in favour or otherwise. On the contrary, our job is to advance in the search for conditions of possibility under which the integration will be favourable to the social purpose that we are seeking. In that sense, we oppose the ideas of globalization or localization as valid ideas per se.

A Latino American Requiem 713

K 33,3/4

714

Extension of future perspectives We have invited you to celebrate Stafford’s life with a Requiem. For this Requiem, we have proposed a change in the use of resources to dedicate them to the transformation of the world. A starting point could be to produce and promote tasks to dissolve the structure of peripherization of the world-system. In this sense, our task is to use systemic methodologies as tools and not to try to find the most appropriate one to work on systems as objects of knowledge. These tools will be useful if they help to dissolve problems such as hunger, violence, corruption, or exclusion, all of which could be understood as emergent properties of the world-system. A starting point to effect/affect our task is to recognize and declare ourselves as citizens of a continent in which Stafford inspired and actually undertook many projects. This is a region that emerged from the cross-breed of modernity to produce a system that we now want to (and need to) dissolve. The problem that we hope to dissolve is ethico-political. Our task is to accept that the teachings of our masters, such as Stafford Beer, have helped us to understand their models. These methodologies are the stairs that have brought us up to where we are now. Thus, now we “must so to speak throw away the ladder since we have climbed upon it” (Wittgenstein, 1922). The burial of the systems movement on the part of research groups that assume a peripheral position could contribute to the construction of “a new age of curiosity” (Foucault, 1994), a curiosity that “evokes the care one takes for what exists and what might exist (. . .) a certain determination to throw off familiar ways of thought and to look at the same things in a different way” (Foucault, 1994). References Aristotle (1998), E´tica Nicomaquea, Editorial Porrua, Me´xico. Ashby, R. (1964), An Introduction to Cybernetics, Methuen, London. Baraba´si, A.L. (2003), Linked, PLUME Penguin Group, Cambridge. Beer, S. (1966), Decision and Control, Wiley, Chichester. Beer, S. (1979), The Heart of Enterprise, Wiley, Chichester. Beer, S. (1981), Brain of the Firm, 2nd ed., Wiley, Chichester. Beer, S. (1983), “The will of the people”, Journal of the Operational Research Society, Vol. 34 No. 8, pp. 797-810. Beer, S. (1985), Diagnosing the System for Organization, Wiley, Chichester. Beer, S. (1992), “Recursion zero: metamanagement”, Systems Practice, Vol. 14 No. 1, pp. 51-6. Beer, S. (1993), “Requiem”, Kybernetes, Vol. 22 No. 6, pp. 105-8. Beer, S. (1994), Beyond Dispute: The Invention of Team Syntegrity, Chichester, New York. Beer, S. (1996), “The culpabliss error: a calculus of ethics for a systemic world”, Memorandum Marking his Seventieth Birthday. Bernstein, R. (1993), The New Constellation, The MIT Press, Cambridge. Boff, L. (2000), El despertar del a´guila, Editorial Trotta, Madrid.

Braudel, F. (1992), Civilization and Capitalism, 15th – 18th Century, University of California Press, Berkley, Los Angeles. Brugger, W. (2000), Diccionario de Filosofı´a, Herder Barcelona. Chaucer, G. (2003), The Canterbury Tales, taken from Handout #1 – The Prologue to The Canterbury Tales on 6 September 2003, available at http://www.horizonshelpr.org/english/ canterbury/handouts/Handout1_part2.pdf Conant, R. and Ashby, R. (1979), “Every good regulator of a system must be a model of that system”, Int. J. Systems Sci., Vol. 1 No. 2, pp. 89-97. Cortina, A. (1998), Hasta un pueblo de demonios: e´tica pu´blica y sociedad, Taurus, Madrid. Dreyfus, H. and Rabinow, P (1982), Michael Foucault: Beyond Structuralism and Hermeneutics, The University of Chicago, Chicago. Espejo, R. (1980), “Cybernetics praxis in government: the management of industry in Chile 1970-1973”, Journal of Cybernetics, Vol. 11, pp. 325-38. Espejo, R. (1990), “Complexity and change: reflections upon the cybernetic intervention in Chile, 1970-1973”, Systems Practice, Vol. 3 No. 3, pp. 303-13. Espejo, R. (1999), “Aspects of identity, cohesion, citizenship and performance in recursive organisations”, Kybernetes, Vol. 28 No. 6/7, pp. 640-58. Espejo, R. and Reyes, A. (2001), “The state of the state”, Systems Practice and Action Research, Vol. 14 No. 2, pp. 135-7. von Foerster, H. (1984), Observing Systems, Intersystems Publications, Salinas. von Foerster, H. (1997), “Celebration of Stafford Beer’s 70th birthday”, Espejo and Schwaninger, To be and not to be, that is the System, Carl Ouer, Heilderberg. Foucault, M. (1990), The Use of Pleasure, The History of Sexuality, Vintage Books, New York, NY, Vol. 2. Foucault, M. (1994), Essential Works of Foucault 1954-1984, The New York Press, New York, NY. Foucault, M. (2002), La hermene´utica del sujeto, Fondo de Cultura Econo´mica, Buenos Aires. Fuller, R.B. (1975), Synergetics: Explorations in the Geometry of Thinking, Macmillan Publishing Co. Inc., New York, NY. Gergen, K. (1991), The Saturated Self, HarperCollins Publishers Inc., New York, NY. Giddens, A. (1984), The Constitution of Society, Polity Press, Cambridge. Hardt, M. and Negri, A. (2000), Empire, Harvard University Press, Cambridge, MA. Heidegger, M. (1962), Being and Time, Harper and Row, New York, NY. Heidegger, M. (1977), The Question Concerning Technology, and Other Essays, Harper Torchbooks, New York, NY. Heidegger, M. (2000), Carta sobre el humanismo, Alianza Editorial, Madrid. Heims, S. (1991), The Cybernetics Group, MIT Press, Cambridge. Lakoff, G. and Johnson, M. (1999), Philosophy in the Flesh: The Embodied Mind and its Challenge to WesternThought, Basic Books, New York, NY. Levinas, E. (1989), De otro modo que ser o ma´s alla´ de la esencia, Sı´gueme, Salamanca. Llina´s, R. (2003), El cerebro y el mito del yo: el papel de la neuronas en el pensamiento y el comportamiento humanos, Editorial Norma, Bogota´. Luhmann, N. (1996), Confianza, Anthropos, Me´xico. Luka´cs, G. (1985), Historia y Consciencia de Clase, Ediciones Orbis, Barcelona.

A Latino American Requiem 715

K 33,3/4

716

Marx, K. (1969), “Thesis on Feuerbach”, Marx/Engels Selected Works, Progress Publishers, Moscow, Vol. 1, pp. 13-15. Marx, K. and Engels, F. (1975), La ideologı´a alemana, Grijalbo, Barcelona. Maturana, H. and Varela, F. (1984), El a´rbol del conocimiento: las bases biolo´gicas del entendimiento humano, Editorial Universitaria, Santiago de Chile. Maturana, H. and Varela, F. (1995), De ma´quinas y seres vivos, Segunda edicio´n, Editorial Universitaria, Santiago de Chile. Mead, M. (1968), “Cybernetics of cybernetics”, in von Foerster, H. et al. (Eds), Cybernetics of Cybernetics, Spartan Books, New York, NY, pp. 1-19. Mearleau-Ponty, M. (1975), Fenomenologı´a de la Percepcio´n, ediciones penı´nsula, Barcelona. Pierce, C.S. (1931-1958), Collected Papers of Charles Sanders Peirce, 8 Vols, in Charles, H., Paul, W. and Arthur, B. (Eds), Harvard University Press, Cambridge, MA. Pert, C. (1997), Molecules of Emotions, SCribner, New York, NY. Prigogine, I. and Stengers, I. (1984), Order Out of Chaos, Batman, New York, NY. Reyes, A. and Zarama, R. (1998), “The process of embodying distinctions – a re-construction of the process of learning”, Cybernetics and Human Knowing, A Journal of Second Order Cybernetics, Autopoiesis and Cyber-semiotics, Vol. 5 No. 3, pp. 19-33. Searle, J. (1995), The Constitution of Social Reality, The Free Press, New York, NY. Sen, A. (2000), Libertad y Desarrollo, Planeta, Bogota´. Spencer-Brown, G. (1994), Laws of Form, limited edition by BoobMasters (Ashland, Ohio) for Cognizer Co., Portland, Ore. Spinosa, C., Flores, F. and Dreyfus, H.L. (1997), Disclosing New Worlds, MIT Press, Boston. Stiglitz, J. (2002), Globalisation and its Discontents, The Penguin Press, Harmondsworth. Wallerstein, I. (1974), The Modern World System: Capitalist Agriculture and the Origins of the European World Economy in the Sixteenth Century, Academic Press, New York, NY. Wallerstein, I. (1991), “Development: Lodestar or illusion?”, Unthinking Social Science, The Limits of Nineteenth – Century Paradigms, Polity Press, Cambridge. Weber, M. (1964), Economı´a y Sociedad, FCE, Me´xico. Wilde, O. (1986), De Profundis and Other Writings, Penguin Books, London. Wittgenstein, L. (1922), Tractatus Logico-philosophicus, Routlege and Kegan Paul Ltd, London. Wolfram, S. (2002), A New Kind of Science, Wolfram Media, Inc., Champaign, II. Zarama, R. (2001), “Creating a new auditor at the Contralorı´a General de la Repu´blica de Colombia (CGR)”, Systemic Practice and Action Research, Vol. 14 No. 2, pp. 181-202. Further reading Papert, S. (1979), “Epistemologı´a de la Ciberne´tica”, in Piaget, J. (Director) Tratado de Lo´gica y Conocimiento Cientı´fico Vol. V, pp. 47-63.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Fighting for science

Fighting for science

Gerard de Zeeuw Faculty of Science, University of Amsterdam, Pieter Jacobszstraat, Amsterdam, The Netherlands Keywords Cybernetics, Systems theory

717

Abstract What deliverables did Stafford Beer envision when he developed his “science of effective organisation”? This paper answers this question as: the organisations that use the distinctions of Beer’s viable system model. Such organisations are part of daily life, but develop to become knowledge by continuously striving to identify experiences that falsify their existence. They will be irreducible in the sense that any acceptable model of the organisation will be the organisation itself. The notion of knowledge involved is made explicit in the paper as a tribute to Stafford Beer’s pioneering work. It allowed Stafford Beer to introduce and develop insights that began to be developed by others only much later. Civilisation . . . worked well in a more leisurely age, but now its relaxation times no longer match the rate of perturbation. . . . Let us use knowledge. . . . to produce freedom as an output Designing Freedom, Stafford Beer 1994, pp. 92-100.

1. Introduction That the past is with us is no surprise – except when suddenly we realise its crooked influence. As Robinson points out in his review of Molotch (2003), the “dimensions of the solid rocket boosters that carry the space shuttle into orbit” are directly constrained by the width of existing train tunnels and rail track, which the “British got from the ruts of Roman chariots pulled by horses”. Stafford Beer will be remembered for many things, but particularly for his heroic fights against the “mess” the world is in – due to the way the past presents us with highly worrying problems (Beer, 1994b, p. 6, 1994c). These include that organisations fail due to previous, sometimes Roman designs; these also include the negative effects of their collapse (Beer, 1994c, p. 5). He also will be remembered for his approach to these problems as a practical philosopher (like Machiavelli, one of Stafford’s sources of inspiration), using three powerful weapons. The first was thinking clearly and writing beautifully, the second aiming for action – rather than only for understanding (Beer, 1994d, p. 13). The third was applying Ashby’s (1964) Law of Requisite Variety. This law distinguishes two sources of experiential variation – one the ability of humans to act and thereby differentiate and chain-link events, the other processes in their environment, often involving other humans. The notion of effective action refers to the possibility of matching both variations and of stabilising their match (usually within a finite time horizon). Matching obviously will be most difficult when the variety considered is large – the product of dense interactions of “men, materials, machines and money” (Beer, 1994d, p. 31), often signalling organisational difficulties.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 717-725 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523661

K 33,3/4

718

Continuous efforts are required to increase and reduce variety on either (source or sink) side, as well as the use of complex tools and having to stay alert every moment. It is a joy to find how easy it is to wield such tools – if one were a Stafford. What is keeping others from doing the same is the weight of history. Throwing off its shackles and anticipating on the future is the aim of the “science of effective organisation” (Beer, 1994c, p. 13), all this standing on the shoulders of Wiener’s (1961) “science of communication and control in the animal and the machine”. Possibly due to the central role played by the Law of Requisite Variety, one important aspect does not seem to have been properly recognised. Any such science will require a significantly expanded formulation of its deliverable, i.e. knowledge. To celebrate Stafford’s work, this paper aims to provide such a formulation as well as how it is justified. The basis of the formulation is the premise that the Law of Requisite Variety is a special case of what science is looking for. In other words, identifying situations that satisfy the law is equivalent to acquiring knowledge. The latter consists, therefore, not only of what can be said about daily life and thereby is set apart from it, but also of what can be recognised as being still in daily life. This definition has begun to be acceptable only recently. Knowledge still tends to be linked mainly to what “underlies” or is “behind” the experiences of daily life and hence is set outside of it. That it may be already “in” daily life began to be realised by contemporaries like Ernst von Glasersfeld, Humberto Maturana, Gordon Pask as well as Polanyi in 1966 and Von Foerster in 1970. In their view, knowledge does not link to a distant reality. It results from continued comparison of the experiences that people report (in some language [1]), and from combining them (Varela, 1979). The combinations that will be preferred are those that require only minimal maintenance and hence interference, and thus serve as resources to effective action. In Section 2, core aspects of knowledge are identified. Section 3 discusses resolving the fundamental problem of testing, leading to the formulation of an extended form of knowledge. This form makes the deliverables explicit that Stafford Beer envisioned for his science of effective organisation (Section 4). Section 5 concludes the paper. 2. Pre-constructions Stafford Beer frequently tells of organisations and institutions that continue to function in spite of ineffective decisions and errors (Beer, 1994a). He identifies many behavioural and organisational aspects that once may have been acceptable and necessary, but are no longer. These include privileges as well as unequal distributions of wealth, position and education (Beer, 1994c). We may see organisations as past their success, as post-constructions that will end up in the dustbin of history. We may also see them as pre-constructions

– intended to stay but at present lacking maintenance. Their repair requires that we understand their functions. Unfortunately, the repairs may be applied to “the wrong side of the equation” (Beer, 1994c, p. 31), thus continuing the failures. The notion of a pre-construction refers to contracts, infrastructures and even progeny – as well as to other elements in daily life that increase people’s capacity to offset the unexpected. It also includes knowledge, which is pre-constructed in that its use is to remain free and accessible over time – and invariant, as is any form of prediction (Beer, 1994b, p. 87; Rosen, 1993). To build pre-constructions, effort has to be spent now instead of later. The notion of knowledge being free to use by any one any time clearly is ideal. Future usage usually requires some effort still, for example, in the form of parameter estimations to adapt to local conditions. Accessing knowledge in the future also may need training and schooling. One cannot expect knowledge to exhaust all future efforts, therefore, but only to minimise these efforts. We do not move all future efforts to the present happenings for other reasons as well. Exhausting such efforts would mean that no future daily life would be left. Conversely, if we are to continue having such life, only part of the efforts we spend in the present can be spent on trying to minimise efforts in the future. This part is called (scientific) research. Reducing future efforts implies experiencing the effects of the reduction in the present as well as in the future, and hence experiencing their combination in the present. The experiences to be combined are called primary, so the experience of their combination is termed secondary. Links between primary and secondary experiences constitute pre-constructions. The establishment of links between experiences is quite common to daily life. We use links to categorise people, for example, by the links they favour, as in the case of poets and politicians. We also use links, including poems, to induce experiences like love. We drive bicycles by creating (secondary) steering experiences to match (primary) experiences of disturbances. Knowledge refers to special types of links. Scientific research began in the 17th century as a search for links that are: (1) between reportable experiences (sometimes also of sensory experiences, but there is a limitation: one can share only what is reported [2]), (2) minimal and hence unique by not being reducible to link other sets of primary experiences. The original role model is Galileo using his magnifying glasses to look at the sky – thereby turning them into a telescope. His primary experiences consisted of reports or rather drawings of lights that varied in position next to the planet Jupiter. Considerable argument was still needed to link these to the (freely accessible) secondary experience of “seeing planets around Jupiter”. This link proved a sensation. It made visible what others like Kepler and Brahe had conjectured already. It also demonstrated the possibility of links that

Fighting for science

719

K 33,3/4

720

had a new authority in the sense of being stable against access and use by anyone, and hence, of being at least as convincing as the links claimed by priests and princes. This authority can be characterised as democratic. It may seem strange to emphasise democracy in relation to knowledge. Usually the notion of democracy is used in the context of sharing power, in contrast to the power owned by social institutions. Galileo’s links were democratic in that they could be shared by anyone who had access to a telescope (and used it properly). This is what makes them democratic [3]. 3. Testing collectives However, making links is easy. Problems arise when the aim is to acquire knowledge. One has to select which links satisfy the criteria of being unique and democratic, and minimise or exhaust later efforts. In more familiar terms, what needs to be done is to show that the variety of secondary experiences absorbs the variety of the primary experiences (as attenuators; Beer, 1993, p. 26). This terminology may obscure the difficulty of testing to some extent. To compute “variety” one needs (finite) sets, whereas primary experiences may refer to classes with an infinite number of elements. To assess whether a secondary experience will “absorb” this infinity requires a pre-construction itself, or a pre-construction of a pre-construction. This indicates how serious is the difficulty. What is to prevent an infinite regress of pre-constructions? How would one distinguish between future efforts that are needed to correct errors, or estimate parameters, and efforts that are meant to test a pre-constructed link, e.g. in case new primary experiences become available due to the use of the link? To answer such questions, we may model the testing of links by mapping them onto a universal turing machine (UTM). Such a machine is disturbed only by its input. It reads symbols from a tape (possibly containing blanks [4]), it may change what it reads by applying operations selected from a finite (behavioural) table and it prints the results on the tape (Hodges, 1997). Scientific links between secondary and primary experiences are defined as an effectively calculable function if its (mapped) values can be (re-)produced or printed reading from a blank tape and using operations from the table. This identifies such links, but does not provide a rule to decide whether a function is calculable, or to stop calculating given a finite number of experiences. We still need some practical way of testing, therefore, although we currently may recognise knowledge as what is effectively calculable. For this we seek recourse to Popper (1959). He requires a (any) person to do the testing – who must strive to recognise what belongs to the primary experiences, but cannot be produced by the secondary ones. Any such experience would falsify the link. This (any) person is called an actor, following the earlier argument. As a term, this does not add much to the argument yet, except to reformulate the

problem. The actor has to match the variety of many machines, each printing tapes using a different secondary experience (mapped onto the behavioural table). This is achieved by stopping any machine any time when it first prints a falsification. Although there is no rule to stop all machines, apparently there is a rule to decide when to stop some of them. The advantage of introducing a person who strives to falsify, is that this number is finite at any one time. This helps, but still does not solve the problem of testing, as the number of machines to be tested may still be infinite. We do not get out of considering all possibilities. Stopping the regress of pre-constructions is especially important when we attempt to know what cannot be mapped onto operations on physical things, as represented by a UTM. In this case we deal with links that prove not calculable. When we ignore this aspect the difficulties, which the 17th century approach to testing had to face, will be aggravated, possibly to the point where the acquisition of knowledge. Fortunately, even in this case it appears possible to pre-construct (and test) links that may serve as knowledge. One may think of links between primary and secondary experiences that differ from any calculable link (e.g. the result of applying Cantor’s diagonalisation procedure; see Deutsch (1998)). Such links can be pre-constructed, even though they are not effectively calculable. This type of construction may be exemplified by introducing a second actor, in addition to the researcher as the first actor, who aims to know the experiences produced by the latter, and who generates what the researcher tries to know [5]. The resulting interactions are not effectively calculable, but may be pre-constructed, even though only as an expanded form of knowledge. In this case, the number of primary experiences may be more than infinite (or beyond the bounds of the infinite). There is a way, however, to reduce this number. We may combine the two actors in a collective (Von Foerster, 1970). If a third actor tries to gain knowledge about the collective, he or she is also to be inducted. No one aiming to acquire knowledge is left outside. Therefore, all acquisition of knowledge is to be done by the resulting collective. This suggests that the collective coming into being is the same as the process of acquiring knowledge: the collective pre-constructs its own deliverable. The co-ordination that helps the inside actors to maintain the collective (its secondary experience) functions like a communication language. The collective will also be charged with testing itself. It should strive to test the links on which it is based (and thereby self-organise itself). This means that its members must try and generate primary experiences that are not linked to the collective’s co-ordination (or language), hence may falsify its existence as well as suggest adaptations to that experience. The resulting collective may seem familiar as well as strange. It is familiar in that the notion of knowledge management refers to knowledge addressed inside an organisation (Hammer and Champy, 1993; Von Krogh et al., 2000). It is

Fighting for science

721

K 33,3/4

722

strange in that it refers to the development of collectives that are knowledge, and hence can be addressed as such both inside and outside [6]. This result is gratifying in that knowledge as a special type of collective does not appear to be in conflict with earlier notions, but rather to reconstruct the latter at higher levels of aggregation. More precisely, it expands these notions, as they are still part of the same quest, the search for links that can be mapped onto calculable functions (on a UTM). The fact that this result signals a relatively general type of expansion is demonstrated by the expansion due to the Ashby model, which also tests for knowledge but not by striving to falsify. It computes the variety of finite sets, for example the variety of the finite state spaces of organisational models (as instances of a UTM), or of their disturbances – which excludes testing for infinity. It is well known that the search for results that are machine calculable has proved very successful. It identified links that are minimal and unique, which has greatly improved the effectiveness of action. In the above, it was argued that it is possible to achieve the same in the case of the non-calculable – by identifying or initiating systems that self-test by becoming knowledge. 4. Striving systems The latter exemplify what Stafford Beer since early times describes in terms of his viable system model (VSM). It is not a contingent model of an organisation. It points to a common deliverable: maintainable collectives that exhaust or minimise future efforts (Beer, 1993, p. 101, 1994d, p. xi). It supports all organisations, or enterprises, to keep testing to what extent this aim is realised (Beer, 1994c, d, p. xii). Tests take various forms, as argued. The model typically requires that actors are addressed and helped to continuously diagnose which links are insufficiently exhaustive and what is needed for redress (Beer, 1993, p. 1). Testing does refer to organisations that strive to experience what may falsify their co-ordination and minimise future testing. To use the model, the first step is to identify the activities that a given organisation is to realise as well as where and when its primary experiences are inadequately matched, so additional knowledge and hence testing (or striving) is needed. The first level of analysis chosen in the VSM is referred to as system one. The resulting collectives may be successful, but usually over time start to change and oscillate too much (Beer, 1994c, p. 6, 1994d, p. 468). Help is needed to facilitate acquisition of knowledge. It comes from system two, a system to research how resources may be recognised, e.g. by modifying the constraints on the activities of system one (the “resource bargain”; Beer, 1993, p. 38). The collective into which systems one and two are inducted may need extra policing (to deal with events that have not been brought into the present

yet) – to be identified by an additional system (system three). The collective of the three systems is expected to maintain the stability of its “inside-and-now” by striving to adapt its internal co-ordination (Beer, 1993, p. 107). Over time there will be a need to deal effectively with changes in the wider environment of the collective, about which further knowledge has to be acquired (for example market research, R&D). Two more systems should be set up and inducted into the collective to explore the collective’s long-term and ethical strategies (Beer, 1993, pp. 115, 125). A collective that includes these five systems is expected to facilitate and speed up the process of finding primary experiences that may falsify its links. Such a collective should show closure (Beer, 1993, p. 125, 130). It will be able to consistently extend its (separate) existence as knowledge (Beer, 1993, p. 1), and thereby increase the effectiveness of the activities it supports. In the 1970s, the notion that an organisation may achieve being knowledge given proper forms of testing was a long way from the (academic) public’s mind. It was developed and implemented as the basis of Stafford Beer’s VSM however, far in advance of later understandings [7]. Given the latter, the model may be expected to lead to further successes.

5. Conclusion The aim of this paper was to celebrate the contributions of Stafford Beer, and especially to explore the deliverables of his “science of effective organisation” (Beer, 1993). It proved possible to identify them as organised knowledge, as systems that are knowledge, on the basis of an argument that expands on previous developments concerning the nature of knowledge. This kind of knowledge does not represent some separate reality. It only requires that it proves possible to move efforts to the present that may support action in the future, with minimal dissipation. Interestingly, such efforts appear to have a lower limit, suggesting that the experiences involved are part of irreducible classes characterised by calculable functions. When this characteristic is not present, such classes may still be constructed. In this case, the resulting knowledge takes the form of a system, part of daily life, which autonomously strives to test itself on being knowledge. It is irreducible to any model but itself (De Zeeuw, 1995, 2001). Although recognition of such systems is relatively recent, it was fully anticipated in Stafford Beer’s viable system. There is much for which Stafford will be remembered, in particular his ability to penetrate difficult problems and deal with them, whatever their complexity. What has been highlighted in this paper is something else – his sophisticated insights into the nature of knowledge and their implementation in daily life as striving systems.

Fighting for science

723

K 33,3/4

724

Notes 1. Experiences on which to test knowledge (or prefer some combinations) thus include “I see that the sun sets”, and “I smell honeysuckle”, but not statements like “the sun sets” or “I have a sensation that I interpret as the setting of the sun” (Tarski, 1944). 2. Humans are limited in reporting smells, sounds, feelings, etc.; much less so in reporting observations. 3. To state something with authority one did not need to be at the top of the social ladder anymore (Fienkelkraut, 2001). 4. Reporting an observation is assumed sufficiently “immediate” so it may be mapped on, and linked to, reading one symbol. 5. Such a second actor is the source of possible innovation. 6. The part of knowledge that is outside a collective is what it presents to outside users (including inside users who behave like outside users). The part of knowledge that is inside refers to the distribution of labour over the members of a collective, implying constraints or qualifications on activities, and hence defining competences. The latter type of knowledge appears related to tacit knowledge (Polanyi, 1994); in this case presented by an organisation rather than by a (human) body (Beer, 1993, p. 108), but not to local knowledge (Turnbull, 2000). Note that tacit knowledge is supposed to be exhaustive. 7. The acquisition of knowledge as a system implies knowledge about values: how to initiate them, how to maintain them. For example, to maintain a knowledge system, members must be honest, or non-corrupt.

References Ashby, W.R. (1964), An Introduction to Cybernetics, Methuen and Co. Ltd, London. Beer, S. (1993), Diagnosing the System for Organizations, (Original 1985), Wiley, Chichester. Beer, S. (1994a), Decision and Control. The Meaning of Operational Research and Management Cybernetics, (Original 1966), Wiley, Chichester. Beer, S. (1994b), Brain of the Firm, (Original 1972), Wiley, Chichester. Beer, S. (1994c), Designing Freedom, (Original 1974), Wiley, Chichester. Beer, S. (1994d), The Heart of Enterprise, (Original 1979), Wiley, Chichester. Deutsch, D. (1998), The Fabric of Reality, Penguin Books, London. De Zeeuw, G. (1995), “Values, science and the quest for demarcation”, Systems Research, Vol. 12 No. 1, pp. 15-25. De Zeeuw, G. (2001), “Constructivism: a ‘next’ area of scientific development?”, Foundations of Science, Vol. 6, pp. 77-98. Fienkelkraut, A. (2001), In the Name of Humanity, Random House, London. Hammer, M. and Champy, J. (1993), Re-engineering the Corporation, Nicholas Brealey, London. Hodges, A. (1997), Turing, Phoenix, London. Molotch, H. (2003), “Where stuff comes from. How toasters, toilets, cars, computers and many other things come to be as they are”, Routledge, New York. Reviewed in Robinson, S. (2003), “Dreams that stuff is made of”, Science, Vol. 301, p. 1675. Polanyi, M. (1994), The Tacit Dimension, (Original 1966), Doubleday, Carden City, New York, NY. Popper, K.R. (1959), The Logic of Scientific Discovery, Unwin Hyman, London.

Rosen, R. (1993), “Some random thoughts about chaos and some chaotic thoughts about randomness”, J. Biol. Syst., Vol. 1 No. 1, pp. 19-27. Tarski, A. (1944), “The semantic conception of truth and the foundations of semantics”, Philosophy and Phenomenological Research, Vol. 4. Turnbull, D. (2000), Masons, Tricksters and Cartographers, Harwood Academic Publishers, Amsterdam. Varela, F.J. (1979), Principles of Biological Autonomy, Elsevier, New York, NY. Von Foerster, H., (1970), “Molecular ethology. An immodest proposal for semantic clarification”, in Ungas, G. (Ed.), Molecular Mechanisms in Memory and Learning, Menum Press, New York, NY, pp. 213-48. Von Krogh, G., Ichijo, K. and Nonaka, I. (2000), Enabling Knowledge Creation. How to Unlock the Mystery of Tacit Knowledge and Release the Power of Innovation, Oxford University Press, Oxford. Wiener, N. (1961), Science of Communication and Control in the Animal and the Machine, (Original 1948), MIT Press, Cambridge. Further reading Watzlawick, P. (1989), “Mu¨nchhausens haren en Wittgensteins ladder”, Uit de greep van de werkelijkheid, Van Loghum Slaterus, Deventer.

Fighting for science

725

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

726

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Implications for Beer’s ontological system/metasystem dichotomy Maurice Yolles Liverpool John Moores University, Liverpool, UK Keywords Cybernetics, Management Abstract Stafford Beer developed managerial cybernetics, but there were many facets of his work. Most of his work concerned epistemology, and little concerned ontology. Not all of the aspects or implications of his work has been fully recognised, and an attempt shall be made to explore one of these. In particular, this paper explores his paradigm by considering some of the epistemologically and ontological angles. Some of the implications for Beer’s work will also be shown to have led to the creation of a virtual paradigm capable of exploring his achievements “externally”, after Go¨del.

Introduction

Kybernetes Vol. 33 No. 3/4, 2004 pp. 726-764 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523670

The 1960s saw the maturing of a few great thinkers who, in the 1970s and 1980s, led the way into forms of constructivism that paved the way for a new emergence into the social sciences. These included people like Foucault (1974), Habermas (1970), Piaget (1977) and Vygotsky (1978). It is rarely recognised that it also included Beer who developed his own constructivist approach[1] conditioned by the inconsistency theory of Go¨del (Beer, 1959, 1979, p. 311). Both Piaget and Vygotsky began their work with an interest in child development, adopting related, but distinct approaches that result in related, but differentiable axioms for their respective paradigms. Vygotsky was interested in the social processes that enabled learning to develop. Piaget used a cybernetic approach for his interest in how the interactions between children contributed to their learning processes, and this developed into an examination of the relationship between subjects and objects (rather like Foucault (1974) whose interest lay in subjectification and objectification as political processes). Habermas (1970) was interested in the cultural, social and behavioural aspects of people, and in the subjective transfer of meaning through the process of communication. Beer was interested in how social communities were able to survive, realising that regulation was central to this. Like Habermas, he also recognised the need for communications. However, he was guided by the formal logic of systems by Whitehead and Russell (1910) and Go¨del’s (1931) incompleteness theorem that illustrated the limitations of language. His interest in this limitation led him to the development of a new cybernetic paradigm with clear practical application for the management of coherent social communities, seen

as systems with controlled operations. The control emanated from a metasystem that communicated internally through a metalanguage. While the logical systems theory that Beer admired explored systems through the use of metasystems, his interests were very much centred on applied science. The paradigm that he developed created an ontological dichotomy defined in terms of the system and metasystem. The term dichotomy was not the one normally used by Beer, but in fact it is rather harmless because it means a “division into two”[2] that can be argued here to represent two ontological species of a given generic entity. The generic entity may be seen as a self-organising body and its ontological species are the system and metasystem that interact as an intimate ontological couple. That is, they each have validity claims about reality that operate in a way that mutually relate: one validity claim to reality is manifested in the other relative to its validity claim to reality. Hence, a thought in the metasystem may be manifested differently in different systems. Survival, for Beer, related to viability, which he saw as occurring through emergence. This is illustrated by a comment by Denis Adams, a close academic colleague of Beer, when he recently said in a private communication that Beer “was very interested in viability which I see as an emergent behavioural property of a complex system (that we can never establish without doubt) and how we think it may work. But as a result of observations and thinking about how different systems (activities seen as if they were systems) seemed to have a varying behavioural emergent property, he was able to ‘see’ characteristics that the organisation of a viable system should have. These were systemic characteristics in that they had to be abstracted and described from an interacting dynamic whole system.” Beer’s pragmatic interest lay in mapping his analytical ontological conceptualisation of the metasystem/system onto practical situations. While he developed the basis of a new ontology through the creation of the dichotomy, his pragmatic interest centred on epistemology, which become manifested through his viable system model (VSM). To illustrate this, again quoting from Denis Adams. The thrust of Stafford’s work in VSM is epistemological in nature; “what do we know and how do we know that” is how he defined epistemology to me “in a nutshell”. So the VSM helps you to think about a situation in terms of viability, and the VSM is describing the communication and information flows round a system that is doing the same (what do they know and how do they know that) for the sub-system behaviours.

The relationship between the dichotomous parts is recursive. By this we mean that the ontological system/metasystem couple can be embedded as a whole within either the system or the metasystem to give it a new ontological context and an epistemological consequence. Beer’s approach also appears to be constructivist in a way that is consistent with the ideas of not only Piaget, but Vygotski too. Beer was true to the ideas of the incompleteness theorem, and this paper discusses his developments not in

Beer’s ontological system 727

K 33,3/4

728

Figure 1. Nadler and Tushman’s perception of systemic organisational behaviour

the language of his paradigm, but consistent with his view, in another cybernetic language that provides a way of exploring his managerial cybernetics externally. The system paradigm, and beyond Elementary systems thinking emerged from the work of the Gestalt psychologists who emphasised that the study of the mind should be seen as a whole rather than as a collection of psychological parts. The approach led to the notion of holistic thinking (Ellis, 1938). The idea of using a system to understand the phenomena is normally attributed to work in the 1930s by Ludwig von Bertalanffy, a German Biologist. He gave the name general systems theory to a discipline devoted to formulating principles that apply to all systems (Bertalanffy, 1951). Others like Boulding (1956) and Churchman et al. (1957) developed these, and applied them to organisational theory. The traditional concept of a system still used in some areas of organisational theory (for instance, in the Organisational Development (OD) methodology (Yolles, 1999) used in Human Resource Management) is an input-output device. Thus, for instance, Fogel (1967) looked at the human as an information-input/decision-making/decision-output processor, and Nadler and Tushman (1977, 1979) used the notion for organisational behaviour (Figure 1). Further conceptualisations of the idea of the system developed, harnessing the idea that it is not only a processor, but also has the property of synergy and wholeness (Ackoff, 1971). The concept of the system, however, also had a cybernetic dimension. In a foundation paper, Arturo Rosenblueth, Norbert Wiener, and Julian Bigelow (Rosenblueth et al., 1943) were interested in the teleological properties of systems, those that relate to their identity and degree of autonomy and

coherence. In particular, they were interested in biological, physiological, and social systems, and their control and feedback processes. These authors formed the Teleological Society, and after Wiener coined the term cybernetics, they changed its name to the Cybernetic Society. While the conceptual base of cybernetics still centred on the single concept of the system, it was to become transformed by Beer with the introduction of a new frame of reference that involved a second conceptual arm, the metasystem. The notion of the metasystem is credited to Whitehead and Russell (1910) in their logical study of formal systems. Recent theoretical developments of this work has led to Metasystem Transition Theory (MTT) by Turchin and Joslyn (1999) as a means by which higher levels of complexity and control are generated, and by Palmer (2000) as a general theory of metasystem engineering. Another development that had relevance to metasystem theory was the incompleteness theorem of Go¨del (1931), who was concerned with the completeness (if an argument is valid, then it is provable) and soundness (if an argument is provable, then it is valid) of logical systems, and showed that any attempts to prove that a logical system is sound (and therefore having validity and truth) will result in a paradox unless reference is made from outside the system. Beer became interested in the use of the concept of the metasystem as a practical way of explaining the viability of coherent social communities through self-regulation, self-organisation and control. As shown in Figure 2, part of his direction for this came from Go¨del’s (1931) inconsistency theorem (Beer, 1979, p. 311). While he does not seem to have used the term metasystem in his very early work, he did use the term metalanguage throughout. He noted that a system uses a language to communicate about what it does – its operations. However, language is defective because there are always propositions about the language itself that cannot be expressed in the language. Consequently, another language is required that is “over and beyond” the language being used at the time, and this is a matalanguage. It is parallel to the notion supported by Beer of “looking at problems themselves and not at their content” (Van Gigch, 1987, p. xv). In particular, in Beer (1959, p. 169), we are told that “a control system cannot discuss itself and that a higher order [system] is needed in which to describe the behaviour of a system expressed in a given language.” Indeed, he takes this a little further by considering that control is practically linked to operations through its local management. Beer (1966, p. 425) also says that “Since the normal occupation of management is to be expert in using the practical language of the firm’s operations, there is the danger that the management will never speak the metalanguage in which its own structure can be discussed.” While the term metasystem was clearly implied in his earlier work, it does not appear to have been used in his books until later, when Beer (1972) explicitly adopts the term metasystem as the residence of the metalanguage. It is, he says, a “second order system” from which language about the system itself and its language can originate.

Beer’s ontological system 729

K 33,3/4

730

Figure 2. The development of the applied use of the metasystem by Stafford Beer

The nature of the metasystem is that it operates as a control impulse domain, which in human systems occurs through the process of “thinking”, and according to De Bono (1977) has come to be associated with beliefs. Without the metasystem, the system would behave with spontaneous responses that are programmed into its structure through what Yolles and Dubois (2001) refer to as strong anticipation[3]. In the context of the human being, like Beer, De Bono relates the concept of the metasystem to the social community when he suggests that without the metasystem a person would act according to its own personal systems, which might be based on immediate gratification, self-indulgence and impulse. Thus, the metasystem lies outside individual systems and overrides these factors in favour of society and a longer time base.

De Bono provides an illustration of the use of the metasystem: an individual may only collect enough food for his immediate needs, but the metasystem may require him to collect enough to store for the winter as well. For De Bono, to some extent, the success of societies has depended on the strength and nature of the metasystems they have created. By creating a management approach incorporating the system/metasystem, Beer produced a new paradigm for management science and indeed in social science. To see this it is appropriate for us to discuss the idea of paradigm change. It has been only within the last 30 years or so, largely since the work of Kuhn (1970), that we have considered how paradigms change their form. Incremental change involves the development of base[4] concepts and their structured relationships, creating new knowledge. Paradigms also change dramatically as new base concepts arise that alter their frames of reference, i.e. as new conceptual extensions enter their frames of reference (Yolles, 1996, 1999). In doing so, paradigm holders expand their capacity to explain and therefore diagnose the phenomena that they perceive. Such dramatic change has also been referred to as paradigmatic revolution[5] or metamorphosis. It occurs because of a perceived need by paradigm holders to respond to inherent inadequacies, anomalies or paradoxes (e.g. Zeno’s paradox[6]). Such metamorphosis can be part of an evolutionary process within which a new species of paradigm arises that has its basis in an existing paradigm[7]. Metamorphosis is not spontaneous, and paradigms first pass through a “virtual” stage (Midgley, 2000; Yolles, 1996, 1999). VST is an example of this; its original development occurs because of a perceived need to respond to the problem of paradigm incommensurability (Burrell and Morgan, 1979; Yolles, 1996, 1999), and at that time other approaches seemed unable to adequately respond to it[8]. VST can be historically related to the principles of managerial cybernetics. Thus, Beer created a new paradigm whose frame of reference moved from a single base conceptualisation of the operational system, to two: the operational system and conceptual metasystem. We have already said that Beer’s interest in this lay primarily in the epistemological basis of their relationship that was associated with structured communications and meaning. His creation of the metasystem/system dichotomy was clearly an ontological one, but his interest here did not appear to extend to the proper ontological considerations that Casti (1989) would consider is important. The epistemological dimension is reflected in conditions that enabled viability to develop, supported by structured processes of communications that effectively related to semantic communications embedded in what Schutz and Luckmann (1974) would call lifeworld processes. The ontological analysis can be understood by following a lead taken by Habermas (1987) in his three worlds model, and where he explores the validity claims about reality for each world resulting a distinct ontological characteristics that differentiate the realities. Therefore, in Beer’s terms, we are in a position where we should examine the validity claim about reality for both

Beer’s ontological system 731

K 33,3/4

732

system and metasystem. This is illustrated in Table I and derives from Beer (1979, p. 311, 57, 70, 120). The system and metasystem are ontologically coupled, the connection occurring through the boundaries that differentiate them. To elaborate on this, following Yolles and Guo (2003), the system and metasystem each have boundaries that condition their validity claims about reality, and the boundary of one domain is differentiated from that of the other through its ontological horizon[9]. This horizon maintains a content that varies depending on the cognitive perceiver that provides an entry into what may be meaningfully reflected on, spoken about, or acted over. The two domains are ontologically related, and their horizons meld[10] when the domains are seen as an emergent whole. However, this can only occur if the boundaries that create the horizons also harbour ontological migrations[11] that condition that melding. Thus, ontological horizons both distinguish and connect differentiable validity claims about reality. This notion provides entry into the understanding that the boundaries have themselves transformation attributes. Epistemologically the metasystem can be perceived to operate by housing a worldview that the members of an organisation (or more generally social community) hold to, and it is the repository of the knowledge that the community accepts to perform its operations. Social communities are autonomous when they define, create and manage their own futures. A corollary to this is that autonomous systems are recognised to exist only when they have a worldview(s) that has associated with it a language that can economically represent and effectively use its own knowledge. Viability and recursion Beer (1979) wanted to provide an approach that can make autonomous organisations viable, where viability is the ability to maintain a separate existence and to survive as an identifiable entity through appropriate self-regulation and self-organisation. The metasystem has a role in this. An organisation is composed of a set of local systems, and each has its own set of operations. While local management manages each set of operations, the different local systems with their operations are collectively managed through Types of reality System

Metasystem Table I. Realities for Beer’s system and metasystem

Nature of reality A first-order system composed of interactive operational objects that together form a whole, the perception of which is conditioned by a cognitive knowledge-based frame of reference. It is relative to individual subjectivity of groups that have developed normative perspectives A second-order system that operate through concepts, thinking and beliefs, from which knowledge derives. The local individual or group belief-based creation of concepts and their patterns are held in worldviews that establish a frame of reference, and determine what is known and associated meanings

the metasystem. These must operate together systemically with coherence (Beer, 1979, p. 120), and for Ackoff (1971) with synergy and wholeness. An important aspect of organisations that has impact for their viability is that they should be seen in terms of a set of nested system ontological focuses of reality that are also logically distinct levels of examination. This means that the ontological model can be applied and re-applied to different focuses of examination, each focus bringing in local complexity that cannot be seen from a higher focus. Each time this occurs, the nature of the ontology model becomes relative to its host focus. One ontological focus can be used as the reality from which the whole system is referenced (the frame of reference), and be referred to as the referencing focus with its vertically embedded subordinate focuses. This nesting is sometimes referred to as a logical hierarchy, but the term can lead to confusion in people’s minds by mistaking the meaning with that of a socio-political hierarchy typical of bureaucracies. The nature of the nesting of focuses is that at any referencing focus of examination, the system has a set of operations that can be expressed in terms of the deeper set of subordinate focuses. Recursion is often a concept used within methodology (Yolles, 1999). For instance, it is an essential part of Stafford Beer’s VSM when it is used as the basis for methodological inquiry into organisations that may not be viable, and it can be used to correct the faults that it finds. Recursion is also quite intimate to the earlier works mentioned, for instance, Whitehead and Russell (1910). The system/metasystem couple can occur at every focus of examination, and the ability to apply the notion of the couple to distinct focuses is a form of recursion. The notion of recursion can be recalled to mean the application of a whole concept or set of actions that occurs at one logical level (focus) of consideration at a lower logical level (or focus) of consideration. What is particularly interesting here is that the recursive use of the system/metasystem dichotomy provides not only a shift in the degree of relative complexity accessed relative to the reference focus, but also in the relativistic process the metasystem must be capable of being assigned a different worldview(s), with its different knowledge and its own metalanguage. Implicitly, this means that Beer has developed an approach that, because of its knowledge implications, is fundamentally constructivist as well as providing an important entry into an understanding of the nature of the autonomous organisation. To see this, let us first explore the notions that underpin constructivism. Knowledge acquisition paradigms Guba and Lincoln (1994) are interested in the methodology, and identified four knowledge acquisition paradigms that underpin them: positivism, postpositivism, constructivism, and critical theory. However, Anderson (1993) provides a distinct view of the classifications proposed by Guba and Lincoln. Taking this into consideration, this distinction can likely be

Beer’s ontological system 733

K 33,3/4

734

differentiated into two classes, positivism and constructivism, there being a variety of forms of the latter (Table II). More generally, we can define ontological species of knowledge acquisition paradigm. They are normally embedded within concepts of either realism or relativism, as illustrated in Table II. In doing so, we shall consider the considerations of Guba and Lincoln initially, and their four species of paradigm. For them, constructivism is relativistic, and relativists believe that reality is defined by what it is locally perceived to be, and a definable social community usually defines that locality. Other perspectives adopt forms of realism. Epistemological species can also be differentiated, by distinguishing between knowledge acquisition paradigms that adopt objectivism or subjectivism. Guba and Lincoln hold that positivism and postpositivism are objectivist, while constructivism and critical theory are subjectivist. We should note that their view of constructivism is defined in terms of the ontology of social constructivism and the epistemology of cybernetic constructivism as identified in Table II. Their postpositivism and critical theory both support

Postpositivism

Aspect Ontology Description

Table II. Comparative axioms for each species of Knowledge Acquisition Paradigm

Positivism

(cognitive constructivism)

Constructivism Cybernetic Beer (1972, Radical 1979) (Piaget, 1977)

Social (Vygotsky, 1978)

Bounded Relativism Social Personal Personal relativism relativism or social relativism Reality is Reality may Reality is Reality is Reality exists Nature of Reality socially exist, but is local and local and imperfectly. Is reality exists, constructed mediated by cognitively and may be probabilistically cognitively individual demiurgic demiurgic: apprehended apprehendable being shaped senses shaped by by socially related factors socially (e.g. political, related factors cultural, economic) Epistemology Objectivism Subjectivism Description Dualist/ Modified Transactional/ Transactional/ Transactional/ Transactional/ objectivist dualist/objectivist subjectivist subjectivist subjectivist subjectivist Findings are Findings are Findings Findings Through Knowledge Value free, individual, and socially are value are value creation and through falsification, connected to the created mediated mediated findings objectivity, interrelationship through through replicable and findings between subject normative consensus probably true must be values and object replicable resulting from to be true interaction Naı¨ve realism

Realism Critical realism

Critical theory Habermas (1971)

Historical realism

realism, and many realists are also instrumentalists (or logical positivists) believing that reality is defined by the readings that have been acquired through measuring instruments. One area of importance to management processes is communication. A positivist/postpositivist view of interpersonal communication looks at the technical phenomena and expresses the efficiency of transactions. A constructivist view sees a communication episode as part of cultural framework with subjective outcomes. Critical theory takes a rather more bleak perspective on what communication is for, and its conceptualisations may be explored best through the work of Habermas (1987) on Communicative Action, where participants to a process of communication pursue their plans cooperatively on the basis of a shared definition of the situation. Similarly, Luhmann’s (1995) work on social science is concerned with the information bound within a communication, noting that its nature must take on board the notion that it means something very different for the sender and the receiver. Positivism has an ontology that is naı¨vely realistic – that is, there is a reality that may be apprehended and that we can see it as it “really” is. Its epistemology adheres to the notion of objectivity, and the possibility of finding universal truths. Those who hold positivistic views see reality to exist autonomously from any observer, and inquirers can be objective and non-participant observers to the events that they see. The events can be represented by observer independent measurables called data that represent the “facts” of a single objective “reality”. Thus, for instance, a given investigation should always produce the same result for any observer if the theory about it is “true”, and if it is undertaken “scientifically” (though what is scientific is defined by positivists in terms of a set of propositions for practice that are consistent with a positivist epistemology). The truths set-up as a pattern of propositions represents knowledge. Through deductive reasoning, the approach usually embeds an attempt to test theory in order to improve both understanding of a situation and ability to make predictions about it. Positivism has a long tradition. It is sometimes referred to as mechanistic thinking as it paralleled the machine age western industrial revolution. In the last century, it was manifested within psychology as behaviourism, and it has some relevance to the learning theory of organisations (Yolles, 1999). Positivism purports that statements that emerge from a theory should be positive and testable, but this can be problematic in a complex world. This is because statements that may first appear to be simple may be quite complex and convoluted, and the creation of testable hypotheses may be highly problematic. Postpositivism arose as a counter to positivism to this. The problems of quantum mechanics, chaos and complexity have been captured in postpositivism with its “participatory interminglings” that links the observer and that being observed, rather than perceptions of objective things standing apart from human subjectivity (Fischer, 1998). In short, the

Beer’s ontological system 735

K 33,3/4

736

traditional understanding of the physical world as a stable or fixed entity is no longer adequate. However, there is no standard definition of “postpositivism”. Its base assumption is that reality exists, but can never be fully understood or explained, given both multiplicity of causes and effects and problem of social meaning. Objectivity can serve as an ideal, but requires a critical community of interpreters. Critical of empiricism, it emphasises the social construction of theory and concepts, and qualitative approaches to the discovery of knowledge. There is a critical tradition in postpositivism that is explained by Myers (1999). A critical realist believes that there is a reality independent of our thinking that science can study. While positivists are realists, the critical realism of postpositivists supports the notion that observation is fallible and has error and that all theory is revisable. That is, the critical realist is critical of our ability to know reality with certainty. Unlike the positivist, the critical realism of the postpositivist believes that the goal of science is to seek the truth about reality. Owing to the fallibility of measurement, the postpositivist supports the notion that multiple measures and observations are needed, each of which may possess different types of errors that can be related and therefore eliminated. The postpositivist also believes that all observations are theory-laden, so that scientific inquiry is inherently biased by their cultural experiences and worldviews. To deal with this, they reject the relativistic idea of the incommensurability of different perspectives. That is the idea that we can never understand each other because of our different worldviews developed through our individual cultural experiences, leading to our individually different patterns of knowledge that deliver meaning to us from our experiences and communications. Data information that defines replicated findings that create knowledge are considered to be true only probably. Myers further contends that most postpositivists are constructivists who believe that we each construct our view of the world based on our perceptions of it. This view will be developed further through the conceptualisations supported by Anderson (1999). It supports the notion that since perception and observation are seen to be fallible, our constructions of reality must be imperfect. Postpositivists do not believe that any individual can see the world perfectly as it really is. All scientific inquiry is biased and all observations are theory-laden. Objectivity in postpositivism is a social rather than individual characteristic involving critique across a subject area. Unlike Myers, Guba and Lincoln (1994) do not admit that postpositivists are technically constructivists. They agree that like positivists who support the notion of an objective reality, postpositivists believe that this may only be apprehended imperfectly and probabilistically, and only an approximate image of reality may be possible. A distinction between constructivists and postpositivists, however, is that the former believe that they can construct their own reality while the latter are constrained in this by a positivist reality. Another distinction

is that in positivism, replicable findings are assumed to be true. In postpositivism, however, the concept of falsification arises, where findings are examined critically to see if they can be shown to be false. Supportive of Guba and Lincoln, Fischer (1998) paints a picture of postpositivists as “interpretive consensualists” rather than constructivists. From an epistemological perspective, he tells us that empirical data that are accepted by consensus becomes knowledge through interpretative interaction with the perspectives of others. It is only by examining such data through conflicting frameworks that the presuppositions giving it meaning can be uncovered. The crucial debates now become centred on their underlying assumptions. Such deliberations produce new understandings in a process better framed as a “learned conversation” than the pursuit of empirical proof. Emphasis shifts from the narrow concerns of empirical-analytic theory to the development of “a rich perspective” on human affairs. Within this context, knowledge is the evolving conversation that is more accurately understood as consensually “accepted belief” than as proof or demonstration. Horkheimer and others developed critical theory in relation to the political disappointments at the absence of revolution in the West, the development of Stalinism in Soviet Russia, and the victory of fascism in Germany (Habermas, 1987; Held, 1980, p. 116). It was intended to explain mistaken Marxian evaluations without breaking with its fundamental intentions. The form proposed by Habermas (1970) is expressed in terms of his theory of knowledge constitutive interests. Critical theory is a blanket term that may be defined to include both postmodernism and poststructuralism since their epistemology supports the notion that inquiry is value determined (Guba and Lincoln, 1994). Epistemologically, critical theory is transactional and subjectivist. An inquirer and the situation being inquired into are assumed to be interactively connected through the values of the inquirer and any others involved (the inquiries). Any data information, taken as findings that create knowledge, are therefore value mediated. As such a properly constituted value laden personal interpretation of data information from findings made through personal values is seen to be valid. The ontology of critical theory tells us that while there may be a reality separate from experience, it can only be known through experience making it relative to the viewer. Since all individuals have distinct experiences, this leads to the notion that each view of reality is unique. In particular, it holds that reality is virtual as opposed to being tangible, and is shaped by social, political, economic, ethnic and other factors that crystallise over time. Myers (1999) tells us that inquirers who adopt a perspective through critical theory assume that social reality is historically constituted and that it is produced and reproduced by people. Although people can consciously act to

Beer’s ontological system 737

K 33,3/4

738

change their social and economic circumstances, critical inquirers recognise that their ability to do so is constrained by various forms of social, cultural and political domination. The main task of critical inquiry is seen as being one of the social critique, whereby the restrictive and alienating conditions of the status quo are brought to light. A critical framework focuses on the oppositions, conflicts and contradictions in contemporary society. It may be seen as the existence of any entity (a thing, object or event) that is perceivable and that can be experienced in some way. As such phenomena are what we shall call “cognitively demiurgic”: to be formative or creative, and deriving from the notion of one who fashions the material world out of chaos. It is consistent with the notion of creative observation as defined by Roy Frieden (1999) that physical reality is a result of an interaction between a viewer and acquired information. Unlike constructivism, it is not consciously constructed. However like it, reality is locally made. Following a lead offered by Carter et al. (1997), the underlying assumptions for critical theory can be identified as: . all thought is mediated by power relations that are socially and historically constituted; . the relationship between “concept and object” (relating to say an idea about something, and the thing itself) and “signifier and signified” (relating to say an indicator of something about a thing and that which is understood about it) is socially (and culturally) mediated rather than being fixed and stable; . language is central to the formation of subjectivity; . certain groups are privileged over others; . oppression (race, class, gender, age, for example) is reproduced when subordinates accept their status or situation as natural, necessary, and/or inevitable; . empirical data are interrogated with the intent of uncovering contradictions and negations in objective descriptions; . information always involves acts of human judgement and interpretation; . power is the basis of social groups; . there is no such thing as neutrality; . inquiry includes political action to redress injustices found in the inquiry process; . purposes focus on facilitating change and emancipatory action; . conceptual context, informed by the assumptions listed above, is explicit; . research questions focus on uncovering, provide a space for introspection, seek out multiple realities; . internal and external validity is replaced by critical trustworthiness.

Habermas (1987) has developed a view of critical theory that eventually developed into a theory of communicative action, concerned with the relationship between behaviour and communication. His earlier work identifies as part of its conceptual base, the notion of cognitive interest that is defined in terms of three attributes that have embedded within them learning domains that generate knowledge. The three attributes are: work, interaction and power (Habermas, 1970, 1971). They determine how knowledge can be identified and whether knowledge claims are warranted. It forms the basis of his theory of Knowledge Constitutive Interests (KCI) that forms a basis for his three worlds model. MacIsaac (1996) differentiates between these three forms of knowledge in the following way. . Work knowledge broadly refers to the way one controls and manipulates one’s environment. Commonly known as the instrument of action, knowledge is based on empirical investigation and governed by technical rules. The appropriateness of action is defined by the criterion “effective control of reality”. It is through empirical and analytical approaches that hypothetical and deductive theories can characterise the learning domains. . Practical knowledge occurs through communicative action within a process of social interaction. Social knowledge is created through consensual norms, and these define reciprocal expectations about the behaviour of others. While social norms can be expressed through empirical and analytical propositions, their validity is subjective and connected to intention. The determinant of nature of appropriate action is “clarification of the conditions for communication and intersubjectivity”, that is, the understanding of meaning rather than causality. . Emancipatory knowledge is individual self-knowledge or self-reflection. It involves an interest in one’s history, and biography is expressed in terms of self-image, roles, and social expectations. Ones desires (libidinal), institutional and environmental forces limit emancipation by constraining our options and rational control over our lives. Knowledge is gained by self-emancipation through reflection leading to a transformation in perspective. In critical theory, there is no absolute real world that can be separated out, because viewers create it within their frame of reference, and interact with their creation in a way that creates local or demiurgic phenomenology. There is therefore no separation between viewers and the behavioural world around them. Since what constitutes reality is determined through worldviews, it changes as worldviews change. In each worldview, we build our view of what we perceive to be the world through our mental models, created through a collection of conceptual extensions that form our patterns of knowledge. We may believe that we share the mental models with others, but mostly they will

Beer’s ontological system 739

K 33,3/4

740

be incommensurable (Yolles, 1999). This is because the mental models involve conceptual extensions, the meanings of which are not individually shared. This is because the meaning of the conceptual extensions that make them up is either not known or is qualitatively different. This results in a mismatch in meaning the models supposedly shared. We are never aware whether the shared models are related, except by attempting to draw meaning from others’ explanations provided through language, or comparing what we expect from the behaviour of people in a situation with what we perceive that they are doing. Let us now move on to forms of constructivism as supported by Anderson (1999), whose classification of constructivism is different from that of Guba and Loncoln. He explains that the species of constructivism adopt the base notions of Dewey (1938). According to Doolittle and Camp (1999) and Von Glasersfeld (1984), constructivism has four epistemological axioms: (1) knowledge is the result of cognitive processes, (2) cognition is an adaptive process that enhances the viability of behaviour for a given environment, (3) experience becomes meaningful through cognitive processes, and (4) knowing is created through biological/neurological as well as social, cultural, and language-based interactions. Thus, constructivism recognises that knowledge acquisition is a personal process that is based on experience. Therefore, the knowledge acquired by an individual will be personal, be connected with experiences, and have a validity that is related to the ability of an individual to relate it convincingly in terms of reality. Interestingly, these axioms are also valid for the critical theory paradigm, which may therefore be seen as yet another species of constructivism. However, this species of knowledge acquisition paradigm has a number of subspecies, named by Anderson (1999) as: cognitive, social and radical constructivism. Cognitive constructivism, according to Dole and Sinatra (1998), accepts the epistemological axioms (1) and (2) above, is linked to the learning process, and holds that external structures that exist in external reality can be accurately represented as internal models. Hence, structures and processes that are internally formulated can correspond to those of the real world, and reality is knowable to the individual, and like positivism supports the notion of naı¨ve realism. Knowledge construction is therefore considered to be primarily a technical rather than a subjectivist process of knowledge creation. Interestingly, from the brief descriptions provided here, it is not an easy matter to differentiate clearly between Anderson’s description of cognitive constructivism and Guba and Lincoln’s postpositivism, and ultimately a decision occurs through a matter of fancy or context. In Table II, we relate postpositivism with cognitive constructivism.

Both radical and social constructivism are closely related, each adopting all four axioms. Their distinctions are differentiated in that Vygotski (1978) sees knowledge processes being dependent initially on social processes in what we shall call naı¨ve knowledge acquisitors, while Piaget (1977) allows knowledge acquisition to be totally a subjective process. Radical constructivism is principally due to Piaget, whose propositions were designed to enable him to explain the capacity of children to learn (Doolittle and Camp, 1999). Knowledge acquisition is an adaptive process that results from active cognising by the individual learner, and while social interactions represent a source of knowledge, it occurs through internalisation by the individual. As such, knowledge has an internal nature, and the idea that, while an external reality may exist, it is unknowable to the individual knowledge is internal. This is because our experience with external forms is mediated by our senses, and our senses are not adept at rendering an accurate representation of these external forms (e.g. objects, social interactions). Therefore, knowledge is constructed from experience, and does not represent an accurate representation of external reality. The adaptive nature of knowledge supports the notion that knowledge cannot represent objective truth, and that internal knowledge is a viable model of experience rather than a reflection of external reality. Piaget theory of psychological constructivism is cybernetic, and holds that understanding is constructed through the interrelationship between the object, that is differentiated from self and towards which one acts without personal attachment, and the subject that is associated with personal attachment. This distinction is similar to that supported by Foucault (1974), and we shall return to this in a moment. Piaget’s perspective on cognitive learning may be constrained through that of Vygotsky’s (1978) social constructivism (Doolittle and Camp, 1999). Rather similar to radical constructivism, knowledge is a social phenomenon and is the result of social interaction, language usage and social discourse. However, the major departure from Piaget is that knowledge is seen to be a shared rather than an individual experience. The social collective seeks and finds truth through their collective interactive dialogue. Truth is therefore socially constructed in the collective consciousness defined by cultural co-participants. The social interaction occurs within a socio-cultural context, so that knowledge is bound to a specific time and place. Truth is socially constructed and agreed upon through common participation in cultural practices. Social constructivists are concerned not so much with the mental constructions of knowledge creation, but rather the co-construction of meaning within a social activity. In this sense, social constructivism is more concerned with meaning than structure. Cullen in 1999 notes that the social constructivist notions of the construction of knowledge focus on its social origins, and appear to have direct relevance to learning in organisational settings.

Beer’s ontological system 741

K 33,3/4

742

Vygotsky (1978) builds upon a foundational principle that all cognitive learning occurs at a social level prior to becoming individual. As such, others mediate cognitive learning, social dialogue is an important component of learning; and cultural tools (beliefs, artifacts, systems) are accessed and acquire meaning in social contexts. However, the mediated cognitive learning approach does not reflect particularly on the processes of knowledge intensification, only on the conditions that enable knowledge intensification to develop. We can refer to this as the mediation proposition. This leads us to the question of where Beer’s conception fit into the pattern of knowledge acquisition paradigms with his system/metasystem recursive process. Following Pickering (2002), Beer’s interest in the success or failure of organisations was a function of their adequacy in coping with their environment as the real world. However, this real world was classified by him as an “exceedingly complex system”, meaning that it was not exhaustively knowable, however much, one mapped it and theorised it, one would always be surprised by it. Hence, ontologically speaking, reality may exist, but it is not knowable. In this sense, it is similar to critical theory. We can consider Beer’s understanding of epistemology by recognising first, that the metasystem is the harbour of knowledge and the place where it is acquired. Consequently, the recursive principle is consistent with the notion that knowledge, embedded in language, is subjective to subordination as relative movement occurs from one subordinate focus to another, relative to a superior focus. In particular, Beer (1972, p. 228-9) implies, but states in Beer (1979, p. 311), quite clearly the constructivist nature of his work when he says: “systems are to be recognised subjectively; and their purposes exist only in the mind of an observer (or group of observers, who have themselves agreed on the conventions of their joint observation)”. This would seem to link with the constructivist conceptualisations of both Piaget and Vygotsky. As a result of this, we have presumed to interpret Beer’s perspective as given in Table I, with similarities to Habermas, Piaget and Vygotski. However, there is a difference between the constructivism of Piaget and Vygotsky, and that of Beer, Beer is a realist in the sense that ultimately reality is independent of our thinking about it (and this is implicit in his thinking when he developed his Viable Systems Model - see for instance Beer, 1979). However, he is also a constructivist, and his relativism is bounded by his realism. This ontology may be referred to as reality relativism. The system/metasystem dichotomy in viable systems Organisations that are viable can adapt to a changing environment. Variety in the environment of an organisation is determined by more or less distinguishable entities (elements, events or states) that occur within it. This is not problematic since to be adaptable, one has to see what it is that one must adapt to. These entities can be expressed in terms of time, space or purpose. The distinguishable entities may:

be constrained through relatively stable causal relationships between them in time and space, and appear to have a lack of constraint or be chaotic, when they appear to be loosely related such that one event or state cannot be clearly associated with another.

Beer’s ontological system

The variety of a system can be defined (Beer, 1979, p. 3) as the number of possible states that the system is capable of exhibiting. The basic condition of the complexity of a system is determined by its variety. Variety can therefore be seen to act as a measure of complexity. As environmental variety changes, so will environmental complexity. Organisational and social problem situations are often seen to arise with changes in complexity. We often see this as a natural development with, for example, the rise of new technologies and their consequence for existing labour mechanisms. The context of a situation that exhibits variety is important when discussing complexity. Thus, what we mean by variety will be dependent upon the context within which the system is placed by an inquirer. In this light, we can say that when we talk of the number of possible states in a situation that defines variety, then we are also talking about the worldview of an inquirer. A viable system is one that can be seen to be self-dependent, and thus take on an independent existence. At present, a system can be viewed as a set of hierarchies that together form a complex whole. In the same way as it is possible to explore the viability of an organisation as a whole, the viability of each focus can also be explored as a part of the system as a whole. This leads to a question posed by Beer. “If a viable system is one ‘able to maintain a separate existence’, how is it that a viable system contains viable systems which are clearly not separate from the viable system in which they are contained” (Beer, 1979, p. 118). The answer is that often parts of the system that might be identifiable as self-standing viable systems have other social, cultural, propositional, operational, or human constraints that do not enable them to separate out to work as independent viable systems. Having said this, it should also be noted that organisations might be responsible for their own demise due to their pathology. This is tied into their autopoietic nature, which should relate to ontological focuses and indeed the referencing focus that may define the whole. Autopoiesis should not be applied to an organisation’s metasystem. An example of when this might occur is when part of the metasystem attempts to control for the sake of control. Seeing control as a product of the organisation destroys the viability and autonomy of the broader system. In Beer’s terms, a system in this condition can be described as pathologically autopoietic (Beer, 1979, pp. 408-12). Ultimately, the pathology of a viable system concerns the failure of its cohesiveness. Beer (1959), in his development of managerial cybernetics, explored the nature of viable systems as he created his VSM. Viable systems participate in the autonomous development of their own futures. A viable organisation

743

.

.

K 33,3/4

744

participates in automorphosis[12], when it is responsible for and participates in changing in its own form, and thus enabling it to maintain appropriate operational behaviour under a changing environment and survive. The form is determined by its structure that both facilitates and constrains that behaviour. Its refinement over the OD methodology is that strategic decisions are not simply seen as an input to the system (Yolles, 1999). Rather, they derive from its metasystem that is responsible for manifesting and maintaining system structure. While OD sees the system itself as the transformation, the management cybernetics that underpins VSM invents a metasystem, and it implicitly supposes a transformation between the system and metasystem. Thus for instance, in OD strategy decisions are seen as inputs to the system, while in VSM they derive from the metasystem. In this way, the metasystem formally becomes one aspect of a structured inquiry. When decision-making is part of a formalised determinable process in an organisation, so the metasystem is also formalised, and decisions are made within it with respect to the perceived needs of the organisation at the level of focus concerned. This does not mean, however, that there may be another informal metasystem from which informal decisions derive. The metasystem ultimately operates through and is defined by the worldviews that determine the nature of the organisation. When a worldview exists formally it may be called its paradigm (Yolles, 1999). VSM is a generic model of the organisation that promotes principles of communication and control that help it to maintain its viability (Schwaninger, 2001). It is axiomatic in VSM that any organisation that can be modelled as a viable system can also be modelled as a set of five subsystems. They each represent an interactive function that act together as a filter between the environment and organisation’s management hierarchy, and connect management processes and their communications channels. The filter is sophisticated because it attenuates (reduces the importance of) some data while simultaneously amplifying other data. The filtered data are converted into information that is relevant to different levels of management within the organisation. A final control element addressed in the model offers auditing tools to make sure that the correct data are being collated. The audit channel mops up variety by sporadic or periodic checks. However, making sure that the appropriate data are assembled is only one of its functions. The VSM is defined in terms of five entities, referred to as system one (S1) to system five (S5), plus S3*, each with related communicative relationships (Yolles, 1999). S1 is defined as the system of operations (with its local management), and S3, S4 and S5 compose the metasystem. The meaning of each of the systems is described in Table III. While the epistemology of the VSM is taken care of by Beer, there is an ontological discussion. In exploring this we see that S1 is ontologically associated with the system, and S3-S5 are ontologically associated with the metasystem. However, S2, S3* and the

System

Ontology

Nature of system

1 Operations

System

System 1 is concerned with the system in focus (“the system”) and its behaviour. “Operations” provide a representation of what the system does and produces; it is usually broken down into functional units, and interacts with the environment through futures/planning. It is the system that is itself the subject of control. S1 interacts with the environment directly and through S4. There may be a number of perspectives from which to see system 1, and it may be seen from more than one by an organisation. For instance, system 1 could be seen in terms of product line, technology used, location, cycle time of products, customers, distribution channels, etc. System 2 can provide effective control. It concerns aspects of culture and is interested in limited synergy across divisions of an organisation. It tries to harmonise the culture and structure of the enterprise whilst also trying to reduce chaos and introduce order. It amplifies the control capability to try to induce self-regulation into its behaviour, which is in the implementation of operations. It can be seen as predominantly anti-oscillatory. It implements non-executive decisions like schedules, personnel and accounting policies and other areas governed by (legal and other) protocol. The aspect of culture it addresses is that of house style rather than the values/identity questions of S5 This function is concerned with effective regulation of the dynamic internal to the organisation. Integration/control is in charge of the functional units of the system. It controls and monitors what is going on. It is responsible for the implementation of policies, resource allocation, and the control and monitoring of the implementation activities. It determines information needs. It is involved in synergy related tasks Investigation, evaluation and validation of information flow between S1 and S3, noting that the link to S1 occurs only to pick up the information deficit associated with S2 This function is important to the identity of the organisation. Futures/planning involves issues of development and strategic planning. It observes the organisation from both internal and external views. It does this by gathering information from the environment and the system itself. It does all the future orientated tasks: research and development, training (except the orientation and maintaining skills at S2), recruitment, public relations, and market research. Consistent with the information gathering activities, it is also connected with the creation of knowledge

2 Coordination

In void

3 Integration/ control

In void (early work) or metasystem (later work)

3* Audit

In void

4 Future/ planning

Metasystem

(continued)

Beer’s ontological system 745

Table III. Nature of the VSM systems

K 33,3/4

746

Table III.

System

Ontology

Nature of system

5 Policy

Metasystem

This is concerned with the establishment and maintenance of a coherent context for the processes of the organisation. It relates to what the organisation sets out to do. It defines the direction of the organisation. It requires an accurate overview that represents the various dimensions of activity. Policy provides the systematic capability to choose from the different problem situations or opportunities thrown up by the environment. It is concerned with identity and cohesion and balances the present and future and internal with external perspectives

environment are normally portrayed within an ontological void (Table III) as illustrated in Beer (1979, p. 253). It must be stressed that this has no impact at all on the practical power VSM, concerned with the meanings associated with each S1-S5 and their associated messaging (epistemological migrations) along channels of communication (ontological migrations). Beer’s constructions, when considered from an ontological perspective, provide entry into a rich development that almost immediately connects with the work of Schwarz (1997) and Yolles (1999). Schwarz has developed a principally ontological theory that explains how persistent viable systems are able to maintain themselves, change and die. However, it does very little to engage with human activity systems, and it explains processes rather than provides for diagnosis as in VSM. In what follows we shall create a synergy between the conceptual constructions of Beer and Schwarz. It will result in a new paradigm that is capable of engaging the power of VSM while at the same time can draw on Schwarz’s elegant explanations. In creating this synergy such that it is true to Beer, however, we shall need to develop some additional conceptualisations that extend those of Schwarz into human activity systems. The three domains VSM Consistent with Table I, the needs of Table III, and with the construction of Schwarz, we shall define the three domains (Yolles and Guo, 2003): cognitive or existential (that can house a metasystem), the virtual or organising (that in terms of VSM would house System 2, System 3 and System 3*), and the phenomenal of behavioural (that can house the operational systems and their environments). Domain epistemology is shown in Figure 3, which is originally derived from the need to give a relatively practical explanation of change in China as it joins the World Trade Organisation (Yolles and Guo, 2003). Each domain has boundaries that condition their validity claims about reality. The boundary of one domain is differentiated from that of the others through its ontological horizon. In exactly the same way, as we considered the system/metasystem couple, this horizon maintains a content that varies depending on the cognitive perceiver that provides an entry into what may be

Beer’s ontological system 747 Figure 3. Influence diagram exploring the relationship between the phenomenal, virtual and existential domains

meaningfully reflected on, spoken about, or acted over. The three domains are ontologically coupled, and their horizons meld when they are seen as an emergent whole. However, this can only occur if the boundaries that differentiate the domains maintain ontological migrations that condition the melding process. Thus, ontological horizons both distinguish and connect differentiable validity claims about reality. Earlier, we introduced the ontology of the system and metasystem, and in Table IV, we consider the ontology of the three domains. Consistent with the notions of phenomenology, the three domains have boundaries that condition their realities. The boundary of one domain is differentiated from that of the others through its ontological horizon. In exactly the same way, as we considered the system/metasystem couple, this horizon maintains a content that varies depending on the cognitive perceiver that provides an entry into Types of domain Phenomenal or behavioural

Virtual or organising

Existential or cognitive

Nature of reality Material objects or events in interaction, the perception of which is conditioned by a cognitive knowledge-based frame of reference. It is cognitively demiurgic (meaning formative or creative), deriving from the notion of one who fashions the material world from chaos, and consistent with Frieden (1999) and Husserl (1950, p. 108) Symbolic or logical relational images that relate to phenomenal reality and involve purposeful organising. It is local to the experiences of the perceiver. Images of value and belief are maintained, partly represented through ethics and ideology. The domain is conditioned by a cognitive knowledge-based frame of reference The local belief-based creation of concepts and their patterns held in worldviews that establish a frame of reference, and determine what is known and their related meanings

Table IV. The three domains and their realities

K 33,3/4

748

what may be meaningfully reflected on, spoken about, or acted over. The three domains are ontologically related, and their horizons meld when the domains are seen as an emergent whole. However, this can only occur if the boundaries that create the horizons also harbour ontological connections that condition that melding. Thus, ontological horizons both distinguish and connect differentiable realities. These domains have properties (Table V), this notion inspired by Habermas’s (1970) Theory of KCI relating to the phenomenal domain, related to Beer’s ideas on the system/metasystem couple, and extended to the other domains. The idea of cognitive interests was Habermas’s, and in addition we have included the notion of cognitive purposes and influences. Additionally, we have adopted the notion of sociality properties that describes some of the capacities of the organisation as a whole (Yolles and Guo, 2003). It is interesting that some of Habermas’s conceptualisations in his KCI theory are directly reflected in Beer’s notions. Thus for instance, when Habermas talks of technical and practical interests, they are principally reflected in Beer’s system (S1) that links practical management control with the operational system. Beer’s principles of variety and requisite variety imply a host of soft issues that are associated with organisational processes, and are necessarily related to the emancipation and critical deconstraining of Habermas. However, Beer’s language is less than transparent in this respect and requires deep reading. It has therefore permitted critics of VSM to make incorrect statements about the inadequacies of VSM, for instance by Checkland (1980) who says that it misses the human meaning aspects of individuals (surprising realising the epistemological nature of VSM), and from Ulrich (1981) who suggests that tools of inquiry should have an ethical dimension. The three domains of Figure 3 exist in a first-, and second-order ontological couple that is expressed through its boundaries. The relationship between the phenomenal and virtual domains defines a first-order ontological couple. This is more usually referred to as autopoiesis (Maturana and Varela, 1979; Schwarz, 1997; Yolles and Dubois, 2001), but in fact autopoiesis is only an example of the ontological migrations that can occur when two ontologies are coupled together. It leads to the simple notion that the autopoietic capacity for a system can be directly related to its ability to manifest phenomenally its own virtual images through the self-production of usually structured intentional behaviour. We say usually because organisations operate through normative behaviour that is consistent with their expectations, and normative behaviour is normally regulated through structure. This does not mean that structure is a necessary condition for regularised behaviour to occur. Having said this, it is probably possible to express any mechanisms through which regularised behaviour occurs in terms of either implicit or explicit structure associated with the organisation in focus. The second-order ontological couple that we have referred to connects the existential or cognitive domain to the first-order ontological

Cognitive properties

Kinematics (through energetic motion)

Sociality properties Orientation Possibilities (through (determining trajectory) potential development)

Cognitive interests Phenomenal or behavioural (conscious) domain

Technical

Practical

Critical deconstraining

Work. This enables people to achieve goals and generate material well-being. It involves technical ability to undertake action in the environment, and the ability to make prediction and establish control

Interaction. This requires that people as individuals and groups in a social system gain and develop the possibilities of an understanding of each others subjective views. It is consistent with a practical interest in mutual understanding that can address disagreements, which can be a threat to the social form of life

Degree of emancipation. For organisational viability, the realising of individual potential is most effective when people: liberate themselves from the constraints; imposed by power structures and learn through precipitation in social and political processes to control their own destinies

Cognitive purposes Virtual or organising (subconscious) domain

Cybernetical

Rational/appreciative

Ideological/moral

Intention. This is through the creation and strategic pursuit of goals and aims that may change over time, enables people through control and communications processes to redirect their futures

Formative organising. Enables missions, goals, and aims to be defined and approached through planning. It may involve logical, and/or relational abilities to organise thought and action and thus to define sets of possible systematic, systemic and behaviour possibilities. It can also involve the use of tacit standards by which experience can be ordered and valued, and may involve reflection

Manner of thinking. An intellectual framework through which policy makers observe and interpret reality. This has an aesthetical or politically correct ethical orientation. It provides an image of the future that enables action through politically correct strategic policy. It gives a politically correct view of stages of historical development, in respect of interaction with the external environment

Beer’s ontological system 749

Table V. The three domains, their cognitive properties, and organisational (continued) patterning

K 33,3/4

750

Table V.

Cognitive properties Cognitive influences Cognitive (non-conscious) domain

Kinematics (through energetic motion)

Sociality properties Orientation Possibilities (through (determining trajectory) potential development)

Social

Cultural

Political

Formation. Enables individuals/groups to be influenced by knowledge that relate to our social environment. This has a consequence for our social structures and processes that define our social forms that are related to our intentions and behaviours

Belief. Influences occur from knowledge that derives from the cognitive organisation (the set of beliefs, attitudes, values) of other worldviews. It ultimately determines how we interact and influence our understanding of formative organising

Freedom. Influences occur from knowledge that affect our polity determined, in part, by how we think about the constraints on group and individual freedoms, and in connection with this to organise and behave. It ultimately has impact on our ideology and morality, and our degree of organisational emancipation

couple. An example of this ontological migration is autogenesis, that represents the self-production of the rules of production, and that can therefore be expressed in terms of the creation of principles that are able to guide self-production. The ontological migrations shown in Figure 4, will be discussed later. The use of autopoiesis within the context of social communities It may here be noted while we have referred to autopoiesis in a social context, there is an argument that this is not appropriate to this. Following

Figure 4. Relationship between normative belief system in a social community and patterns of knowledge that it develops

Mingers (1995), Maturana and Varela (1980) developed the concept of autopoiesis[13] within the sphere of biology applied to living systems. They do not see social systems as an appropriate application because they are not living systems and cannot self-produce the components that comprise them. Beer (1980) notes that the purpose of Maturana and Varela (1980) is “to understand the organization of living systems in relation to their unitary character. This formulation of the problem begs the question as to what is allowed to be a called a living system, as they themselves admit.” From an epistemological perspective, Beer (1980, p. 68) does not see that the need for systems to be living stands in the way of social systems being seen as autopoietic: The fact is that if a social institution is autopoietic (and many seem to answer to the proper criteria) then, on the authors’ own showing, it is necessarily alive. That certainly sounds odd, but it cannot be helped. It seems to me that the authors are holding at arms length their own tremendously important discovery. It does not matter about this mere word “alive”, what does matter is that the social institution has identity in the biological sense; it is not just the random assemblage of interested parties that it is thought to be. When it comes to social evolution then, when it comes to political change: we are not dealing with institutions and societies that will be different tomorrow because of the legislation we passed today. The legislation – even the revolution – with which we confront them does not alter them at all; it proposes a new challenge to their autopoietic adaptation. The behaviour they exhibit may have to be very different if they are to survive: the point is that they have not lost their identities.

Beer (1980, p. 71) consequently shows that he is not neutral to whether or not social systems can be autopoietic, as he also argues epistemologically that: . . .any cohesive social institution is an autopoietic system because it survives, because its method of survival answers the autopoietic criteria; and because it may well change its entire appearance and its apparent purpose in the process. As examples I list: firms and industries, schools and universities, clinics and hospitals, professional bodies, departments of state, and whole countries. If this view is valid, it has extremely important consequences. In the first place it means that every social institution (in several of which any one individual is embedded at the intersect) is embedded in a larger social institution, and so on recursively – and that all of them are autopoietic. This immediately explains why the process of change at any level of recursion (from the individual to the state) is not only difficult to accomplish but actually impossible – in the full sense of the intention: “I am going completely to change myself ”. The reason is that the “it”, that self-contained autopoietic “it”, is a component of another autopoietic system. Now we already know that the first can be considered as allopoietic with respect to the second, and that is what makes the second a viable autopoietic system. But this is in turn means that the larger system perceives the embedded system as diminished as less than fully autopoietic. That perception will be an illusion; but it does have consequences for the contained system. For now its own autopoiesis must respond to a special kind of constraint: treatment which attempts to deny its own autopoiesis. Consider this argument at whatever level of recursion you please. An individual attempting to reform his own life within an autopoietic family cannot fully be his new self because the family insists that he is actually his old self. A country attempting to become a socialist state cannot fully become socialist; because there exists an international autopoietic capitalism in which it is embedded, by which the revolutionary

Beer’s ontological system 751

K 33,3/4

752

country is deemed allopoietic. These conclusions derive from entailments of premises which the authors have placed in our hands. I think they are most valuable.

In exploring the argument against social system autopoiesis, Mingers (1995, p. 123) defines the ontological argument that inhibits social systems being seen as autopoietic. For this he identifies the following three “problematic” elements. (1) Centrally, autopoiesis is concerned with the processes of production – the production of those components that constitute the system themselves. (2) It is constituted in temporal and spatial relations, and the components involved must create a boundary defining the entity as a unity – that is, a whole interacting with its environment. (3) The concept of autopoietic organisation specifies nothing beyond self-production. It does not specify particular structural properties and thus should not need to be modified for social systems. There is a concern explored by Mingers that these elements cannot be legitimately applied to social systems, presumably because it is unclear how this can occur directly in terms of individuals and groups. While he discusses Luhmann’s (1986) approach to autopoiesis in social systems later in his book, he does not appear to highlight that because it may not be the individual or the group that is self-produced, but the components that enable the social group to exist. Thus for instance, Luhmann’s model centres on communications and the self-production of communications. In the same way, social systems produce patterns of knowledge, myths, behaviour, and other things to which autopoiesis can similarly be applied. Mingers (1995, p. 125) notes, however, that “a more radical approach” is to apply autopoiesis to concepts or ideas, though why this radical is unclear. In another vein, Mingers (1995, p. 124) notes that the fundamental problems of autopoiesis for social systems are not significant if they are applied metaphorically “in helping our thinking, or that a more generalised version, such as Varela’s idea of organisational closure, could be fruitfully applied.” Having said this, Mingers also indicates that metaphors produce merely metaphoric results, and thus they have no greater claim on our attention. Consistent with this view, Beer (1989) suggests that comparisons deriving from metaphor should not be taken too seriously. These representative views about the limitations of metaphor relate to those that are on a par with simile, which take experiences from one domain and apply them to another directly. However, unlike simile, metaphor is often purposefully abstracted and elaborated, leading to more profound and significant comparisons. For Brown (2003), metaphor is very important to the development of science, facilitating mature knowledge and understanding. Based on his characteristics of metaphor, we list the following.

(1) Metaphors, like simile, begin with literal everyday experiences in a source domain that is necessarily local and culturally based. (2) Metaphors are mapped from the source domain to a sink domain (where it is used). The aim is to enlarge and enhance understanding of situations in that sink domain. These understandings ultimately derive from direct experiences that enable us to create more abstract conceptualisations. (3) A given metaphor may highlight certain features of the source domain and may obscure others. Obscured features are often implied or inferred through context, and this can make the metaphor a powerfully creative force in scientific reasoning. (4) Although metaphors invite comparisons of two disparate things, the more interesting metaphors do more than this. They stimulate creation of similarities between the source and sink domains, such that the latter is seen in an entirely new light. (5) Metaphors in science serve an explanatory role and are a stimulus to new inquiries. They may be very simple and evocative initially, then grow more detailed as research findings support or disconfirm inferences drawn from the initial metaphor. (6) Metaphors may be elaborated, when they are extended and abstracted, and also perhaps individually or in plural convergence so can form models. These models may have associated with them metaphorical entailments that influence how they are understood and applied. Models commonly form a basis for theory creation. They may constitute primary[14] propositions, and when this occurs they need to be evidenced. As an example of this, we note Beer’s (1989) reference to his VSM that he considers to be a generic[15] model for the social domain, and rather than talking about evidencing it, he equivalently refers to it as being testable and verifiable. Scientific principle may be thought of as a literal representation of an elaborated metaphor, a statement that we shall explore briefly. While metaphors are grounded in experience, scientific principles are grounded in facts. However, what is fact? In one of Beer’s writings, he said that facts are “fantasies that you can trust”, where we can take trust to be a firm belief [16], and where fantasies at there best can be a “subjective interpretation of information”. Trust, however, occurs through belief, and it should therefore be realised that it can vary from individual to individual, from group to group, or from time to time. In other words, it is a cultural phenomenon. From a constructivist perspective, this must mean that since scientific principle are grounded in fact, and fact is culture relative and not absolute, scientific principle must also be relative. A simple illustration of this arises from a brief examination of the conflict between the supporters of the wave and particle theories of light (Hoffman, 1947).

Beer’s ontological system 753

K 33,3/4

754

The distinction between metaphor and scientific principle therefore becomes less differentiable, as can be illustrated through an example. The system is a conceptual construction that constitutes an elaborated metaphor. It operates as an abstracted ideal, and used non-literally in a sink domain aids the process of inquiry and the creation of intervention strategies for improvement where this is desirable. This is a constructivist view that Beer (1980) supported, when he tells us “a system is not something presented to the observer, it is something recognized by him”. Sometimes the knowledge and language of an extended metaphor becomes so embedded in the sink domain that it becomes a frame of reference, and any scientific principle that develops become grounded in the metaphor(s). When this happens, it becomes very difficult to distinguish between extended metaphor(s) and resulting scientific principle. Hence, metaphors can be as important as the scientific principles that rest on them. In developing our viable systems theory conceptualisations, we create a frame of reference that like the system should be thought of as an abstract metaphor. As such, the ontological constructions that appear here should not be seen in positivist terms, as might be the case if we were attempting to create literal causative models. They operate to assist the formation of explanations about social community pathology, which may more pragmatically enable an inquirer to reflect on ways of creating intervention strategies for improvement. The abstract metaphor that we are using is a development of philosophical questions that ask what is the nature of reality (ontology) and knowledge (epistemology). Systems concepts are normally framed in an epistemological frame of reference. Thus, it may be asked, “how can we improve a given complex situation for a social community”, where the notion of improvement implies the acquisition of knowledge such that this can happen. It is a rarity for systems concepts to be defined in ontological terms. The reason is that reality is usually taken for granted because it cannot apparently provide a route, like epistemology, for improvement. However, one of the reason that our social communities are pathological is that we each individually have our own realities, and when we form into bounded groups these too ascribe to new normative bounded realities. These realities form with the development of local paradigms that are the concern of epistemology. In this sense, epistemology and ontology can only be divorced analytically, not practically or pragmatically. However, the analytic and pragmatic approaches are different sides of the same coin, especially if the analytic approach is explicitly intended for use to satisfy the pragmatic one. Mingers (1995, p. 151) discusses the metaphorical use of autopoiesis by Morgan (1986) whose thesis is that the organisation is influenced by its own internal self-image or identity. They are continually concerned to recreate and maintain their image and identity by projecting themselves onto environments, and what they monitor is a reflection of their own concerns and interests. While there is more to the theory than this, its basic tenets are consistent with the

notions embedded in the ontological arguments of Eric Schwarz (Yolles, 1999), who developed his abstract analytic ontology that we are applying in principle to social community. Schwarz is concerned with the ontological perspective that explains the dynamic that enables autonomous systems to maintain their viability, and his constructions explored the nature of autonomy in terms of autopoiesis and its second-order form autogenesis. We assert that there is a relationship between autonomy and autopoiesis in social communities, an argument that comes, for instance, from Jessop (1990). He defines autopoiesis as a condition of radical autonomy that enables a system to define its own boundaries relative to its environment and its own operational code. It implements its own programmes, reproduces its own elements in a closed circuit, and obeys its own laws of motion. When it has “autopoietic take-off”, its operations can no longer be directly controlled from outside, though there may be a variety of indirect controls that in part constitute its “environment”. When we talk of autonomous systems, we are often interested in autopoiesis, and conversely when we talk of autopoiesis, we are normally concerned with radical autonomy. However, it can be argued that the characteristics that constitute the condition of radical autonomy may have a subjective dimension. It must be stressed at this juncture that even though we consider that this construction, like that of the system, is a metaphor, this does little to weaken the importance of theoretical arguments to pragmatic approaches of inquiry. Recursion There are epistemological implications to our ontological construction that relates to the notion of recursion, so important to VSM. Maturana (1996) explores the nature of reality, regarded as: a proposition that we use as an explanatory notion to explain our experiences. . .. [beyond this] it is that which in our living as human beings we live as the fundament of our living. Under these circumstances, reality is not energy, not information, however powerful these notions may appear to us in the explanation of our experiences. We explain our experiences with our experiences and with the coherence of our experiences. That is we explain our living with our living, and in this sense we explain human beings as constitutively the fundament for all that exists, or may exist in our domains of cognition.

Explaining our experiences with our experiences is a recursive phenomenon, enabling whatever images of reality that we perceive to be embedded within other images, like two mirrors at an angle reflecting an image of an object to infinity. This is effectively a recursive frame of reference, and each image represents a new validity claim about reality that is contextualised by the validity claim in which it is embedded. This idea allows us to talk about recursion, by which we mean that each of the three domains can, through the local context of its own validity claim about reality, recursively host the set of three domains. When this happens, the host domain has a validity claim that is ontologically distinguished. When the domain hosts other relative domains

Beer’s ontological system 755

K 33,3/4

756

Figure 5. Embedding the three domain model into the existential domain

within it, they are capable of formulating finer, more local validity claims about reality. Let us illustrate this. Phenomenal reality can be apprehended by a unitary consciousness from which a single person responds to his or her phenomenal experiences. Alternatively, a socially plural consciousness with distinguishable complexities may be defined, for which coherent social behaviour occurs phenomenally. This is enabled through phenomenal structures that anticipate [17] a plurality of commonalities and norms, and an expectation for behavioural adherence to them. It is within the virtual domain that images of these arise that enables the phenomenal structures and behaviours to be manifested in the first place. They are defined in the conceptual domain through the knowledge that constitutes such commonalities and norms. This is only possible because of the recursive nature of the domains within the conceptual domain, through which the commonalities and norms are manifested through the interaction of a plurality of consciousnesses. It may be noted that the commonalities and norms that have arisen to create a paradigm for the group arose originally through the creation of a virtual paradigm in the virtual domain at another level of recursion. In this case, the paradigm itself with its shared concepts and their structured interconnections that constitutes a pattern of normative knowledge would have been associated with the phenomenal domain. We can apply the concept of recursion to our three domains model. Unlike Beer’s VSM, this is not intended to diagnose the system. Rather, like the work of Schwarz (1997), its purpose is to provide explanations for the complex organisation that relate to its operational behaviour. This may or may not provide additional ways of diagnosing the organisation. We illustrate the notion of recursion and its significance for explanation in Figures 4 and 5. Figure 4 shows the relationship between normative organisational beliefs and the patterns of objectified knowledge it has. Figure 5 shows how recursion can occur in this model. The postulated model in Figure 4 is not claimed as valid, and the notion that the relationship between an

organisation’s paradigm and its patterns of knowledge is an autopoietic one is sheer hypothesis. It postulates that an organisation can maintain its own patterns of knowledge as structures that can be represented in the phenomenal domain. Likely these patterns are explicit and can be expressed as propositions that underpin the organisation’s modus operandi. They derive from the dominant paradigm (if one exists) that the organisation maintains and from which it operates. In viable organisations, the relationship between the paradigm and patterns of knowledge may well be expressed as an autopoietic process. Earlier, we indicated that this relationship between the virtual and phenomenal domains is a first-order ontological couple. The second-order ontological couple links from the existential domain to the first-order couple. To illustrate recursion, in Figure 5, we have embedded the three domains in the existential domain to explain how the normative belief system arises in the first place from a plurality of them connected with the individuals that make up the social community. Normative processes develop during communication between participants of an organisation through the lifeworld[18]. The recursion in Figure 5 postulates how a normative belief system emerges in an organisation from a plurality of individual belief systems. Through autogenesis common, principles of lifeworld interaction about belief systems emerge that enable a plurality of competing images of what belief system is to hold to be managed. It is through autopoiesis that these competing images are self-produced as a normative organisational belief system. The normative belief system that results is now reflected as the existential domain for Figure 4. In Figure 6, we develop a further model that deals with co-evolutionary development. This model derived from Yolles, explores the relationship

Beer’s ontological system 757

Figure 6. Indication of the ontological relationship between adaptation and co-evolution “Man is a prisoner of his own way of thinking and his own stereotypes of himself” Beer (1975, p. 15)

K 33,3/4

758

Figure 7. Situation during pathological autopoiesis with bounded variety options

between adaptation, self-organisation [19] and co-evolution, the first two of these concepts are a serious concern of Beer. Interestingly, this representation now gives autopoiesis a simple form of expression, illustrating that it is an organisation’s ability to manifest its internal images of itself and its future into phenomenal/behavioural reality. It may be noted here that Figure 4 would also appear to give clear meaning to Beer’s (1975, p. 15) statement that “Man is a prisoner of his own way of thinking and his own stereotypes of himself”. The new ontology of the three domains model provides the capability of more easily appreciating the notion of pathological autopoiesis, a term that is easily open to a variety of interpretations. This is primarily because autopoiesis is an ontological condition, and if one does not engage ontological arguments the notion of autopoiesis can become convoluted and unclear. Viewed from the ontological perspective of Schwarz, the meaning of pathological autopoiesis is very clear. Since autopoiesis is the capacity of a social community (or individual) to establish/produce its image of itself and its future as a pattern of behaviour, pathological autopoiesis must mean that the social community gets locked into this, thereby decoupling the ontological connection (autogenesis) to autopoiesis as shown in Figure 6. We have represented this situation of pathological autopoiesis in Figure 7 as a development of Figure 6. The pathology leads to a stationary image of oneself and the future with whatever embedded variety it may have in it. Adaptation can occur, but if none of the possibilities available within that image are adequate to deal with the changing environment, then a lack of capacity for adaptation occurs. In general, while it might appear that an evolutionary process is under way, this is not the case since a host of variations available to the community will be called on, but no new evolutionary ones will develop. As a consequence, there is no possibility for a co-evolutionary process. This type of

situation therefore explains the onset of the eventual demise of a species of social community, when all of its variety has been used up without success. The legacy of Stafford Beer and the dynamics of paradigm change We have discussed the contribution that Beer has made to organisational theory through his introduction of constructivism, adopting the theoretical ideas of the metasystem and recursion, and giving them practical capacity. In arguing this, we have also discussed the idea that paradigms change, and in doing so that pass through a virtual stage that they may not survive. This brings us to an interesting juncture, which is how do we perceive the legacy of Beer’s conceptualisations. The problem we have here relates to what Iles and Yolles (2002) and Yolles (2000) call knowledge migration that explains the epistemological distance between the semantic implications seen in a communication by a message source and semantic inferences applied to a communication by a message sink. This epistemological distance results in the acquisition of distinct information and the creation (not re-creation) of knowledge that is catalysed by the communication, not embedded in it. When people communicate they send messages that carry meaning, and thus embeds knowledge, in coded form. To encode the message the source of the message uses their current patterns of knowledge to encode the message. The message sink does something similar. In a paradigm that adopts the epistemology and ontology of positivism, knowledge migration is simply knowledge transfer. However, in the critical theory approach adopted in this paper, every communicator has their own unique pattern of knowledge defined by their experiences and contextualised by their culture. This means that the knowledge that is assembled by each message sink is not a reconstruction of the knowledge of a source, but is rather a knowledge re-creation facilitated and catalysed by the message, and it is unique. This is complexified by the idea that every message has a horizon of meanings, those things implied by those who know, but not made explicit. There is a problem therefore, when a new paradigm arises. It is that each person who interacts with it is likely to create what Yolles (1999) calls a doppelganger virtual paradigm. It is a new species of the genus that has re-interpreted or recreated the new paradigm. Unless this shift is substantive in that it has the capacity to introduce a new conceptualisation that fundamentally alters the frame of reference of the original paradigm, contestations can be fed back through lifeworld processes that involve response and debate. Sometimes this is not sufficient, and contested differences become elaborated, and result in conflict. This process is explained by Yolles (2001). This process of paradigm contestation can become exacerbated with the demise of the father of the paradigm. In the case of Beer’s cybernetic theory of management, there is no possibility of a feedback control process, and the consequence is that bloody paradigmatic revolution can result. This author

Beer’s ontological system 759

K 33,3/4

760

wonders therefore, whether the legacy of Beer’s ideas will fragment into not only a set of virtual paradigms, but also whether the result will be destructive bloody conflict. To ensure that this does not happen, the operational research and systems community needs to establish a metasystem in which the operational subsystems are the species of virtual paradigms. For the sake of simplicity, we can call this a metaparadigm. I pose this as a challenge to the OR and systems community. Notes 1. To illustrate that Beer was a constructivist, and held such principles at least in the same period as those whose names are assigned to this, we will be obliged to explore the notion of constructivism in this paper, and it has led to an appreciation of an apparent conflict of view in the literature. 2. Webster online dictionary. 3. The complement to this is weak anticipation that can be associated with strategy. 4. Swanson (2001) differentiates between base and auxiliary concepts. Auxiliary concepts describe base concepts. 5. “. . .scientists, just like the rest of humanity, carry out their day-to-day affairs within a framework of presuppositions about what constitutes a problem, a solution, and a method. Such a background of shared assumptions makes up a paradigm, and at any given time a particular scientific community will have a prevailing paradigm that shapes and directs work in the field. Since people become so attached to their paradigms, Kuhn claims that scientific revolutions involve bloodshed on the same order of magnitude as that commonly seen in political revolutions, only the difference being that the blood is now intellectual rather than liquid. . .the issues are not rational but emotional, and are settled not by logic, syllogism, and appeals to reason, but by irrational factors like group affiliation and majority or ‘mob’ rule” (Casti, 1989, p. 40). 6. Zeno’s paradox is concerned with the impossibility of moving between two points A and B in space. To reach B from A, one must travel half the distance to it to a point say a1, and to go from a1 to B you must reach a point half way to it at a2. This argument is recursive as you move to a3, a4, a5,. . .. To count the full distance that you have travelled you must add all of the half distances that form an infinite series, suggesting mathematically that you can never reach B. The solution to the paradox is to introduce time as a new analytically and empirically independent conceptual extension that operates as a limiting factor on the summation. The introduction of this new conceptualisation has meant that a new paradigm has been created with new propositions and beliefs, and it is thus incommensurable with the previous paradigm since it creates a new conceptual extension through which new ways of seeing can be created (Yolles, 1998). 7. This happens in all paradigmatic environments, whether they relate to the cultural basis of an organisation – for instance, in the privatisation of public companies (Yolles, 1999), or of a discipline of science as that being considered here. 8. For example, see Yolles (1999), referring to the work of Flood and Jackson (1991). 9. The idea of ontological horizon may be developed by referring to Ladrie¨re (2002). 10. According to the American Heritage Dictionary 4th edition 2000 online, meld means to merge or blend (e.g. a meld of diverse ethnic stocks). In our context, it relates to a process of de-differentiating that is a consequence of emergence.

11. An ontological migration enables validity claims about one reality to be migrated to another ontologically coupled reality. 12. The term self-organising is normally used here, but within the context of this paper, it can be misleading in that it can be supposed to be part of an “organising” domain, rather than what it is, associated with system structure and its manifest behaviour. It is for this reason that we refer to it as automorphosis, or self-change-of-form, relating to the concept of morphogenesis (Yolles, 1999). 13. Mingers (1995) notes that this word autopoiesis, also referred to as self-production, comes from auto as self as opposed to alloi as other, and poiesis as bringing forth, in this context with respect to production. 14. Propositions constitute knowledge. Axioms are base propositions that are cultural statements of belief, need no demonstration, and underpin the primary propositions that may be elaborated and perhaps generalised abstractions of a metaphor. Secondary propositions are derived consequences from primary propositions and may describe a particular characteristic of them. Following Keynes (1973), such secondary propositions may be claimed to support rational belief. 15. Beer’s language is different, and rather than talk of a model being generic, he rather uses the more formal logical word homomorphism. 16. Collins Reference Dictionary, 1992. 17. When we say anticipation, we are actually referring to “strong anticipation” (Yolles and Dubois, 2001), relating to the nature and relationship of the boundaries of the three domains and their validity claims about reality. 18. According to Habermas (1987), lifeworld is a transcendental site where speakers and hearers meet for intersubjective affairs like dealing with validity claims, settle disagreements, achieve agreements. It has both teleological and communicative aspects of a management situation. Lifeworld defines patterns of the social system as a whole, and is associated with culturally transmitted background knowledge. 19. The term automorphosis is used here for self-organisation. The reason is that the virtual domain can be called as an organising domain, and it is better to be sure that self-organisation is part of the phenomenal domain and relating to self-change of form. References Ackoff, R.L. (1971), “Towards a system of systems concepts”, Management Sciences, Vol. 7 No. 11, p. 27. Anderson, J.R. (1999), Rules of the Mind, Erlbaum, Hillsdale, NJ. Beer, S. (1959), Cybernetics and Management, English U. Press, London. Beer, S. (1966), Decision and Control, Wiley, Chichester. Beer, S. (1972), The Brain of the Firm, Wiley, Chichester. Beer, S. (1975), Platform for Change, Wiley, Chichester. Beer, S. (1979), The Heart of the Enterprise, Wiley, Chichester. Beer, S. (1980), Autopoiesis: The Organization of the Living, pp. 63-72 and a preface to Maturana, H. and Varela, F.J., Autopoiesis and Cognition, Boston Philosophy of Science series, Vol. 40, available at: http://www.cogsci.ed.ac.uk/,jwjhix/Beer.html Beer, S. (1989), “The viable system model: its provenance, development, methodology and pathology”, Journal of Operational Research Society, Vol. 35, pp. 7-26. Bertalanffy, L.V. (1951), “General systems theory: a new approach to the unity of science”, Human Biology, Vol. 23, pp. 302-61.

Beer’s ontological system 761

K 33,3/4

762

Boulding, K.E. (1956), “General systems theory – the skeleton of a science”, Management Science, Vol. 2, pp. 197-208. Brown, T.L. (2003), Making Truth, Metaphor in Science, University of Illinois Press, Urbana and Chicago. Burrell, G. and Morgan, G. (1979), Sociological Paradigms and Organisational Analysis, Heinemann, London. Carter, V., Dohrer, T., Duque, G., Fulton, T. and Shipley, C. (1997), Critical Theory and Qualitative Research, available at: http://www.ed.psu.edu/ci/cs597/definition.html (note that this Web site appears to have demised, but despite this it is still necessary to refer to the source that inspired the summary that we provide here). Casti, J.L. (1989), Paradigms Lost, Abacus, London. Checkland, P. (1980), “Are organisations machines?”, Futures, Vol. 12, p. 421. Churchman, C.W., Arnoff, E. and Ackoff, L.R. (1957), Introduction to Operations Research, Wiley, New York, NY. De Bono, E. (1977), The Happiness Purpose, Penguin Books Ltd, Harmondsworth, ISBN: 0140220585. Dewey, J. (1938), Experience and Education, Macmillan, New York, NY. Doolittle, P.E. and Camp, W.G. (1999), “Constructivism: the career and technical education perspective”, Journal of Vocational and Training Education, Vol. 16, p. 1. Dole, J.A. and Sinatra, G.M. (1998), “Reconceptualizing change in the cognitive construction of knowledge”, Educational Psychologist, Vol. 33 Nos. 2/3, pp. 109-28. Ellis, W.D. (1938), A Source Book of Gestalt Psychology, Routledge and Kegan Paul, London. Fischer, F. (1998), “Policy inquiry in postpositivist perspective”, Policy Studies Journal, Vol. 26 No. 1, pp. 129-46. Flood, R.L. and Jackson (1991), Creative Problem Solving: Total Intervention Strategy, Wiley, New York, NY. Fogel, L.J. (1967), Human Information Processing, Prentice-Hall, Englewood Cliffs, NJ. Foucault, M. (1974), “Human nature”, in Elders, F. (Ed.), Reflexive Water: The Basic Concerns of Mankind, Souvenir Press, London. Frieden, R. (1999), Physics from Fisher Information: A Unification, Cambridge University Press, Cambridge, MA. Go¨del, K. (1931), Kurt Godel: Collected Works, Oxford University Press, Oxford, UK. Guba, E.G. and Lincoln, Y.S. (1994), “Competing paradigms in qualitative research”, in Denzin, N.K. and Lincoln, Y.S. (Eds), Handbook of Qualitative Research, Sage, Thousand Oaks, CA, pp. 105-17. Habermas, J. (1970), “Knowledge and interest”, in Emmet, D. and MacIntyre, A. (Eds), Sociological Theory and Philosophical Analysis, MacMillan, London, pp. 36-54. Habermas, J. (1971), Knowledge and Human Interests, Beacon Press, Boston. Habermas, J. (1987), The Theory of Communicative Action, Vol. 2, Polity Press, Cambridge, MA. Held, D. (1980), Introduction to Critical Theory: Horkheimer to Habermas, Hutchinson, London. Hoffman, B. (1947), The Strange Story of the Quantum, Penguin books, Middlesex. Husserl, E. (1950), “Ideen zu einer reinen Phanomenolgie und phanomenologischen Philosophie”, Vol. 1, Husserliana, Vol. 3.

Iles, P. and Yolles, M. (2003), “International HRD alliances in viable knowledge migration and development: The Czech Academic Link Project”, Human Resource Development International, Vol. 6 No. 3, pp. 301-24. Jessop, B. (1990), State Theory, Polity Press, Cambridge, MA. Keynes, J.M. (1973), “A treatise on probability”, in Moggridge, D. (Ed.), The Collected Writings of John Maynard Keynes, Vol. VIII, Macmillan Press, London. Kuhn, S.T. (1970), The Structure of Scientific Revolutions, University of Chicago Press, Chicago, IL. Ladrie¨re, J. (2002), “Technical universe in an ontological perspective”, Society for Philosophy and Technology, Vol. 4 No. 1, available at: http://scholar.lib.vt.edu/ejournals/SPT/v4nl/ LADRIERE.html Luhmann, N. (1986), “The autopoiesis of social systems”, in Geyer, F. and van der Zouwen, J. (Eds), Sociocybernetic Paradoxes, Sage Publications, London. Luhmann, N. (1995), Social Systems, Translated from the German edition on 1984, Stanford University Press, California. MacIsaac, D. (1996), The Critical Theory of Jurgan Habermas, available at: http://www.physics. nau.edu/,danmac Maturana, H. (1996), “Metadesign”, Instituto de erapia ognitiva, Publicaciones, available at: http:// www.inteco.cl/articulos/006/doc_ing4.htm Maturana, H. and Varela, F.J. (1979), Autopoiesis and Cognition, Boston Studies in the Philosophy of Science, Boston. Maturana, H. and Varela, F.J. (1980), “Autopoiesis and Cognition”, Boston Philosophy of Science series, Vol. 40. Midgley, G. (2000), Systemic Intervention: Philosophy, Methodology, and Practice, Kluwer Academic/Plenum Publishers, New York, NY. Mingers, J. (1995), Self-Producing Systems, Plenum Press, New York, NY and London. Morgan, G. (1986), Images of Organisation, Sage, CA. Myers, M.D. (1999), Qualitative Research in Information Systems, available at: http://www. auckland.ac.nz/msis/isworld/index.html Nadler, D.A. and Tushman (1977), Feedback and Organisations Development: Using Data Based on Methods, Addison-Wesley, Reading, MA. Nadler, D.A. and Tushman (1979), “A congruence model for diagnosing organisational behaviour”, in Kolb, D., Rubin, I. and McIntyre, J. (Eds), Organisational Psychology: A Book of Readings, 3rd ed., Prentice-Hall, Englewood Cliffs, NJ. Palmer, K.D. (2000), The Advance of the Systems Engineering Discipline Through an Understanding of Meta-Systems Theory, Meta-System Engineering Futures, available at: http://archonic.net/incose2002/fs00b02.htm Piaget, J. (1977), “The development of thought: equilibrium of cognitive structures”, Viking Press, New York, NY. Pickering, A. (2002), Cybernetics and the Mangle: Ashby, Beer and Pask, available at: http://www. soc.uiuc.edu/faculty/pickerin/cybernetics.pdf Rosenblueth, A., Wierner, N. and Bigelow, J. (1943), “Behaviour, purpose and teology”, Philosophy of Science, Vol. 10S, pp. 18-24. Schutz, A. and Luckmann, T. (1974), The Structures of the Lifeworld, Heinamann, London.

Beer’s ontological system 763

K 33,3/4

764

Schwaninger, M. (2001), “Intelligent organisations: an integrative framework”, Sys. Res., Vol. 18, pp. 137-58. Schwarz, E. (1997), “Towards a holistic cybernetics: from science through epistemology to being”, Cybernetics and Human Knowing, Vol. 4 No. 1, pp. 17-50. Swanson, G.A. (2001), “Management observation and communication theory and organisational information”, Sys. Res., Vol. 18 No. 46.5, pp. 429-41. Turchin, V. and Joslyn, C. (1999), The Metasystem Transition, available at: http://pespmc1.vub. ac.be/MST.html Ulrich, W. (1981), “A critique of pure cybernetic reason: the Chilean experience with cybernetics”, J. Appl. Sys. Anal., Vol. 8, p. 33. Van Gigch, J.P. (1987), Decision Making about Decision Making: Metamodels and Metasystems, Abacus, Tumbridge Well, Kent, UK. Von Glasersfeld, E. (1984), “An introduction to radical constructivism”, in Watzlawick, P. (Ed.), The Invented Reality, W.W. Norton and Company, New York, NY, pp. 17-40. Vygotsky, L.S. (1978), Mind in Society: The Development of Higher Psychological Process, Harvard University Press, Cambridge, MA. Whitehead, A.N. and Russell, B. (1910), Principia Mathematica, Cambridge University Press, Cambridge, MA. Yolles, M.I. (1996), “Critical systems thinking, paradigms, and the modelling space”, Systems Practice, Vol. 9 No. 6, pp. 549-70. Yolles, M.I. (1999), Management Systems: A Viable Approach, Financial Times Pitman, London. Yolles, M.I. (2000), “From viable systems to surfing the organisation”, Journal of Applied Systems, Vol. 1 No. 1, pp. 127-42. Yolles, M.I. (2001), “Viable boundary critique”, Journal of Operational Research Society, Vol. 51, pp. 1-12, ISSN 0160-5682/00. Yolles, M.I. and Dubois, D. (2001), “Anticipatory viable systems”, International Journal of Computing Anticipatory Systems, Vol. 9, pp. 3-20, ISSN 1373-5411. Yolles, M.I. and Guo, K. (2003), “Paradigmatic metamorphosis and organisational development”, Sys. Res., Vol. 19. Further reading Bateson, G. (1972), Steps to an Ecology of Mind, Ballantine Books. Beer, S. (1985), Diagnosing the System, Wiley, Chichester. Bruning, R.H., Schraw, G.J. and Ronning, R.R. (1999), Cognitive Psychology and Instruction, Merrill, Upper Saddle River, NJ. Piaget (1969), The Mechanisms of Perception, Basic Books, New York, NY. Staver, J.R. (1995), “Scientific research and oncoming vehicles: can radical constructivist embrace one and dodge the other?”, Journal of Research in Science Teaching, Vol. 32 No. 10, pp. 1125-8. Von Glasersfeld, E. (1998), “Why constructivism must be radical”, in Larochelle, M., Bednarz, N. and Garrison, J. (Eds), Constructivism and Education, Cambridge University Press, Cambridge, MA, pp. 23-8. Yolles, M.I. (2002), “Viable boundary critique: a reply to Bryant”, Journal of Operational Research Society, Vol. 53, pp. 1-3.

File name{KYB}Issues/33_3-4/Part 5.3d 09:51 – 1/4/04 – Total page(s) 2

PART V Selected contributions by Stafford Beer published in Kybernetes 1991-2002

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Reflections of a Cybernetician on the Practice of Planning Stafford Beer President of the World Organisation of Systems and Cybernetics

Reflections of a Cybernetician

767

Keywords Cybernetics, Management Abstract The author gives his reflections as a cybernetician on the practice of planning in a presentation directed at a professional audience of planners.

According to the mathematical theory of games, there are games of complete information whose outcomes are therefore theoretically predictable. Such a game is noughts and crosses or tic-tac-toe. There simply is no strategy that guarantees a win. Another such is chess. All the information there is lies on the table. Thus, in theory, the Black player ought to say to the White player: “You have the advantage, because you have the first move. I cannot regain the advantage, because this is a game of complete information. Therefore, I resign”. This does not happen, because even in these rigorous cicumstances someone is goind to misapprehend or make a mistake. Children do that with tic-tac-toe. In chess, please note that you have only sixteen players on your team and that these cover only six functional roles. Even so, something unexpected is going to blast you! Switch to planning. Is it not true that we try to make this game rigorous? We would like to play a game of complete information. That would be professionally sound, and of course a political delight. But there are thousands of players, and hundreds of functional roles. Not only do people misapprehend and make mistakes, but there are enemies. An enemy is one who necessarily withholds information: he is not going to tell you his hidden agenda; he defies you to discover what he is really up to. It follows that even in the ideal circumstances of a game with apparently complete information, planning suffers from intolerable constraints. The main one of these is the compartmental nature of the organisations for which planning works. Plans ramify across larger-scale systems than planners normally consider, because handling the next larger system is never their job. Planning a stadium is one thing; the effect on the city’s transportation system is another; the effect on a nearby dry cleaner whose clients can no longer park is still one more. Hospital planners consider an isolated system which is in fact a small part of health care, within which it ought to be fully integrated – but it is Based on an address to the Canadian Institute of Planners, 1987 National Conference, Toronto and published in: kybernetes, Vol. 20 No. 6, 1991, pp. 8-13, q MCB University Press, 0368-492X. Correspondence address: The Complementary Set, 34 Palmerston Square, Toronto, Ontario, Canada M6G 2S7.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 767-773 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523689

K 33,3/4

768

not. Economic plans do not work because they are isolated by departments, and geared to information systems that are perpetually out of date. The profession will I hope consider how to handle the extremely intractable problem underlying these defects which I express like this: . If the plans we create are designed to fit the existing administrative structure, they cannot properly consider the total system as an entity anew. . But if the plans we create are not designed to fit the existing administrative structure, then who will consider them, who will approve them, who will implement them? . Planning is homologous with the organisation it seeks to change. . That is the problem. So much for the planning game of ostensibly complete information. But real life is a game of incomplete information. Such a game is rock-scissors-paper. There is no secure way of predicting an opponent’s move. To model this on a chess board, you need to play a variant of chess called “kriegspiel”. In this game, each player records his own moves, but has to infer or merely guess the position of his opponent from move to move. There are three chessboards, and only the umpire has the “correct” description of the game. Now, surely, we are getting nearer to a real-life account of planning. We have departed swiftly from the public’s favourite question: “Who are the planners – those stern-faced, humourless, men and women armed to the teeth with computers?” We have moved on to our own question: “Who is the umpire?” The answer to that is City Hall, whereas the answer to that ought to be the public. But, says City Hall, we represent the public. To this the planners would do well (game-theoretically) to reply: “Oh, good. We represent a game of complete information”. Both claims refer not to a continuing reality, but to models, and both models are seriously defective. The City Hall represents the public the day after each election. The planner represents a game of complete information the day after the study is printed, bound and transmitted. Thereafter, anything may happen to either of them, and something certainly will. Consider this analogy. You have a watch that does not keep time. I have a broken watch that does not work at all. Who most often has the right time? I do, of course. You are never tight. I am exactly tight twice a day. The politician is always right when an election is called, if she has any political sense. (I say “she” because I am English — and look what just happened over there. . .). The planner is always right the day that the plan is published, if he/she has any competence. But at other times, the rightness or not cannot be recognised because there is no longer a politically demonstrable data point. It’s great to know that my watch is smack-on twice a day, but how do I know (since my watch is broken) when exactly to shout out loud: “I’m right!”?

The politician does it by polling. The planner has no such device available. Letters to the editor constitute a rather biased sample – even before the editorial policy selects them. My argument begins by saying that because of these considerations, only an idiot would actually implement a plan, if that is the product of planning. This is so for two reasons. If the plan when published is conceived out of a model that assumes complete information, it is mistaken. If it admits to using incomplete information, then obviously more information will start to be available before the ink can dry. A common way of dealing with this in business and government alike is by some form of sensitivity analysis. We may simulate the future under a variety of assumptions, and try to determine how vulnerable our policies are to plausible possible futures. My own experience is that the implausible future that we failed to consider is the future that actually happens! Another aspect of my experience, which includes involvement in some rather famous and influential simulations, is unhappily that the data are suspect. At the worst, some data have been invented. Some others have been extrapolated. But even at best, they may not relate to equivalent epochs of time. Let me give you a graphic image of this. Have you ever looked up at the night sky and realised that you are looking at a totally false image of the universe? You are seeing the moon as it was only a couple of minutes ago; you are seeing that new supernova as it was a thousand years ago. What you see is not the universe as it ever is or was, you are seeing a slice of time. So it often is with simulations. And if the different chunks of a total system are chronologically out of step with each other, then we have a planning universe that represents an illusion of something that never was nor will be. It seems to follow that the plan that is written down and set upon the shelf is not the real product of planning. The major aphorism of planning is that the product of planning is the process of planning itself, which is necessarily continuous. I am not the only person to have proclaimed this aphorism and maybe you all live by it anyway. But I offer you three consequences, which put together as a package may suggest something new. (1) Whatever is in the public mind as being “the plan” has to be adaptive. By adaptation in the planning mode we may talk about timely updating, about flexibility, about constant review and revision as events unfold. (2) If we are to maintain faith with the public, then the continuous process of adaptation ought to be open to the public, and the corrective feedback of the public should also be continuous. Something like fifteen years ago, in advocating this view to the government of Alberta, I actually designed a building for the purpose. The planning offices constituted a ring, and were interconnected across the courtyard that the ring enclosed. An outer ring consisted of the “shop windows” of each component of

Reflections of a Cybernetician

769

K 33,3/4

770

planning, in which the inner ring would display their work – not by fixed exhibits, but by constantly changing “What-if” evaluations to which the visiting public could then respond. You will realise that all these arguments point to the need to operate in real time, but the very idea is completely foreign to the bureaucratic way of doing things. Well, I believe this can and must change, so please consider my third point. (3) In the management of business and industry, folk are pleased with themselves if they can get last month’s figures out by the second Thursday of this month. It is then too late to do anything but applaud or cry, and hope to learn any lesson that’s on offer. In government and economic planning, the lag is usually about nine months. We have to employ an army of econometricians to forecast where we already are! The result that I have seen in many countries of the world is that the economy, by some weird mechanism, locks onto an imagined cycle that is exactly out of phase with the actual cycle. Therefore (in so extreme a case) all decisions are quite precisely wrong. You will correctly see in this my advocacy for real-time data input – but that not yet the point. After all, even if we get yesterday’s figures today, there is still nothing we can do about yesterday: it is all over. No, my point is that now we are suddenly able to do something about tomorrow. That is because if we can inspect yesterday’s figures for incipient instability, then we may succeed in checking that instability. Here we run headlong into the notorious problem of the political will to take any action about any such thing. The post-war “baby boom” of the 1940s caused a demographic tidal bore that has progressively overwhelmed all our institutions from nursery school onward. Quite soon we must expect to see its catastrophic effect on health and social services, hospitals, homes for the aged, and so on. This is because the average rate of illness begins to rise long before retirement age. I have heard no evidence that all this has entered the planning process, although SERF (the Socio-Economic Resource Framework pioneered in Statistics Canada and the Uiversity of Waterloo) has elaborate instrumentation on the bore’s inevitable progress. In this case the incipient instability has a probability that approaches certainty. Let’s hope that we can generate the political will to take advantage of a real-time measure of incipient instability – just as we have to hope that we can muster the political will to do anything at all that is new. Well, we shall never convert existing bureaucratic systems to this end – the upheaval would be too vast. What we can do is to model the holistic system concerned, recognise in that model the key features of its systemic vulnerability, create indices that slap measures on these features and feed them into our management system daily. Then we pass the stream of data through an elaborate statistical filter,

using mathematical probability theory, in order to detect incipient instability wherever it is evidenced. It takes a long time to comprehend fully, but it is not all that difficult to do. Many companies, services, and even some governments have implemented the process. I put it into Chile at the national level for President Allende 20 years ago [1], and between 1986 and 1988 I did the same thing for President Sanguinetti of Uruguay. Let us see what planning looks like under such a system of management in real time. My first call was for a continuously adaptive planning process; and I had already said that our best efforts at simulation are defeated by defective data – in particular those data that do not share a congruent time frame. Then imagine siphoning off the whole relevant gamut of yesterday’s data into today’s planning simulations all the time! My second call was for a planning process open continuously to public inspection, with no ancient exhibits gathering dust. Under this scheme, immediate data could be observed flowing in to modify a physical model in the planning “shop window”. Statistics Canada used to have a display in the main entrance consisting simply of a large scoreboard indicating the current number of Canadians. Of course, it kept changing as you watched it. I am afraid that this was only a simulation of a real-time system, since Tunney’s Pasture was not wired in to the nation’s maternity wards – but visitors were enthralled by its appearance of real time. And real time itself was my third call. I suggested earlier that these three points, taken as a package (which is how they were designed), might offer a new vision. Here it is. We see planning as a continuous, visible activity, operating as a kind of hinge between the inside world of the organisation, such as City Hall, and the outside world of the public interest. The whole thing is operating adaptively in a real-time flux, and the hinge keeps flexing and unflexing as internally arising information is aligned with externally arising information and feasible alternative outcomes are suggested to both. At any moment that seems politically appropriate, the executive commits resources so that the future will be different from the future that this activity expects. It is the undertaking to spend that constitutes the planning act, and not the business of preparing for it or reporting on it. In organising itself to play this role, the planning office needs to create and to be self-conscious of two models especially. Firstly, the model of the world outside that it holds has to have as much complexity as that of the interactions possible between the executive and the public. Here I am invoking the Conant-Ashby Theorem of cybernetics [2]. It means that the regulating power of the planning process is only as good as the model it holds can generate, for regulatory power cannot derive from any other source. Power itself can derive from any source – from a gun, from a crazy executive. But the regulatory power of the planning process, if it is based on a demeaned level of complexity,

Reflections of a Cybernetician

771

K 33,3/4

772

is automatically fixed at that level. This theorem, like all good theorems, sounds tautological. Yet people continually try to refute it in practice – and then wonder why their plans do not work. The other model that is needed is a model by the office of itself. There is an ever-present need to check on the identity of such activity. The more imaginative we are, the more creative, the more excited and even euphoric we become, the more necessary it is to apply a logical closure that says: “It is I who am doing these things”. Without this degree of self-reference effort becomes dissipated at best, and can soon become almost lunatic. There have been many recent examples of this in public planning in Britain, as you may know. The closure imputed by self-reference is a major preoccupation of contemporary cybernetics. In the world of organisations, as of psychiatry, it is this healthy state that avoids the contradictory outlooks of schizophrenia. For example, is Toronto in the business of preserving its comfortable old identity as a (rather large) village, or is it soon to become one of the world’s major cities, and a leader in some kind of excellence? Moreover, if the latter, which kind of excellence do we have in mind? Maybe you could achieve any of the possible alternatives but you cannot achieve them all. These are genuine dilemmas that I have noted here. If you wish a business example: Is a mutual life company a system for selling insurance with high panache, whereupon you will have lots of premiums and had better do something about investing the money; or is it a system for making money on the international stock market, whereupon you will need huge funds, and might as well sell some insurance to acquire them? I once met this as a genuine dilemma here too [3]. Finally, as I mentioned before, there is the over-riding issue of political will. Our kind of democracy is exceptionally short-term, and that is inimical to planning. Governments do not last long, appointments to office last perhaps a matter of months, and always the media shorten the time-span of decisions just as they shorten the time-span of public attention. Next, our society is extremely materialistic, and is unlikely to accept planning suggestions that conflict with selfish aims. For instance, there is an opinion among social scientists that a big mistake has been made in segregating the elderly. The idea of bringing them back into the family through newly designed housing is mooted and could begin with new designs and conversions of existing property to provide “granny flats”. But I heard one Minister of Housing say that people were already opposing this step because it might reduce property values. So, of course, would the provision of many social amenities in “posh” districts ranging from abattoirs in rural areas to homes for AIDS victims in the cities. As I said, we confront the issues of political will that planners cannot be expected to affect. No wonder that planning comes to be seen as a permanent

and indistinguishable feature of the bureaucratic background. . . But are we right in eschewing responsibility? It is a matter worth debate. Let us at least jerk planning into real time and public prominence! Impossible? No, it isn’t. That flat allegation leads me to share with you my own secret as to planning and the future. Never forecast anything until it has already happened. You will have 100 per cent success and no one will spot the trick – because, in this “information age”, there is such a huge lag between accomplishment and public awareness. For example, I first implemented this real-time design, with its mathematical measures of the importance of instability, and its inferences for planning, without benefit of computers, 40 years ago. I am still doing it. Notes 1. The story of the work for President Allende in Chile is recorded in the last five chapters of the author’s Brain of the Firm, 2nd edition, John Wiley & Sons, 1981. 2. The imporant Conant-Ashby Theory was first published in International Journal of Systems Science, Vol. 1 No. 2, 1970. It is now available in Ashby’s collected works, Mechanisms of Intelligence, edited by Roger Conant, Intersystems Publications, Seaside, California 1981. In both cases the title is: Every Good Regulator of a System must be a Model of that System”. 3. The dilemma faced by the mutual life insurance company is fully discussed in the author’s Chapter 9, “The Evolution of a Management Cybernetics Process” in The Viable System Model: Interpretations and Applications of Stafford Beer’s VSM, edited by Espejo and Han den, John Wiley & Sons, 1989.

Reflections of a Cybernetician

773

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

774

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

World in Torment: A Time Whose Idea Must Come Stafford Beer Toronto, Ontario, Canada Keywords Change management, Cybernetics, Team working Abstract Presents the full text of the Presidential Address by Stafford Beer to the Triennial Congress of the World Organization of Systems and Cybernetics, New Delhi, India, January 1993. Introduces the components of contemporary change and discusses the diagnostic approach of management cybernetics. Outlines a summary theory of autonomy and considers autonomy at the global level. Offers an account of the cybernetics of chronic societary triage, developing an analysis of triage through category A, B and C partition. Produces a summary theory of team syntegrity and discusses the power and the use of the team syntegrity model. Finally, outlines an action plan for World Syntegration.

You will remember the beginning of humankind. Our first parents were quick to get themselves into trouble. They were expelled from the Garden of Eden. I understand that Adam took Eve’s hand, and said: “My dear, we are living in a time of transition.” Perhaps people have always felt like that. We certainly do today. Have you ever tried to list the components of contemporary change? It is easy enough to cite the marvels of modern science and technology – how the computer, and television, and medical science have changed our lives. If you start with such matters, it becomes a “profound insight” to observe that there has been a change in the rate of change. But that was obvious 20 to 30 years ago, for I was writing books about it then.

Kybernetes Vol. 33 No. 3/4, 2004 pp. 774-803 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523698

Components of Contemporary Change Today my list is different. At the top is the spectacular advance in human misery. I estimate that more human beings are enduring agony today than ever before; the number could be greater than the sum of sufferers throughout history. I speak of starvation and epidemic; war and terrorism; deprivation, exploitation, and physical torture. I repeat the word agony; I am not talking about “hard times”. Second on my list is the collapse of the civilization which we have known in our lifetime. We are looking at the rubble which remains of two competing empires. Soviet communism has accepted its own demise; Western capitalism has not accepted it yet. But I am not making a forecast. I am examining the facts which are under our noses. This is the full text of the Presidential Address to the Triennial Congress of the World Organization of Systems and Cybernetics, New Delhi, India, January 1993. First published in: Kybernetes, Vol. 22 No. 6, 1993, pp. 15-43. q MCB University Press, 0368-492X

Out of “political correctness” no one talks about the exploitation of either nature or indigenous peoples any more. They talk instead about “sustainable development” – but there is no such thing. Not only can development not be sustained; even the existing fabric cannot be sustained any longer. These two spectacular transitions, of human agony and societary collapse, are connected – not only at the phenomenal level, but also in their etiology. It is not credible that most people prefer to live under these twin conditions. It follows that we are governed by an oligarchy – by the few; it is an oligarchy of power, greed, and terror. In the most extraordinary way, we are blind to this. To take the major example: none of the phenomena which I have mentioned would be observed in its current and virulent form, if there were no powerful modern armaments. There are always pacifists around, thank God. But many friends tell me that pacificism cannot work in practice. Do not say this to me, who heard Gandhiji speak in the flesh. It is not for nothing that the Sanskrit word ahimsa is so negatively translated as “non-violence”: satyagraha as its complement is not understood in the West at all. It means “holding to the truth”. Ahimsa does not entail any lack of involvement or strong engagement. But no serious political platform anywhere has proposed to make the manufacture of armaments illegal. To the contrary, this manufacture is essential to the conduct of the existing world economy, and is the major instrument of vicarious foreign policy by those who command it. The Diagnostic Approach of Managerial Cybernetics What are we to say about the management which procreates this disastrous mess? Without jumping to conspiracy theories, or citing the illegal activities which now constitute the world’s biggest industry; we can at the least say that humankind now manages its own affairs with breathtaking incompetence. This was not always so. Small tribes managed themselves very well indeed, and without destroying their habitats. Something has been going on which seems, at least in part, to be a function of size. Why should size make a difference? Look at it this way. The number of internal relationships inside a complex system grows exponentially with linear growth in the system. And, thanks to increasing complexity in the relationships detectable between the systemic elements themselves, induced by higher technology, we have been witnessing a variety explosion where the exponential function is itself an exponent. After centuries of riding on horseback, we have achieved speeds of 28,000 miles per hour: escape velocity. Each of us knows the dramatic change in the rate of change in computing – from the abacus to the chip. Overridingly, the population explosion seems to be hyperbolic, never mind exponential. The variety, the measure of complexity; of the system which we need to manage is a new universe of galaxies, compared with the single solar system

World in Torment

775

K 33,3/4

776

which we had to manage when the industrial revolution began. It has all happened within 200 years. And surely that revolution was, in cybernetic terms, the coenetic variable to which we may trace systemic change in technology, economics, and the social order alike over the period. What can the cybernetician, having recognized the coenetic variable, say about the management of this explosive transition, which is more than the sum of the relatively isolated analyses of technological, economic, and societary change management? First of all, the brain has not changed in this time. It remains, as McCulloch found it, a three-pound electro-chemical computer, running on glucose at 25 watts. Even so, it had a very large number of elements: ten billion neurons, forsooth. It sounded a lot at the time, that is, in the 1950s. But now? Why, that is only ten gigabites. Computers, if not brains, can handle that. But here is the rub. In programming a computer, one needs a model. Models are provided by brains. Models are necessarily massive variety attenuators, because they select only those aspects of the world which are relevant to the model’s purpose. Worse still, the models adopted are not the best which we can provide: they are consensual models put in place and held together by ideologies. And an ideology is a very low variety instrument indeed. Vast tracts of political philosophy since the ancient Greeks have been studied in common by the theorists of both communism and capitalism; but the ideologies to which the two superpowers rallied their supporters attenuated this variety in different guises. They have had this much in common: neither had Requisite Variety (as defined by Ashby’s Law) by which to manage. Both are managerially dysfunctional therefore. And neither works. To the political analyst, the two management systems are quite different, and to the politician wholly opposed. Neither side has had the least compunction in imputing moral judgements, and the rank and file have loved it. They have gone to both hot and cold wars about it. To the cybernetician, oddly enough, and starting from that common ancestor, the coenetic variable, much the same process has been going on. In a word, it is dysfunctional over-centrality. A Summary Theory of Autonomy According to the Law of Inter-recursive Cohesion, which I propounded in the 1970s, it is a prerequisite of viability that a system should develop maximum autonomy in its parts, where maximum is defined to mean “short of threatening the integrity of the whole”. Since you may not have encountered this work (more fully treated in [1] ), let me depict any viable system as a cohesive whole having distinguishing parts like this: What is this whole? I am talking about a given societary unit: it could be a village or a company, a state or a multinational, a nation or the whole of humankind. What is systemic identity for this unit? It is a highly complex ethos

of self-recognition, deriving from the interplay of many components – and in particular of the observed system with the multiplicity of its environments, which, indeed, are major forces in formulating that ethos and forging that identity. These complications are discussed in the quoted reference; but we all know what we mean by acknowledging our nationality, for example. It has connotations of which we may be proud, and connotations of which we may be ashamed; it has connotations in other places which are gross distortions of the reality which we recognize as home. What I say next assumes that we understand something about “identity” – in all its complexification. The process or activity which identifies each part of this identity is a circle marked with the measure of its variety, namely V. The management, depicted by the square box which impinges on the process, clearly has lower variety than the process itself. Management, after all, has to amplify its states in order to accommodate the process variety which proliferates around it. Similarly, the variety of the environment, in which the process is embedded, must be attenuated by some means, if the process is not to be engulfed by the proliferating variety of its external world. In any given case, the situation is far more complicated, because we are dealing with continuous loops rather than simple connectivities, and I have published detailed analyses [2] at length. But the essence of the Cohesion Law comes down to this. First of all we have: for each part of a viable system, where the differential sizes of V are intended as quantifiers. Then, in order to obey Ashby’s Law of Requisite Variety; regulators called amplification and attenuation must be in place (I use the standard symbols from electrical engineering), so as to procure variety equivalence, like this: so that the three Vs are of equivalent, and therefore requisite, variety. I have put emphasis on the central process and direct regulation (straight lines) for simplicity. In fact (as indicated by the curved lines) we are dealing, as I mentioned, with a pair of homeostatic loops which produce their effect by mixed regulative strategies. Note that I said that the regulators “must be in place”. I say “must”, because we speak of a law of nature, whereby variety ( just like water seeking its own level) tends to equate. Of course, we may not like the way in which this happens to occur, but we cannot prevent it. Managers may autocratically use threats or even guns as amplifiers of their own variety. And the most effective attenuator of environmental variety is often sheer ignorance within this subsystem of how the environment actually works. Management which is effective and ethical, however, will design the regulators, and put them in place. I have written much about how to do this [2, 3], but I shall now return to the Law of Inter-recursive Cohesion and the deployment of this elementary model. The task of managing the part, the horizontal component of Figure 1, is difficult because the managers do not start out with Requisite Variety. They need ingenuity and skills to put their regulators optimally in place. It follows that any intervention from above is likely to inhibit their professional practice.

World in Torment

777

K 33,3/4

778

Figure 1.

Even when intervention is deliberately designed to facilitate their on-site jobs, the policy or overall plan which intervenes cannot in principle have Requisite Variety in turn. The argument is the same as in Figures 2 and 3, turned through 90 degrees. Then why not leave the subsystemic parts to be self-organizing viable systems in their own right? Well, if the original whole is to maintain its coherent identity, then the relations between the processes which the parts embody need to be continuously reaffirmed, and that is the function of the squiggly vertical links depicted in Figure 1. Autonomy turns out to mean the maximum discretionary action for the part, short of threatening the integrity of the whole. This is a non-emotive definition of a very emotive term: freedom. I have been in trouble before for defining freedom as a computable function of the systemic purpose.

Figure 2.

Figure 3.

Please note the connexions here. If the purpose of the system changes, then its identity is different. If its identity is different; then the relationships defining its connected parts will be different. There is nothing in that so far which is not within the bounds of quiet evolutionary change. Indeed, I may claim to have given a cybernetic account of biological adaptation – which seems to apply to societies and species alike. It is, in my opinion, a start on the general process which characterizes viable behaviour – a process slant on the general process identified by Humberto Maturana as autopoesis [4]. But now I return to our current state of collapse. In the Soviet Union, the belief in central planning was quite clearly a repudiation of the Law of Inter-recursive Cohesion. The higher level of recursion, which was the USSR itself with all its managerial and party appurtenances, undertook massive intervention in the autonomy of the parts. According to my analysis, even as so briefly given here, this was necessarily dysfunctional to viability. It is not relevant to decide whether this was well-intentioned intervention, nor whether its enforcement was ethically conducted. What is relevant is that the human, economic and social frustration of proper autonomy was bound to blow the system apart. This is not said with mere hindsight. I argued this case with Eastern bloc scientists for many years. And, when I had the chance to help President Allende in Chile, 20 years ago, I designed a viable system which was completely autonomous, as here defined. The British Collapse The collapse of Soviet comunism was formal, and it was heralded by the West in a spirit of vainglorious triumphalism. They were wrong; ergo we are right. But, to say this, one must blind oneself to the facts. As in the East, dysfunctional over-centrality has underwritten disaster; diagnosis uncovers the same cybernetic illness; only in the West the symptoms are different. And, as often happens with mortal disease, people refuse to look at those disquieting symptoms. I am British, and through the decade of the 1980s I watched my country destroyed by an ideology. Investment in wealth-producing industry at home was denied in favour of international speculation, and the industrial base all but disappeared. Savings were eaten into by inflation, and by market manipulations which forced the price of housing, for instance, out of reach. Inner cities and urban infrastructure decayed; the army of the homeless was joined by large numbers of mentally deranged, as the health service was contracted; schools and universities were profoundly damaged by cuts. National assets which were publicly owned were sold off at bargain prices, and are steadily deteriorating. Income was redistributed, with huge increases going to the rich at the expense of the poor. Indeed, at the point when the proportion of people existing below the poverty line exceeded 30 per cent, the Government abolished the publication of that statistical index. Unemployment steadily

World in Torment

779

K 33,3/4

780

increased, despite more than 20 changes in the method of calculation to make the situation look better. So much for symptoms. This is no more than a thumbnail sketch. How have these symptoms been concealed? Well, of course, we have taken bitter medicine, and things will soon get better. Never mind the cover-up of a spuriously prosperous and an absurdly over-valued pound which took too long to collapse in its turn. Immense sums were spent on international public relations in order, fairly successfully, to have a fake boom labelled an economic miracle. The creed of greed masqueraded under the respectable title of “traditional values”, but we are left with the rubble. We are now ready to recognize the cybernetic process which impelled this disaster. The first variety attenuation which ran quite counter to British viability was the creed of greed itself. Never before had the great majority of Britons been persuaded that the only criterion which matters is wealth, or that money alone is its measure. Thus Philistinism, disregard of learning, contempt for the weak and, above all, abandonment of the concept of the social good came to be the marks of the decade. In proselytizing egoism, the Prime Minister actually declared: “There is no such thing as society”. This variety attenuation of values enabled a low variety managerial model to seem adequate. Monetarism, and its attempt to maintain stability through manipulating the money supply and interest rates, does not have Requisite Variety to maintain viability. The main reason for this is that the nation in all its locations and in all its facets generates far more variety than such an exiguous regulatory concept can encompass. The nation resists the attempt to define its identity simply in terms of gross national product, particularly when this is underwritten by “funny money” generated between international exchange operators, and especially when what remains of value is inequitably distributed. A second reason for the failure of the Government to control the money supply by not printing money to spend in Keynesian investment, was that they embossed money on plastic instead. The promotion of greed was backed by the facilitation of credit, and it was this massive increase in indebtedness which promoted the fake boom. Lacking Requisite Variety, the Government’s policies did not work. In cybernetic terms, there are two possible ways of restoring Requisite Variety. One is to enrich the identity of the nation by restoring to it the purposes which it used to embrace. To do this, as we saw, the subsystemic relationships between aspects of society must be reinforced and desiderata other than money re-established. But this would contradict the ruling ideological paradigm. It follows that proliferating variety must be restrained, which can be done in terms of this analysis by decreasing autonomy. And that is exactly what happened. The educational authorities and the health authorities which hold decentralized federal power in counties, municipalities and even villages,

were an intrinsic part of the British way of life. They had Requisite Variety in local knowledge and insight; and their existence encouraged care and concern and voluntary action. They were virtually abolished, because they were divested of power; policy making and planning were centralized in London. A national curriculum has been introduced, and school inspection privatized. Small hospitals have been closed as inefficient and London determines most aspects of health care delivery, including dentistry: This wholesale bureaucratization conflicts with the ideology of individual enterprise. The ingenious if disingenuous way round this has been to urge medical practices and schools to opt out of the social framework which has characterized Britain, and commit themselves (still under central direction) to the profit-oriented definition of “care”. But it is local government itself which is the ultimate victim. Local elected bodies, again with Requisite Variety, and under democratic sanction, raised local taxes to implement local policies to satisfy local needs. But they spent more money than the central power would permit. They have been emasculated. It is illegal for them to spend more than central authority decides. Again, this critique does not derive from hindsight. I launched it 20 years ago [5] and have often returned to the attack, [6] for example, since. It is sad to watch the ideologies of the West refusing to accept the evidence of the rubble which was their culture; sadder still is the willingness of the East to sit at the feet of proven failure, to learn how to make the same mistakes – just as if there had been a success. It is strange that two ideologies which have been billed as utterly opposed should both come to ruin from the same systemic disease: the loss of autonomy. Yet its inverse form, the seizure of power, is common enough in the history of humankind. Where does this always and inevitably lead? In the absence of a cybernetically sound structure for viability, what basic machinery underlies its unviable alternative? I want to propose one form, which I call chronic triage.

Autonomy at the Global Level The centralization of power in Britain has proved dysfunctional, just as it did in the Soviet Union. The difference is this. The strong sense of purpose in the dominant Communist Party generated a strong sense of identity in the Soviet Union. This was preserved by the loss of subsystemic autonomy, to the point where the whole system collapsed in an explosion which could not be denied. In Britain the strong purpose of the ideological oligarchy set out to change the existing sense of identity, and did so by juggling with autonomy so as to alter subsystemic relationships. The result was the same: the loss of freedom involved in diminished autonomy, and the ruin involved in dysfunctional over-centrality. What is different is that Britain is still denying the truth, and can get away with it by

World in Torment

781

K 33,3/4

782

claiming the possibility of recovery from a national indebtedness which it denies to its own parts as “a matter of principle”. I hardly dare mention the national indebtedness of the USA, which has the most cybernetically flawed organization of the Western nations (although this is not quite its self-image). This exploding debt cannot in the long run be contained. In the attempt to service it, social penalties are being incurred which threaten the future of the American dream. It would be entertaining, if it were not tragic, that, when policies do not work, the oligarchy does not suppose that they are wrong policies, but decides that they have not been pursued with sufficient vigour. Britain has suffered greatly already from this mistakenly positive feedback, but proceeds to propel itself into a united Europe on the same lethal premisses. We are already witnessing dysfunctional over-centralization in Brussels, and it will get much worse. I go on record as a convinced European, a convinced World Citizen, indeed. But nothing can be achieved by organizations which are cybernetically flawed. The ecumenical tendency to replicate these flawed structures and ruinous policies is most alarming. Power, as we saw, becomes concentrated, and the initiative is lost to local hands. Moreover, the time cycle of development is artificially shortened, and the market distorted to recognize only those needs which current technology determines to be economic. When we build at minimum cost per square foot, for example, we get boxes which are cheap – but may well be dysfunctional. There are practical matters of effective operation, of convenience, even of aesthetics which are often ignored in the economic equation. The biggest casualty of all under the creed of greed is the social good. But Western Society obeys the economic diktat, which is uttered from the centre – from the central bank to the World Bank; from local budget provision to the International Monetary Fund. Which brings me to the United Nations, which ought to embody that social good for the sake of all humanity. Here we have an assembly of the whole world’s nation states, each of which is in some sort of disarray. Yugoslavia has blown apart, having lost its Tito-empowered identity, and rediscovered its participant autonomy. As this process continues, whether in the strong form of the Soviet Republics, or the weak form of disunity which causes difficulties already in Britain and even Canada, the assembly of nations called the UN grows larger. If Africa were to disintegrate, so that natural tribal autonomies emerged as nations, replacing the artefacts of colonial invention, perhaps the roll-call at the UN would quadruple overnight. . . Now, if we apply even the summary theory of autonomy on which this article is based to the United Nations, we find the reverse situation from that so far discussed. The autonomy claimed by the nation states overwhelms in sum the variety displayed by the UN as an entity – if indeed it is an entity at all.

It was breaking new ground when the Security Council managed to muster its variety together in the attempt to change the course of Iraq. The results were most disheartening to anyone who cares more about the agony of human beings than about the price of oil. I used the word triumphalism before; surely it is shocking that anyone should brag about the handful of lives lost in an operation which killed directly and indirectly a quarter of a million people. The Cybernetics of Chronic Societary Triage We manage through a model which we hold in our heads about how things work “out there”. If our model does not have Requisite Variety, then we ought to incorporate learning circuits which will enrich it. But, if we are ideologically attached to our model, so that it is not negotiable, and if we reject all contrary evidence, then it becomes a dysfunctional paradigm. Any oligarchy which has the power to enforce its ideology in this way will quite predictably give rise to triage. Let me explain this term, and offer a cybernetic account of it. Triage comes from the French verb “trier”, to sort. In the early eighteenth century, triage meant the sorting of wool into various degrees of quality of the fleece. In the early nineteenth century, triage was the process of sorting coffee-beans into categories of the best, the satisfactory: and the worst – bad or broken beans. And the term “triage coffee” referred to this worst category. A pejorative connotation had entered the usage; and the triage process had come to mean a sorting into three categories. Please note that the “tri” in triage does not etymologically mean “three”: the threefold categorization was a mere convenience. Today, triage has an enlarged connotation. In a disaster, priority in treating casualties is given to those with the best chance of survival, if given treatment. That is bad luck for the most afflicted; and contrary, indeed, to humanitarian instinct. But the hard decision, given limited resources, may have to be taken – these people are probably going to die anyway. Now I propose to extend this modern use of the word triage from discrete disasters (an acute condition) to a continuing state of affairs (a chronic condition). My argument will be to show why an ideological oligarchy in government tends to sort the people into categories, and with what expected results. Again I use three categories, just for convenience. The measures of variety are of course arbitrary too: no empirical estimates exist, and I am computing in terms of relative complexity. In the following sketch, which is an approach towards a Model of Chronic Societary Triage, the basic conventions of the autonomy model are maintained. The letter V stands for a constant representing the (low) variety measure of the controlling ideology. x, y and z are modifiers which reflect the way in which societary categories A, B and C respectively represent the variety of the controlling ideology v(i). The integer coefficients chosen, having no empirical basis, are, strictly speaking, ordinal numbers. Note, however, that they follow

World in Torment

783

K 33,3/4

784

the basic Fibonacci series (1, 1, 2, 3, 5, 8, 13, 21, 34, 55). This is the common growth progression in biological systems; it typifies the expectation that additive growth leads to geometric development. Thus each number n in the series used is defined by n ¼ ðn 2 1Þ þ ðn 2 2Þ: Ordinal numbers unweighted (1, 2, 3, 4 . . .) offer no sense of the increasing stress in the subsystems as disparity grows, whereas there is abundant evidence that social systems are under such stress – beginning ( perhaps crucially) with the demographic stress of the population explosion itself, and continuing into epidemiological stress – whether physical (as with AIDS) or cultural (as with drugs). Suffice it to say that, in the absence of empirical measures yielding cardinal numbers, ordinal numbers must be used instead, and that there is a basic scientific reason for “intensifying” the series. It will not affect the argument if someone wishes to replace Fibonacci with a pet series of his/her own, so long as it recognizes biological and sociological development facts. Consider the Category A partition. This comprises, by definition, a management group which shares the ideological paradigm i, of variety v, which itself (and therefore) has variety equal to v, disposed on the horizontal axis by the Category A modifier x. We make the usual claim that the variety of the activity regulated must be higher than that which the management itself has available, call it 2v(x). The variety of the environment in turn must be higher still, and we call it 3v(x). Homeostatic loops are in place which amplify the lower varieties to absorb the higher varieties, and attenuate the higher varieties towards the lower varieties. Their “mixed strategies” facilitate the Law of Requisite Variety, and induce transformations (marked T ), which have the effect of reducing variety in all three blocks of the diagram to the basic variety v(x). The equilibriating process, indicated by the two meters for measuring approximate equality, feed into a comparator measuring error e, which generates a feedback function f(e) to the management box. Note that the transformation on which it impinges generates an identity transform, since Category A was defined from the start as matching the controlling variety of the ideological paradigm. The effect of the feedback, then, is constantly to reinforce the identity v(x), stabilizing the complexity which the ideology recognizes as the quantity v, and reaffirming the qualitative x-ness of A. This should have the outcome that the management box learns ever more effective ways of managing the homeostasis. This means in practice the continuous adaptation of design in the amplifiers and attenuators at P and Q, thus completing the learning loop. Note finally in Figure 4 that the vertical connections between the overall culture and the partition by definition exhibit Requisite Variety namely v. This vertical homeostat therefore operates without adjustors; and that very fact constantly reaffirms the shared ideology which defines Category A.

World in Torment

785

Figure 4. Analysis of Triage: Category A Partition

Category B does not share the ideology which it must nevertheless take into account – because the ideology controls the general culture. This necessarily places the horizontal system under stress. All components are struggling with the compromise between the official ideology and the pragmatic policies which they must pursue to remain B-viable. This is a fascinating process to watch – once one recognizes, through Requisite Variety, what to look for. B-people are typically uneasy about the compromises which they make, and can be observed to wriggle under this duress. Many people in the West send their children to private schools because they confer privilege, despite disbelieving in the class and value systems which support them. Third World politicians accept the rules of the international economic game, although many suspect and will say privately that the entailed destruction of their indigenous culture is a tragedy. I agree with them, and am shocked at the willingness of the West to presume to teach low-variety ideological models which have already and demonstrably failed. With whatever mental reservations, however, B-people and politicians tend to go through the motions of compliance with the dominant culture paradigm. Unhappily, when Category A status is fully secured, many forget the mental reservations which they made. They become genuine conformists. Then they begin to revel in the pay-off for inhibiting their personal or cultural variety. This is called corruption. The Fibonacci coefficients 5, 8, 13 are used to make clear the consequences of not sharing, and to indicate that the complexity of the whole horizontal process is greatly increased. Because of the pragmatic compromise, the management box has (as it were) to “run two sets of books”. One is accountable to its

K 33,3/4

786

Figure 5. Analysis of Triage: Category B Partition

masters, and the other to its constituents. These constituents, in the process circle and in the environment, are subject to strong pressures by the management of the horizontal regulators of the two homeostats, since they must use more energy than Category A regulators to restore Requisite Variety with ideological conformity. The learning process generated through the feedback function f(e) is correspondingly more severe, so that B-managers themselves will feel oppressed. The overall “controllers”, the cultural Establishment, characteristically make themselves alert to this. Wisely so, because the oppression nurtures the seeds of revolt. Thus the model posits a comparator which judges the continuous error e between the identity transform which stabilizes ideological conformity for the culture itself, and the outcome of B-managements’s efforts to match it in Requisite Variety. The learning loop this time is closed through adaptation of the vertical regulators via the moderator R. Note that any deviations on the part of the B-partition result in louder, more stern commands from above, coupled with a decreased inclination to pay attention to the “noise” of complaint from below (Figure 5). It is hardly necessary to draw a new figure to illustrate the plight of Category C, because it is structurally identical with the last figure (but see Figure 6, which includes it). The differences lie in the variety mismatches involved. The Fibonacci coefficients are now 21, 34, 55 – and it is this massive exacerbation of ideological inequality which puts Category C into the disaster mode. It is the triage subculture; and, because of the oppressive regulatory apparatus needed to stabilize the system at this level, disaffection and disjunction are inevitable. Moreover, this is not an acute, momentary crisis. It is

World in Torment

787

Figure 6. Graphic Model of Chronic Societary Triage

chronic dysfunction determined by the societary structure as defined. That is why it is idle to complain that the model has invented its own problems by choosing Fibonacci coefficients with explosive results. The loss of Requisite Variety is implicit anyway. Allow me to repeat, because the point is commonly misrepresented: the Fibonacci series chosen to illustrate an observed behaviour of societary systems is not of critical importance. No one challenges the fact that large numbers of people constitute a badly disadvantaged section of society (we trip over them in the streets of London and Washington, and they starve in Somalia). No one denies the alienation of this section of society (the crime rate rises in Manchester and New York, and there have been rebellions against dictatorship all over the world at all times). Second, if the variety of the

K 33,3/4

788

dominant paradigm is v, no one can deny that the variety at C is very much larger, because the excess variety takes so much containing – larger and more powerful police forces are only the start. But force, oppression, tends to be the dominant mode of containment, for the quite straightforward reason that the dominant ideology, manifestly – ostentatiously – lacks the legitimacy of consensual government. Are scientists to ignore the validated predictions of conceptual scientific models because the data are so far imprecise? If I say that a dead man weighed barely 100 lbs, and a scientist weighs the body and its weight is 102 lbs, the man is none the less dead, and the “mere theory” that he had too little to eat is not altogether discredited. Some members of the scientific confraternity prefer to complain about illustrations such as Fibonacci, and “loose” measures of a variety which is clearly explosive, than to see whether cybernetic structures have something to tell us. The next and final figure is given to facilitate discussion of the whole model. The reader is asked to supply the large portions of the schematic which s/he knows to be missing: those details are unchanged. Consider in more depth the role of the modifiers x, y and z, which were said to reflect the way in which the categories respectively represent the variety of the controlling ideology. Since variety is simply a measure of complexity (the number of possible states), it may not be clear why the systematic explosion of variety through the model is alleged to have a qualitative political importance. The reason is that the low-variety model is simplistic in the eyes of all those concerned in Categories B and C. It fails to accommodate or even to register the complications which they face in ordinary life. In a nutshell: folks whose children are starving do not accept that stealing food is wicked, although they know it is illegal. Then the enforcement of the ideology denies the opportunity to question it. The less dramatic version of this argument says that the dominant paradigm does not recognize such a thing as “social good” at level C. People’s lives are grossly affected by cutting bus services, manipulating apparently trivial prices, and so on, in ways unimagined at level A – all things done under the “more economic” excuse. In the limit, which means in Category C, the “modifier” z will fail to modify. C-folk will have to work within a low variety model imposed from above which they do not see as at all relevant: that is to say they become alienated from society at large. A gap opens between B and C which is soon a chasm. Upward mobility for C-folk becomes impossible, even if it were desired; and the C-environment becomes a sink into which B-failures descend. The cosmetic treatment of this socially disastrous gap is painful to observe. The pretence that the gap can be bridged to any important degree is fostered through high-profile “examples” of such success: think of the arts ( pop-stars, say) or sports (transfer fees, say). For most people, there is no hope. The more C-folk are alienated, the more alienated they will become. This is because the vertical learning loop, based on error-controlled feedback

generated at comparator E, uses its instrumentality R to put higher gain on the amplifiers and higher resistance into the attenuators (indicated by the doubling of symbols in the Figure). This represents a strong positive feedback. The unification of society would call for negative feedback at this point, but the oligarchy calls that weakness: a loss of standards or a failure of resolve. It is a matter of observation that the undesirable positive reinforcement of alienation actually occurs. The disadvantaged often reject the help to which they are entitled – for instance, scholarships may be refused or welfare payments go uncollected, because the peer group pressure to remain “one of us” (victims) is strong, and the inertia of the bureaucratic process is too great for disillusioned people to overcome. Then turn to the environments in which the processes of the three categories subsist. In so far as the C-people are totally alienated, there is no interplay between their world and the remainder. There is only remote and perhaps envious observation, enhanced and distorted by the media (situation comedies, game shows, and so forth). Whereas there is an intersection (hatched) between the worlds of A and B in which membership is ambiguous; and, as the vertical arrows are meant to suggest, there is migration between them. People in the B-Category are not so far removed in variety from the dominant ideology that they have no hope of assimilating their modus vivendi into the lower-variety strait-jacket of Category A. It pays them to try, because social benefits always accrue to those who work within the paradigm which designs and underwrites them. Backsliders in A, on the other hand, who increase variety by increasing behavioural options, lapse from the rigour of the shared ideology of A in so doing. Unfortunately, Category C frequently pays penalties (for failing to uphold the ideology) to Category A. Category B also pays into A (to help it to uphold the ideology). The B-people have a repugnance from C-people, and all the societary pressures exhibited by the model press for their upward mobility: For Bs the fear of falling through the gap into C is threatening. They avert their eyes, therefore: serious confrontation of C-topics tends to be taboo. The Bs’ sense of security and entitlement is often tenuous – which leads to self-censorship. They are likely to aspire to A-hood, in any case; but Category A resists the pressure in order to preserve privilege. Remember that the oligarchy is itself operating within Category A. Reflecting on the factors listed in this paragraph in the context of the model suggests sizes of A, B and C will tend to equate. The measure cannot be so crude as income per capita, nor the number of individuals included. It seems likely to be a weighted social statistical index of some kind, depending on the society under study. The point is that there will be a partition equivalence in societary terms which reflects the Requisite Variety of the cybernetic terms. A final comment says that the triage constantly reinforces its own structure, and rapidly becomes indestructible except by main force. That is because, as

World in Torment

789

K 33,3/4

790

I have many times shown [3,7], high-gain systems such as this have large negative real parts, and are soon dominated by their own feedback functions rather than by their inputs. In this case, after all, the feedback loops have been described as societary learning loops – which teach the system how to be itself and nothing else. Its structure is self-referential. Making changes in the inputs, therefore, will have little effect. And this goes a long way towards an understanding as to why social change is so difficult, and especially why liberal measures to alleviate suffering usually and typically fail. The Bottom Line is Not the Bottom Line The exposition of the triage model was introduced by reference to the internal worlds of the empires which we used to call the superpowers, the USA and the USSR. In a worldwide context, it is easy enough to recognize Categories A and B in these two; and, if so, there is little difficulty in placing the so-called Third World into Category C. There is no room here to analyse this classification in detail, but I invite you to do so. Note, in particular, how the attempt to give economic aid from the rich world to the poor world has actually increased the “triage gap”, in that there is a net outflow of wealth from the poor countries to the rich. Note that most B-folk are not aware of this: the averted eyes syndrome, perhaps; even so most people (especially those with personal computers) do not realize that they are impoverished of information, because they are flooded with data. Note also how the attempt to impose the greed ideology through so-called “development” leads to deficit and disaster – just as the model predicts. The World Bank and the IMF share the ideology, and the low variety models behind it. Within an individual country, the triage is usually all too evident. In India, for example, the triage gap remains startling. Of course, it is actually institutionalized: the core group of the C-category is the harijan class. I was in the Indian army in the mid-1940s as a young man. I left India on 23 March 1947 – a significant date, for another Empire, the Raj, had just shut down. With joy, I truly believed that the harijan class would soon disappear. It was not until I elaborated the triage model 45 years later, that is in 1992, that I understood my disappointment. There has been some social progress, I know, but it is very slow; and the structural triage is unaffected. Meanwhile, in the Britain of the 1980s, I watched the triage machine working with great effect. The ruling ideology was very strong, and of very low variety. Category B enjoyed a fake boom, and money was systematically siphoned out of Category C into Category A. When the percentage of people living below the poverty line passed 30 per cent; the Government at last acted. They abolished the statistical index measuring the poverty line itself. Much the same thing is going on in the USA, and with similar impact on the poor, and on health, education and welfare generally. It has become an accepted truth that capitalist management is efficient, compared with the bureaucratic inefficiency

associated with state enterprise. But anyone who has worked in the high echelons of big business knows this “truth” to be false. She/he has experienced the “Waste and Folly in the Private Sector” which subtitles a devastating indictment [8] by Herschel Hardin. This book should be required reading in every business school. We are dealing with a triage pump. The As become steadily more advantaged, and the Cs more disadvantaged; and the trend continues. If my instinct about thirds is even roughly right, then we can see why we have the trend, and why disastrous governments are re-elected. The As retain the Government which favours them, and the Cs are powerless. The Bs hold the power balance; and they are bribed with consumer goods financed by ever-mounting credit – plus the hope of leveraging themselves into Category A. The majority in a triage democracy is likely to opt for the status quo. An interesting contemporary phenomenon is that change is often embraced, because it is seen to be desperately required, on the tacit understanding that it will not involve any actual alteration. Governments tend to oscillate around a central political stance, so that “left” and “right” positions around that mean are almost indistinguishable. Tomorrow the new President Clinton will be inaugurated in the USA. He was elected on a platform demanding change. It remains to be seen whether anyone, the urban Black population for instance, will detect any difference. Since there has been no suggestion of restructuring anything, as distinct from massaging soreness, the model expects that the triage effect will leave conditions much as they are. The theory advanced has isolated two features of the apparently triumphant culture which now dominates the world, which I allege will prove fatal to coherent civilization. The first is the triage machine. In brief, this divides the world into categories (more than three may yet be identified) which condemn the lowest category nations to subjection by a controlling ideology which is now ecumenical in scope. The nations are increasingly defined in terms of chronic triage, and this has the effect (see Figure 6) of confirming their indigent identity by positive feedback. Within each separate nation, a recursive effect is observed: there is no real hope of rescuing the underprivileged category from the workings of the triage pump. The second feature is that this pump is primed by the motivation of greed, which respectable people call the profit motive. Respectability in this sense is concerned with the conventions which underwrite identity, and with the materialism which drives the societary machine. Empirical evidence abounds. I began by drawing attention to the steady rise of misery and agony. The Category A or B nation-states allow this: they do not forswear the manufacture of arms, and they fail to distribute the world’s superabundance of food and medicine. It is left to small bands of dedicated volunteers to try to help. More than inefficiency is behind this: the triage derives from a structural flaw. Identically, at the nation-state level of

World in Torment

791

K 33,3/4

792

recursion, respectable people sit at their dinner-parties discussing the “problem” of thousands of abandoned children roaming the streets outside, with resort only to stealing or prostitution, and chronic societary triage determines that nothing can be done. Again, small bands of volunteers do their best to mitigate the hurt, but society itself is astoundingly detached. And yet, for the disadvantaged third of the nation, and the starving third of the world, and for all the good and caring people who share both national and international concern, which is to say in total more than half the people alive, the ruling ideology is not acceptable. For them, the bottom line is not the bottom line. They know there is such a thing as social good. The Redundancy of Potential Command We know in political terms why more than half the world is disempowered by low variety. In the kinds of democracies which we run, which are better described as elective dictatorships, the individual is media-swamped by highly amplified low variety: she/he is robbed of voice, but also of intention. Thanks to the work in neurocybernetics of the great co-founder of cybernetics, Warren McCulloch, we have the theory of the redundancy of potential command. He refined the theory from his studies of the human brain – which is not organized hierarchically, which has no underlying ideology, and which is not single-valued. The brain cells, or neurons, have much in common with the individuals who make up society. There are a great many, they are unreliable and, when they die, they are replaced in their function by others. So the first thing is to say that neurons and people are organizationally redundant – highly redundant. This is not to disparage individual worth. It is only to say that graveyards are full of indispensable people. McCulloch reckoned that the brain is redundant in the order of 20,000 to one. People are often heard to say that this means that there is a lot of unused brain which could be developed to exhibit extraordinary powers. It is not the point: we need redundancy to make up for unreliability, dysfunction, and demise, and a society does too. Although various regions of the brain normally deal with various recognizable functions, there is a lot of interchangeability which derives from redundancy. And, in any case, brain functions are not localized and dedicated, as are functions in the management hierarchy of the firm or the state. So how are “decisions taken”, if they do not follow a given pathway up a given chain of command? The answer is that a concatenation of neurons operates when and only when information collects which enables a reliable “decision” to be made. Note that what I am calling a concatenation is defined from moment to moment by the presence of relevant information, and not by a pre-ordained structure hungry for facts. Now the number of possible subsets of ten billion neurons is effectively infinite. This means that command is not prefigured, but is potential in any concatenation which shares the needful information. Let us call that an information set, an infoset. It is obvious that not

only are individual neurons highly redundant, the infosets are too – since any infoset has the power of potential command. It interested both McCulloch and me that the most successful human organizations, be they never so hierarchical in appearance, operate on an understanding of the redundancy of potential command. He analysed old battles, particularly those of Nelson, to demonstrate this. Sea-mists and the smog of cannon fire made direct orders impossible to convey by signal flags, and Nelson’s captains took command of whatever local situation they could actually see and interpret. I soon discovered the same phenomenon in industry. Who is really taking the decisions about computers? It is not the bigwigs who have to sign the requisitions, but a subcultural infoset of junior young people who know the latest technology. The Board pretends to understand, but is easily manipulated by those who know the facts. This does not happen in our elective dictatorships, because it is not allowed to happen. I have formulated one of the mechanisms of censorship in the theory of chronic triage. The regulatory homeostats manipulated by mandarins and media prevent inconvenient facts from being properly understood by the people. The attenuation of political debate to mere “sound-bites” began the process; today it is common to see a gesticulating image uttering no sound at all, while voice-over commentary claims to say what is happening. Again, there is increasing resort to plebiscite where a yes-no vote is recorded in relation to a complicated, high variety package of propositions. Notoriously, Napoleon III used the referendum to outflank the democratic process, and today it comes with a spurious claim to be democracy itself. It should scornfully be recognized as the Government’s mandate to repeal the natural Law of Requisite Variety. Do we have to submit to these harsh facts? I answer that we do not so long as we can identify infosets which have redundancy of potential command. Sovereignty does not lie with the nation state, but with the individual. Even political prisoners under torture have found the courage to maintain their personal integrity. Certainly ordinary sovereign citizens may maintain integrity; the problem is how to empower their use of it to political effect. A Summary Theory of Team Syntegrity For many years I have devised alternative models for the infoset, and protocols for its efficient operation. The first account of the definitive model was written nearly ten years ago, although it was first published in 1990 [9]. In the meantime, much experimental work has been done, and a book [10] is in course of preparation. This is not the place, then, to give full details, or to justify the choice of form. What follows is purely descriptive of a particular social invention. It does work efficiently already: what will be newly discussed in the next section is its relevance to a politics of potential command. Please consider a group of five friends. No one is in charge of them. They have various interests in common, which is why they are friends. They talk

World in Torment

793

K 33,3/4

794

Figure 7. Creative Synergy: Five Friends Gather for a Serious Talk

about their interests – round the fire, walking in the park, over coffee, in the pub . . . If they discuss matters of moment to them, they might well come to a group insight, a crystallization of ideas which might be labelled creative synergy. The insight might well include a plan of action to get something done, something which needs more than one activist to achieve. Figure 7 shows the five friends, and how they come together. The insight is pictured as emerging at the centre of the group. Each of these people has other friends. Let us suppose that the focus of the information which the group developed is spread around a larger network of friends. An imaginary whistle blows by the time 30 people are involved: it is quite difficult to maintain an informational focus, especially if it is the source of action plans, with a number much larger than this. And if all 30 people arrange to meet, what are they going to do? Typically, people try to setup some kind of agenda. But that protocol means that in some sense what will happen is partly decided in advance, and even then the discussion is biased by the order of the agenda. We could divide the 30 people into six groups of five, seeking to recapitulate the original process of creative synergy. Five people form an animated group, and everyone gets a chance to speak. Then, when the whole 30 meet, six rapporteurs might report in turn. This is a good idea, and versions of it are often used. But we are talking about integrity: the best way to generate synergy which is also integral is to close the network in upon itself. Then the 30 people are the struts which define the edges of an icosahedron. This is itself a strongly bonded organizational framework, depicted in diagram A of Figure 8. It is obviously not hierarchic: it has no top or bottom or sideways. Each person is a strut which forms an edge, and none can be differentiated from any other. Moreover, the model preserves the fivefold group

World in Torment

795

Figure 8. The Icosahedron: A: Featuring its Twelve Vertices; B: Featuring the Internal Braces

with which we began. There are 12 such groups (the vertices of the icosahedron), and each person in this three-dimensional network belongs to two of them. Think how ideas are free to buzz around, and to return to base – laden with honey. We have constructed a kind of social machine for generating not just six but 12 key foci of concern. Buckminster Fuller [10], for example, published posthumously, noted that all constructions in nature depend on forces of tension, as well as on the compressive forces beloved of architecture. He so often used the phrase “tensile integrity” that he shortened it to “tensegrity” – and invented the geodesic dome on the strength of it ( pun intended). So far so good. But this structural model already clearly gains further strength, if we can tack it together from inside. If we put in the connections such as the four marked in diagram B of Figure 8, we end up with 30 internal cross-braces. Each vertex is the root of five braces which connect it to five other vertices, each of which is one vertex removed from its nearest neighbours. If you wish to live inside a geodesic dome, you cannot clutter up the internal space in this way; but in a model of social integration no such problem arises. Even so, there are procedural and theoretical reasons for not representing the six major axes which connect the vertices from pole to pole. The model now represents a maximal integration of the 30-strong infoset, and it is well structured to maximize synergy too. Following the linguistic convention set by “tensegrity”, this structure is called a model of Team Syntegrity. It remains to specify how to turn this static structure into a dynamic process, and that is the task of Syntegration – a procedure governed by a special protocol. Each person finds that he is a member of two separate teams, trying to bring creative synergy to bear on a concern.

K 33,3/4

796

Each team is aided by five critics, appointed from other teams, and following the pattern of the braces which cross the internal space. A logic had to be developed to make a coherent allocation of critical resources which does not counter the requirement that every participant has exactly equal status. Then each person, a member of two teams, finds that he is a critic of two other teams. The logic also facilitates the self-organizing of the whole group: there is no one who has authority to make appointments. Ideally, it takes five days to carry through the protocol which I have designed and validated by experiment for enabling all this to happen, and to generate action-direction outcomes. But, as shall be seen, it is possible to shrink the exercise to three days.

The Power and the Use of the Team Syntegrity Model Something has to be said about the strength and power of Team Syntegrity. From an engineering standpoint, the strength is apparent. From an aesthetic standpoint, the structure is beautiful. But, as the fourteenth century builder Vignot said about the dome in Milan, ars sine scientia nihil – art without science is nothing (and the reverse is also true). So artists over the centuries have studied the five regular, convex, continuous Platonic solids which the Greeks extolled – the tetrahedron, cube, octahedron, dodecahedron and icosahedron, all of which can be contained inside each other, and generated by one another. Geometric analysis of the icosahedron reveals a symphony of relational harmonies based on the Golden Section a=b ¼ b=ða þ bÞ: This ratio, when p b/a is called x, yields the equation x 2 2 x 2 1 ¼ 0: The positive root is ð1 þ 5Þ=2; that is 1.618. Iterations of rectangles based on this ratio generate the logarithmic spirals which we find in biology, and the Fibonacci series which we invoked earlier. Small wonder that these all-pervasive harmonies gave rise to the Pythagorean number-mystic, then the Hebrew Kabbala, and so on down the centuries. The graph-theoretic mathematics of the syntegrity model advanced were first set out by Dr Assad Jalali [11], and are discussed at length separately [12]. The key insight, however, has to do with the reverberation of ideas within the syntegration process which every participant so far has noticed through experience. The model has 30 edges, 12 vertices, and a valency (k) of five. Then it is a k-regular symmetric, ergodic, connected graph of diameter three, which exhibits involutive automorphism. That sentence, though needing much explanation, is mathematically exact; and the last two words are helpfully evocative. For they suggest the self-consciousness which the reverberating infoset actually exhibits. Many years ago I learned from Heinz von Foerster that awareness depends on an eigenvalue which is the characteristic root of a particular kind of mathematical function which computes itself. That is what

happens in the p “Staffordian Graph” here presented, where the eigenvalues relate to the 5 derivative already seen to underlie the Golden Section. Perhaps the most virile example of an infoset with which society is familiar is the directive management team of an enterprise. Take the leading directors of a company board; add the most respected staff aides; include (possibly) representatives of workers, clients and the community: here are 30 people strongly connected by a motive, a collegiate purpose. Can they afford to meet for an intensive five-day exploration of the future of their enterprise, using the Team Syntegrity model and protocol? If so, they may avail themselves of a service [13] set up for the purpose. If not, they are probably condemning themselves to years of orthodox strung-out committee work which ties up thinking time, exhausts patience, frustrates innovation – and may be too late. For societary purposes, however; we return to the image of a group of friends. These might indeed constitute a municipal council; they are more likely to be a loosely constituted gang of citizens, exercising their rights as sovereign individuals to try to get something done. If they are living in the same area, I call them a neighbourhood infoset. But perhaps they are not geographically contiguous. When Teilhard de Chardin wrote about the geosphere, the terrestrial world of rock, he saw it as enrobed in the biosphere, a multifarious but seamless web of life, which is in turn surrounded by a cloud of knowing – mind, the neosphere. Between the last two integuments I have inserted the technosphere – the infinitely differentiated network of modern communications. And a group of 30 like-minded friends, sharing a purpose but living well apart, may share their intentions and formulate their actions by electronic means via the technosphere. I call 30 such friends a global infoset. World Syntegration: An Action Plan You are now asked to set the team syntegrity model, the Staffordian Graph, in an ecumenical context which concerns itself with the redundancy of potential command. There are people all over the world, sovereign individuals, who have ideas and purposes which they wish to share with others. They do not see themselves as bound by hierarchy (even to their own nation-states) or committed to the processes (even those called democratic) which demand the establishment of political parties, dedicated movements, delegations – or indeed high-profile leadership. These people are the material of infosets: neighbourhood infosets of 30 local friends, global infosets of 30 world citizens. Infosets of either kind formulate themselves, because they constitute potential command posts; they spread epidemically, demonstrating their redundancy; they interact massively, as is the nature of shared commitment. It is not a commitment to some shared manifesto, but a commitment to circumvent folly, wherever it is found; it is a commitment to alleviate suffering; it is a commitment to brotherhood and peace.

World in Torment

797

K 33,3/4

798

This worldwide syntegration does not of course exist. It is a vision. But, although visions may be inspirational, they do nothing much to alleviate suffering, until inspiration is embodied in a plan of action. And, if mounting human misery is the product of a triage machine, as I have argued, and if the triage machine is endemic to the ruling world ideology, so that it cannot be dismantled, then the action plan must circumvent triage altogether. The aim is ambitious: to start a process which invokes the redundancy of potential command as the methodology for a new system of world governance. We have the seed: its germination requires that it be sown in a nutrient medium on a sufficient scale that the plant takes hold and spreads. I seek to enlist the help of Key Collaborators to this end. It is not mere whimsy which asks for 30 such people to enlist, as will become clear. To mobilize action, we need a launch date for the endeavour: the week of 26 July 1993 has been nominated [14,15]. The Programme is quite modest for the potency which it ought to project. Imagine: (1) Thirty neighbourhood infosets meet for three days during that week, using a specially designed syntegrity protocol which will be provided for their use. These 30 meetings are in widely dispersed places around the globe: it would be good to have each continent represented. (2) Thirty global infosets are operational over the five working days of that week, their exchanges borne electronically on the technosphere, using an electronic version of the Syntegrity protocol, for which software will be made available. (3) Each of the 60 (in total) infosets has the same remit for discussion: . Isolate 12 key features which would characterize a world without triage to improve the lot of humankind. . Express each in terms which indicate urgently required action. . State in each case what action this infoset will immediately undertake in furtherance of this willed future. The infoset is absolutely in charge of its output, generated within the Syntegration protocol which is provided. There is one logical test which should be applied to every sentence: .

This sentence could be disputed without absurdity: that is, its negative is plausible.

“Feed the hungry”, “heal the sick”, “educate the ignorant” are injunctions which no sane person would dispute. “Drop enough food by parachute to satiate the black market and rival armies, so that plenty is left for ordinary folk; send the bill to the previous colonial power” is a substantive, because disputable, proposition.

(4) An operations centre is in commission throughout the launch week in Toronto, Canada. Its task is to collect, collate and synthesize the 720 statements arriving from the 60 infosets. Note that 1,800 people are simultaneously active in this manifestation of potential command. Press releases chart progress throughout the week.

World in Torment

Progress to what, if not to some sort of “declaration”? It is not the right question. We are trying to invent a societary learning system to replace the triage machine. If you stop the triage machine, freeze a frame, you capture the same categorization – only the measures of misery get worse. If you enshrine pretended solutions in a Manifesto, you must begin to redraft it the day after it is adopted. If you enact a Constitution, experience shows that you must soon start passing Amendments. Twenty-five years ago, I coined the phrase “the aborting corporate plan” ( [16], for example). The argument was that as soon as any ecumenical planning office had finally bound “the Plan”, and circulated it, new information arrived to make it obsolete. Therefore it must continually abort. The same is true of any set of intentions which would bind world governance in a transitional age. What we need is a continuously aborting corporate plan which resolves into a fluid, adapting, programme of actions at all levels of society, that is in process of constant adjustment. Now freeze a frame: you have simply a snapshot of action-oriented thinking always under development, in continuing operational mode, sponsored by continuous syntegration on the part of all the redundant resources of potential command who care enough to join in. Even the best of our democratic processes results in a consensus by the lowest common denominator. Continuous syntegration is expressive of the highest common factor. As an action-oriented procedure, it follows the fundamental rule of governance: do not decide on the next action, until you have observed the outcome of the last action. McCulloch used to say that this rule embodied the quintessence of cybernetics. Every proposition which emerges from the elaborate processes of Launch Week next July is logos spermatikos. It is formulated to engender its own progeny. The final move of the week-long programme is designed to harness this life-giving energy to facilitate the propagation of continuous syntegration, thus: (5) Consider each set of 30 infosets, the one of neighbourhoods, the other of global sets. We conceive of each as the Staffordian Graph of its own 30 graphs. They are hypericosahedra. Here is a graphic image of what this means. Take the 12 propositions of each infoset. The members decide which are the two most important, or most conveniently inclusive for what follows. Pull the icosahedron apart by these two vertices, so that the construction begins to disintegrate.

799

K 33,3/4

800

Catch hold of each of the ten propositions as they fall away from the cohesive whole, and seek to incorporate it in one of the two chosen statements. The first-generation icosahedra are now struts (edges) of a new Staffordian Graph, each carrying an enhanced proposition at each end. This means that each vertex of the second generation hyperinfoset receives five inputs as starting propositions for its own syntegration. The ordinary infoset has a protocol for generating its initial statements out of the minds of its members; these second-generation inputs are expressions of the group minds of its founding infosets. It is suggested that the human individuals who carry this through should be randomly selected, one each, from the original 30 infosets. It is vital, according to these principles, that they should not be “delegates”, but free spirits who have learned from their initial syntegration. As happens with the generation of species, we should expect repeated generations of infosetters to form and to overlap in experience. But these are small societies, not individuals; and we should expect their membership to change according to the redundancy of potential command. The purpose of forming hypersets is not to create a hierarchy of warriors or of logical propositions which have enhanced status because they have “risen to higher levels”. It is to make consolidated statements available which have been worked over by 900 people, 27,000 people and so on. My own expectation is that entropy in the language system and changing times will soon vitiate this “hyper-hype” of the process. It will fix its own limits. There is no prospect of gnarled ten-level syntegrations formulating tenth-level propositions of unimaginable obscurity, emulating Hesse’s Glass Bead Game. Potential command should see to that . . . Alternatively, graph theoretic discoveries in the Staffordian Graph context may well result in an ability to formulate protocols which work n-dimensionally and therefore are not constrained to 30 infosetters at a time. We shall see. In the meantime, our world-in-torment requires actions of every sort which might help to ameliorate the plight of millions. The triage model, however sketchy, powerfully suggests the need for structural change in the technique of governance. No established authority shows any willingness to undertake it. The non-hierarchic organization which facilitates the redundancy of potential command is well founded in the example of all biological systems – in the brain outstandingly. Both the methodology and the technology are available to implement its societary analogue. No established authority, cowed by its own information technology, shows any willingness to explore it. Global syntegration, however designed, is a necessity in the face of such inertia. The icosahedral model backs my own proposal for global syntegration. But no one knows what it is, and the book [12] is fairly abstract. What we need is a start; a first demonstration, an ostensive definition, of what global syntegration

might look like, be and do. This is the purpose of the action plan. It is not grandiose, but practical. The proposal is directed in the first place to anyone with the courage to take a modest lead, and willing to be a pioneer of the Launch Week. I end by explaining what the role of Key Collaborator in that enterprise involves. The Key is a colleague ready to accept responsibility as a Patron for Launch Week. She/he needs to be in a position to undertake two tasks: given that further information and all supporting documentation will be available months in advance. These are the tasks: . Become a patron of a neighbourhood infoset. Select someone well-known to you who could organize a neighbourhood infoset. You would enthuse this leader, and “hold his/her hand”. How is the net to be cast for volunteers – how do we implement the redundancy of potential command? How can the requirements for rooms and equipment (all fully specified) be met? The Key is not expected to run the infoset personally – but she/he is welcome to do so. . Become a patron of a global infoset. This requires a computer port which can be made available for Launch Week, and a small team to man it – on a round-the-clock basis. This port is the pivot of a global group of 30 volunteers, who are being recruited by the Toronto organizers. The port communicates results to Toronto as they are generated (there are several iterations of the process provided for in the protocol to be disseminated). Thirty volunteers will be allocated from Toronto to your port’s infoset, aiming for a maximum dispersion worldwide. Of course the port team may recruit volunteers itself; but local volunteers who come to use the port facility will be allocated to other infosets. Again, the Key Collaborator may wish to run this port personally, or simply to authorize its use. Some of the 30 Key Collaborators have been recruited, and the plan hinges on their support. More are needed. Obviously, this is a scheme for realizing Requisite Variety by local autonomies, as earlier defined. A Project Director is already operating out of Toronto. Her name is Wendy Walsh. She has a big job to orchestrate 30 Keys: 1,800 individual infosetters are too many to co-ordinate directly. The gap between the ambitions of this plan and the stark realities of the tormented world is vast. No action which we take here can do anything about those realities in the short term. How easy it is to say (and many do): “The world’s not too bad, and we are working on a cure.” Nonsense: 40,000 children dying every day is bad enough, and the cure is not a genuine prospect. How easy, especially for us academic guardians of scientific probity to say (and many do): “Hang on a bit. This needs toning down. Wait for proof, and the approbation of peers. My conscience will not permit such reckless use of

World in Torment

801

K 33,3/4

802

science’s fair name.” Such worthy folk cannot have taken the most fleeting look at the history of science, or of social innovation. I do not hesitate to remind this conference again, as I did six years ago in London [17], that Lord Kelvin, a President of the Royal Society (no less) when the century began, called heavier-than-air flying machines impossible, and opined that X-rays would prove to be a hoax. As to conscience, I agree with Oscar Wilde: “Conscience and cowardice are really the same things. Conscience is the trade name of the firm.” I did say we needed courage. Let us take comfort in the words of still another friend from whom it was my privilege to learn: she is almost uniquely qualified to give this encouragement. “Never doubt”, said Margaret Mead, “that a small group of thoughtful, committed citizens can change the world – indeed, it’s the only thing that ever has.” That is true, is it not? The Buddha, the enlightened, is called Siddharta: “he who has fulfilled his purpose”: but it was a small beginning, not far from here. The Christ is named Jesus: he had 12 friends – and one of them was most unreliable. I draw no comparisons from these references, please understand; but I do draw hope. Let us get up and do something in our own shameful mess of a world. It is better than to make excuses; better than to sit on your tenure for 30 years, and hang your hat on a pension. Notes and References 1. Beer, S., The Heart of Enterprise, John Wiley & Sons, Chichester, 1979. 2. Beer, S., Diagnosing the System for Organizations, John Wiley & Sons, Chichester, 1985. 3. Beer, S., Brain of the Firm, 2nd ed., John Wiley & Sons, Chichester, 1981. 4. Maturana, H. and Varela, F., “Autopoesis and Cognition”, Boston Studies in the Philosophy of Science, No. 42, Riedel Publishing, Dordrecht, The Netherlands, 1980. 5. Beer, S., Platform for Change, John Wiley & Sons, Chichester, 1975. 6. Beer, S., “I Am the Emperor and I Want Dumplings”, Systems Practice, Vol. 2 No. 3, 1989. 7. Beer, S., “The Will of the People”, Journal of the Operational Research Society, Vol. 34 No. 8, 1983. 8. Hardin, H. The New Bureaucracy: Waste and Folly in the Private Sector, McClelland and Stewart, Toronto, Ont., 1991. 9. Beer, S., “On Suicidal Rabbits”, Systems Practice, Vol. 3 No. 2, April 1990. 10. Fuller, R., Cosmography, Macmillan, New York, NY, 1992. 11. Jalali, A., Association Schemes with Opposition, Research Report, European Business Management School, University of Wales, Swansea, 1993. 12. Beer, S., Beyond Dispute: “The Invention of Team Syntegrity”, John Wiley & Sons, Chichester, in preparation. 13. The commercial service which provides consultation and implementation for syntegration is provided by Team Syntegrity Inc, 34 Palmerston Square, Toronto, Ontario, Canada M6G 2S7. 14. The Action Plan put forward in the last section has been devised in collaboration with the World Service Authority, headed by the pioneer world citizen, Garry Davis. For anyone who

does not know the story of this brilliant, farsighted and courageous man, his most recent book [15] is compelling reading. The Year of the Launch, 1993, is the 45th anniversary of his declaration of world citizenship, and his friends will celebrate his birthday during Launch Week. 15. Davis, G., Passport to Freedom, Seven Locks Press, Washington, DC, 1992. 16. Beer, S., “The Aborting Corporate Plan”, in Jantch, E. (Ed.), Perspectives of Planning, OECD, Paris, 1969. 17. Beer, S., “Holism and the Frou-Frou Slander”, Kybernetes, Vol. 17 No. 1, 1988. Further reading McCulloch, W.S. (1989), Collected Works, Four Volumes, Intersystems Publications, California, passim.

World in Torment

803

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Knowing Norbert Stafford Beer Toronto, Ontario, Canada

804

Keywords Literature, Cybernetics Abstract Gives an account of Norbert Wiener’s works and how they influenced the author. The seminal Cybernetics or Control and Communications in the Animal and Machine introduces the scientific world to the notion of cybernetics and indeed, according to the author, split it down the middle. Goes on to describe Wiener’s other major works and the impact they had not only on the author but also on other writers, researchers and scientists. Concludes with a brief description of how the author helped Wiener to arrange a European trip in 1964.

Well he was, wasn’t he? Knowing enough to have drawn comparisons with Liebniz – and that entails transdisciplinary knowledge, which was the key to his insight. My own knowing of him began in 1950. Somehow I acquired a copy of Cybernetics[1], which had been published two years earlier. It irked me for years that I was unable to recover the memory of how that book came to me. It was not a purchase or a loan, but a gift – that’s all. Eventually I had to settle for the fey notion that a Good Fairy had left it on my pillow. Even two decades later I described it like this in a vignette. Difficult, quixotic, immensely stimulating (then and now), Cybernetics split the scientific world (for those who read it) down the middle.Think of it like this: the great man (he really was) holds forth to his friends after dinner, ruins the tablecloth by scribbling mathematics all over it, sings a little song in German, and changes your life. It is tough going. You have to stay the night[2].

Kybernetes Vol. 33 No. 3/4, 2004 pp. 804-808 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523706

The book had indeed changed my life. In 1950 I was running an operational research unit inside a British steelworks. An early assignment had produced a whole new design for internal regulation, with the result that I had been appointed to the managerial role of production controller as well. Cybernetics inspired me to conceive of whole-company regulation, and by 1956 I was running management science research for the whole corporation of United Steel in Sheffield. There were soon 70 scientific staff in a unit that included the word “cybernetics” in its title, and operated out of “Cybor House”. Also in the year of 1956, the world’s first conference on cybernetics was hosted by the city and the province of Namur in Belgium. An international horde of scientists descended on the quiet town: no one could believe then how many were already captivated by these ideas, and few can believe now that an event of this magnitude actually happened. (But see the Proceedings[3]). It is surely important to appreciate how quickly the word had spread. Everyone First published in: Kybernetes, Vol. 23 Nos. 6/7, 1994, pp. 17-20. q MCB University Press, 0368-492X.

there seemed to know Cybernetics; and by this time the social and industrial relevance of Wiener’s thought had been highlighted by his The Human Use of Human Beings[4], a vital text the lessons of which have not been learned to this day, although the book is widely studied. It was a great disappointment that he was not himself present, but he had been much occupied in the Far East during that period. On the other hand, he did not attend the second Namur conference in 1958 either. Perhaps he had not fully grasped the degree of excitement in Europe surrounding the whole movement; or perhaps the personality clashes which seem characteristic of innovative advance in every field had already become a problem. At any rate, many of the Namur delegates were by now familiar with the Josiah Macy Jr Foundation conferences, among others, held annually from 1946 to 1953 (see Heims[5]), and with other major influences, notably Design for a Brain[6]. The result of all this was that I did not meet Wiener personally until 1960, when I called on him at MIT (see [7]). By then my first book, Cybernetics and Management[8], had been published; he knew not only that I had been active in the field, but also that in describing it I gave pride of place to his own achievements. He was on the verge of retirement from MIT at that moment; but as Emeritus he remained very active, and travelled widely. In particular, he became associated with the school of cybernetic thought at Naples. But his base remained at Cambridge, Massachusetts, and I saw him there too from time to time – notably in early 1964. He was then planning another European trip, and wanted to visit some people and some familiar places in England (he had been at Cambridge with Bertrand Russell before the First World War, and was back there again in the early thirties). So I offered my help. But before completing that story, I should like to say a little more about Wiener’s books and their impact. During the fifties, I was quite active in the Royal Statistical Society, particularly in the area of industrial applications (the “or” in Cybor House stood for operational research after all). Extrapolation, Interpolation, and Smoothing of Stationary Time Series[9] was a standard reference in these circles, but most people did not know what to make of Cybernetics. The vignette with which this note begins said that it “split the scientific world”, and I had a ring-side seat as far as England was concerned. The great names in statistics, such as R.A. Fisher, E.S. Pearson, M.G. Kendall, were arraigned in rather hostile camps even in terms of their own expertise. The Operational Research Club, (which antedated the OR Society), met in the rooms of the Royal Society itself, because the founders of OR such as Patrick Blackett and Charles Goodeve were Fellows of the Royal. Of course, I was second generation; and I must say that all those mentioned were extremely nice to me; some others were not. The split over Cybernetics led to warnings – just as I had incurred the wrath of some over my support for the Jesuit Teilhard de Chardin’s Phenomenon of Man[10], which was vitriolically assaulted by the biologist Peter Medewar at Cambridge, and defended by (of all people)

Knowing Norbert

805

K 33,3/4

806

Julian Huxley, a supposed atheist. The difference now was that I was taking cybernetics into the world of affairs. Wiener’s two autobiographical books[11,12] occasioned no alarm; but it has to be admitted that the English savant (if such a normative term is conceivable) tends to treat such works as the mark of egoism, and their content as self-serving. Tant pis for the savant, one could say. But the book that occasioned me nearly as much excitement as Cybernetics itself came out in 1958, and once again led to dispute among the intellectual lite. Few people these days seem even to have heard of Non-linear Problems in Random Theory[13], and the title is hardly “sexy”. The fact is that it triggered some important new ideas for me because of its fresh methodological orientation. In an obituary notice I wrote: “The importance of this typically Wienerian, typically strange, penultimate book seems to have been much underrated so far”. Thirty years later, it is lost altogether. At the time, opponents were offended by Wiener’s mathematical incursions into neurology in particular. Then we may celebrate the fact that he was named honorary head of The Netherlands Institute for Brain Research at Amsterdam in the year that he died. But to me, all worthwhile thinking is essentially transdisciplinary, and this is why Wiener earned his Leibnizian epithet. To be interdisciplinary is always fascinating, and offers intellectual rewards; but to be transdisciplinary is more than this. Even within a single discipline, it is possible that ideas transcend its boundaries – because they turn out to be (in some interesting sense) universal. The same could be said of God, Golem, Inc.[14], with its consideration of science and religion, although these ideas were not so novel for me personally. It is to Wiener the transdisciplinarian that I make this obeisance, noting the courage it takes because of the offence it gives to soar away from the safely commonplace. When they stop occasioning fury and denunciation, innovators and agents of change have stopped. Period. The obituary quotation above referred to the Non-linear Problems as the “penultimate book” – God, Golem, Inc. being the “last”; but Wiener was to have the last laugh about that. How many readers have seen the fascinating little book called Invention: The Care and Feeding of Ideas [15], which was published in 1993. The author is Norbert Wiener. And no, he is not our subject’s own grandson, but the pioneer Wiener himself. The text of this book was found among his papers, dated June 1954. Here, writing for the general public, Wiener makes an impassioned plea against the misuse of science, its subjugation to big business (he calls it Megabuck Science), and much more that was insightful and germane to social reality when he wrote it. Now, 40 years later, far from being quaintly dated, it is more urgently needed than ever. It seems that Wiener changed his mind about how he would treat this material (he contemplated a second novel as the vehicle), and he returned the publisher’s advance of $500 in 1957. Steve Joshua Heims says in his current introduction:

It is just possible that when he went back to the 1954 draft in 1957, he found its implicit optimism unrealistic. As with all his ideas, he would be inclined to worry that his suggestions for fostering new inventions would be taken up most readily by people whose purposes he abhorred.

I can well believe this. Sadly enough though, I also believe that he need not have worried. He should not be worrying on that score now: the problems that he perceived have become worse, not better; and the cybernetics of the societary system that generates them is clearer now to see as fundamentally pernicious. It is to be hoped that, the publication of Invention[15] will reawaken interest in Wiener’s social thought, and encourage a new generation to trace it back through Human Use[4] to Cybernetics[1] itself, so that the mathematical foundations of Wiener’s extraordinary insight will be understood. He was not, after all, a mere fortune-teller. In this retrospective journey newcomers may take advantage of a guide book already mentioned in Part I of this special issue. It is written by Professor P.R. Masani, and provides an excellent account of Wiener’s life and works[16]. May I now complete the story about my offering help to Norbert Wiener in connection with his trip to Europe in 1964. He had engagements in Sweden, and would then come to England. At the time, I lived to the south of London – within an easy car journey of Heathrow Airport. The plan was that I should meet him and his wife, and that they would stay at my home for a few days completing their arrangements for further visits. We prepared the guest room, opening windows and setting flowers in place. It was 18 March 1964. The phone rang. It was Mrs Margaret Wiener, calling from Stockholm. Her husband was dead. Anyone who understood his personal testament in science, and relished the attached Apocrypha of engaging personal anecdote, mourns him yet. References 1 Wiener, N. (1948), Cybernetics or Control and Communications in the Animal and Machine, MIT Press, Cambridge, MA, ; and John Wiley & Sons, New York, NY. 2 Beer, S. (1972), Brain of the Firm, John Wiley & Sons, Chichester. 3 Proceedings of the First (1956), International Conference on Cybernetics, Namur, Belgium. 4 Wiener, N. (1950), The Human Use of Human Beings, Houghton-Mifflin, Boston, MA. 5 Heims, S.J. (1991), Constructing a Social Science for Postwar America, (a history of the Cybernetics Group 1946-1953), MIT Press, Cambridge, MA. 6 Ashby, W.R. (1952), Design for a Brain, Chapman and Hall, London. 7 Harnden, R. and Allenna, L. (Eds) (1994), How Many Grapes Went into the Wine: Stafford Beer on the Art and Science of Holistic Management, John Wiley & Sons, Chichester. 8 Beer, S. (1959), Cybernetics and Management, English Universities Press, London. 9 Wiener, N. (1949), Extrapolation, Interpolation, and Smoothing of Stationary Time Series with Engineering Applications, MIT Press, Cambridge, MA.

Knowing Norbert

807

K 33,3/4

808

10 Teilhard de Chardin, P. (1959), The Phenomenon of Man, Collins, London. 11 Wiener, N. (1953), Ex-prodigy: My Childhood and Youth, Simon & Schuster, New York, NY. 12 Wiener, N. (1956), I am a Mathematician, The Later Life of a Prodigy, Doubleday, Garden City, New York, NY. 13 Wiener, N. (1958), Non-linear Problems in Random Theory, MIT Press, Cambridge, MA. 14 Wiener, N. (1964), God, Golem, Inc. – A Comment on Certain Points where Cybernetics Impinges on Religion, The MIT Press, Cambridge, MA. 15 Wiener, N. (1993), Invention: The Care and Feeding of Ideas, MIT Press, Cambridge, MA. 16 Masani, P.R. (1990), Norbert Wiener 1894-1964, (in the series Vita Mathematica), Birkh user, Basel.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Man in a garrulous silence Stafford Beer

Man in a garrulous silence

President, World Organisation of Systems and Cybernetics (WOSC) Keywords Cybernetics, Culture (sociology), Systems theory

809

Abstract Written in 1982 for a closed meeting of scholars, provides an insight into the thinking of the time and in particular to the contributions of the author to systems and cybernetics. Illustrates the challenge of technology to humankind and sees the coming of a second Tower of Babel which is as counter-productive as the first. Believed it was necessary to consider practical strategies for change. Looks at the concept of human identity, considering “Oneself: a systems viewpoint”, “A Tripartite selfhood” and “The community consequence”. Provides a schematic framework for the discussion of the social cybernetics of the human condition. Considers a basic stance, and the application of rules detected by cybernetic scholarship. Outlines strategies for inducing change and examines new dimensions for planning. Emphasizes the importance of regarding planning as a matter of obtaining recognition of the frameworks that have been introduced and then facilitating choice.

At one time, we read, “the whole earth was of one language, and of one speech”. But this monoglot humankind was very soon scattered over the whole earth. The single language was “confounded” so that people could not understand one another’s speech. The reason was that monoglot humankind had built itself a very high tower – a tower which, because language was confounded there, received the name Babel – meaning “confusion”. It seems from this story (Genesis 11) that our species set about getting things systematically wrong soon after Noah’s flood. The Book of Genesis declares that the tower was built precisely in the interests of solidarity – to “make us a name” – and specifically “lest we be scattered abroad upon the face of the whole earth”. So here is the prototype of the policy that produces just the opposite of the desired effect. Here too is the prototype of the means we always select by which to implement our counter-productive policies – namely technology. “Let us make brick, and burn them thoroughly. And they had brick for stone. . .” As usual, technology ran away with itself: the tower was reaching up to heaven itself. When this became apparent came the judgement: “This they began to do; and now nothing will be restrained from them which they have imagined to do”. Then came the scattering. Even so, and in any case, nothing was “restrained from them”. We have landed on the moon, and probed to the far planets: I do not need to recite the list of marvels that technology has brought true in the twentieth century. However, humankind has understood very little about the proper terrestrial use of all this technology; of our own true nature, of the quality and genesis of human Written in 1982 for a closed meeting of scholars, and first published in: Kybernetes, Vol. 25 No. 7/8, 1996, pp. 14-31. q MCB University Press, 0368-492X

Kybernetes Vol. 33 No. 3/4, 2004 pp. 809-827 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523715

K 33,3/4

810

consciousness, and of the governance of human affairs, nothing much has been learned in 100 years – although I strongly suspect that much known long before has actually been lost. Today’s brash know-how has reconstructed Babel’s Tower: everyone talks to everyone else in a confusion of tongues and electronic images all the time, day in and day out – the globe spinning inside an envelope of noise. The idea was that technology would make the world smaller, and therefore more intimate; that increased communication would mean better understanding; that shared information entailed synergistic strategies to yield collaborators’ surplus. . . These were pious hopes indeed. The second Tower of Babel is as counter-productive as the first for reasons founded not so much in profound psychology or irreconcilable idealogies as in simple arithmetic. When the tribes scattered, each was free to make its own local mistakes – wiping out a few species here and there, creating the occasional dust-bowl, engaging in mighty wars that decimated the contestants. But now that the tribes have again been brought together, with their devastating technologies (the adjective is well-chosen), their joint pollution (in the widest sense) embraces the whole Earth. It seems that a species is being made extinct each day, on a rising scale that reaches 10,000 species a year by 1990. By the year 2000, 20 per cent of the species now on Earth is expected to have vanished – and the gene pool from which new solutions to our problems can be drawn will be thereby and thus far depleted. What I mean by saying that this is due to simple arithmetic is that oecumenical trends rise exponentially because all the component parts are now interconnected, and there is such a thing as compound interest. By the same token, the number of relationships that operate inside the Tower of Babel that is attempting to regulate all this grows as n(n 2 1), so that every attempt to include a further input or a further viewpoint on any matter increases the regulatory complexity involved by almost the square of the number of elements concerned. The growth of a bureaucracy competent only to stifle its own regulatory intentions is built in to this situation, and may indeed be observed growing in this obscene fashion on every side. The expectation that raising the level of aggregation will help (as in national super-ministries, international regulatory bodies such as the EEC, or planetary invigilators such as the WHO) will not work in a dynamic system, because that system continuously proliferates variety – that is to say, it generates new states of its own existence non-stop. The bureaucracy must grow at an equivalent rate, according to the cybernetic canons that govern regulatory systems, and then fails to cope with its own variety – never mind that of the system with which it is supposed to be concerned. It becomes obsessed with itself, in short, and constitutes a vast drain on resources.

All of this is bad enough if we take the situation as given, and examine its trends. But to do so disregards the fact that neither the world system as a whole nor any of its component elements is monolithic: in fact all are volatile systems in highly unstable equilibrium. Any one of these equilibria, the balance of nuclear arms or the internal balances of a nuclear power station, the balance of food production or of the total ecology – any precarious balance in any department of our affairs, might suddenly break down in catastrophic collapse. It is evident that science does not know enough about turbulence in very large systems, despite the work of physicists, does not know enough about the behaviour of large interactive human communities, despite the work of social scientists, does not know enough about the regulation of viable systems, despite the work of managerial cyberneticians, even to predict where, when, or even how a disastrous collapse is likely to occur. For, after all the slogans have been heard and forgotten, the Earth really is one unit with an integrated ecology of rocks and water, winds and tides, plants and animals, and our one human species has taken the whole habitat to the stage where the entire set of equilibria that informs and coheres the whole is susceptible to imminent disruption. I reiterated that, in each department, “we do not know enough”. What we do know, most dreadful news of all, we have so far failed to metabolize in structural change. The froth of talk goes on, a babel of languages, portending a plethora of absurd and dangerous outcomes, amidst which sits the silent spirit of humankind – at best lonely, alienated and afraid; at worst starved, degraded and about to die. Can we, then, do no more than add to the confusion? And is it not still true that the first action of the powerful when faced with frightening news is to kill the messengers who bring the tidings? It is necessary, in considering practical strategies for change, to take account of such baleful questions. We should, I suggest, instantly conclude that elaborate proposals are not likely to be heard, nor understood, nor – finally – tolerated. There is probably not time to reconstruct the regulatory mechanisms that are failing us, even if we could get agreement about that; and this is why so many people, from untutored romantics to hard-headed scientists, are prepared to believe that extra-galactic sources of information will come to the rescue. The best that we can do is to propose a fundamental notion that is clear and simple and direct enough to pass into general currency, yet sufficiently accurate diagnostically to have beneficial effect, and be capable of amplification by civilized means rather than coercion. Part 1: A concept of human identity Oneself: a systems viewpoint Consider firstly a self, one human person, to be called then: Oneself. This Oneself is not a fixed identity, to be labelled by a photograph with a number

Man in a garrulous silence 811

K 33,3/4

812

hanging round its neck. It is a complex system. In physical terms, it is a whole collection of components – whether we consider it mechanically, or chemically, or electrically, or in any other way. It is, for instance, a self-producing, self-reproducing puddle of genes drawn from the gene pool of humankind. Or it is a self-propelling action-man driven by an electro-chemical, three-pound computer, having ten billion switching components, and running on glucose at 25 watts. And so on. Even in these physical terms, and even in solitary confinement, this system called Oneself has an incredibly large number of possible states – a number measurable in principle, and called its variety. Take Oneself out of solitary confinement, and its variety leaps up exponentially as the number of relationships subsisting between itself and the world around it rapidly grows. These relationships refer to other people who are also high-variety systems, and to environments in general, of which the same is true. Far from being a simple, single entity, Oneself is a high-variety system set in a shifting and uncertain high-variety context of systems. The boundary conditions are difficult to draw; measures of any sort are difficult to make, identities are therefore obscure. We try to capture them in the points scored in silly little questionnaires; whether these are drawn up by social scientists, examiners, insurance salesmen, policemen, or popular magazines, they are all of low variety. So are quasi-psychiatric epithets such as “paranoid personality”, and so forth. These approaches have their uses. The trouble is that we end up with low-variety legislation to match the low-variety people invented by our inept techniques, while the real Oneself is left beyond it all – unloved and unexplained. Of all such low-variety regulatory techniques, the only ones that work are those that attenuate the variety of the physical person rather than the social account of the person: namely imprisonment and death. No wonder that they are used so often, and on a genocidal scale. There are potent consequences in the commonplace as well. We say: “He is not himself”; and we say: “she is exerting a woman’s prerogative”. By this we mean quite simply that we do not fully understand either this himself or that herself. Pushed to professional limits, we could say that much mental illness, much crime, much bankruptcy and so forth (all being pathological conditions of Oneself systems) offer names for the inability of psychiatrists or social scientists or economists fully to assimilate the variety generated by their clients. And if the Oneself that is myself does not behave properly in my own eyes and according to the expectations of myself, what then? Can it be that Oneself cannot fully comprehend one’s own burgeoning variety? Yes, it can. And how often does one hear quoted the wise injunction of Delphi: “know thyself”, as if to confirm that most people do not know themselves. We seem to have begun the definition of some of the problems that are associated with being Oneself, and to have done so in just the way our Western culture would expect. Illness is a medical problem, crime is a social problem,

and so forth. We know something about it; if we knew more we could do better; therefore we need research funds, and we need to establish causes and effects; then we shall need to legislate. The same for economic problems. The same for international problems. As to the myself-Oneself, the Western culture suggests that we come to terms with our own weakness, that we educate ourselves, that we submit to medical regimes, that we confess to being miserable sinners, that we look for blame in heredity or upbringing. It does not work out at all well in any of the contexts that I am moving in front of your eyes. After all, our “defence” systems facilitate aggressive wars, our medicine is some 50 per cent iatrogenic (thus causing as much damage as it cures), our penal systems are breeding grounds of crime, and so on. It is because we are wrongly contemplating the difficulties that large-scale systemic interactions invariably evoke. What is invariant about all these difficulties thrown up by selfhood in a universe of selves is that Oneself has a regulatory problem. The variety generated by the human Oneself is far greater than the variety disposed by the regulators available to assimilate it. This circumstance attempts to contradict Ashby’s Law in cybernetics, which says that the variety of the regulator must be at least as great as the variety of the system under regulation. For this to happen in society, the regulators would have to be richer in variety than our culture’s regulators can manage to be – or even are permitted to be, because of the threat to individual freedom that might be entailed. Therefore they do not work; and it is very interesting that no commitment of further effort or research funds, no more casual hypothesizing or even mandatory sanctions, can make them work. They are basically flawed. And when again we come to the Oneself who is myself, the flaw is outrageous, and generates doctrinal explanations such as original sin, schizophrenia, imperfectability, and the generation gap. This is because the variety mismatch between the primitive internal regulator and the cornucopia of the human spirit is so vast. As to the myself-Oneself caught up in this kind of society, which is indeed a perversion of society because it inhibits social solutions in favour of central direction, the human potential is squashed. The variety natural to men and women is drained away from them as their number of possible states becomes more and more circumscribed. This happens by law, insofar as less freedom for Oneself is deemed to be wise for general good. The more general this “general good”, the greater the loss of freedom – since an increasing number of constraints is required in order to meet the minimum requirements of all the parties. But not only laws have this effect of draining away the variety of Oneself and diminishing the chances of realizing human potential. Incredibly, education often reduces the variety it is meant to increase by insisting on particular academic patterns and formats – and by “knowing the answers” to questions that are not yet even appropriately posed. The media, and in particular television, are most potent variety inhibitors. In the first

Man in a garrulous silence 813

K 33,3/4

814

quarter of 1980, the average television set in that culturally trend-setting country, the United States, was switched on for nearly seven-and-a-half hours every day. The variety of those watching is truncated by the act of watching, often in a daze, the brain free-wheeling in its alpha rhythm, when the human being could be doing other things. It is a second time truncated by the content of the televisual message – modifying, as we know it does, future behaviour, and so by definition lowering the threshold of sensibility and diminishing the capacity to adapt. Yet more incredibly, the administration of good works has become a major drain on the recipient’s variety – and therefore on his or her resilience and stability – by becoming institutionalized and bureaucratic. These pressures, expensively exerted on Oneself, are counter-productive to any regulatory norm that cybernetics understands. Then we enquire why it is that the internal regulator is primitive, the answer is always the same. Simplistic accounts of Oneselfhood in itself and in all its contexts have been propounded and accepted. They are simplistic and low-variety because they treat Oneself as an entity instead of a system, and they treat relationships as the clashing of billiard balls instead of recognizing in them the interpenetration of high-variety systems, in which much uncertainty prevails and boundaries become ambiguous. Thus the law says: “Guilty as charged”, or in billiard-ball terms: pot the red. But a higher variety perception then that offered by the statute book says: “I blame the dehumanizing environment”, or in billiard-ball terms: in-off the red. “We are all to blame” expresses a higher variety yet: but the public has potted its opponent theorist’s white, and the red is left in baulk. Thus the model, however clever we are with it, is variety-inadequate to our purpose; and it is the model that needs to be changed, not the talking-about-the-model that passes for progress in the Western culture. This is especially evident in the Oneself-myself case. People become ill, and disturbed, and unmanageable in their inability to match the model they have of themselves to their public performance; but they are taught and exhorted and drugged to modify that performance, which is real and often adaptive, to fit the implanted low-variety model which is neither. Their mentors who purvey the model and who subsequently reinforce it are not aware that life is necessarily richer than the regulatory model they use, nor that the outcome of such cybernetics is wholly predictable. That outcome is the collapse of control; the mentors are content to say that an “inadequate personality” simply “could not cope”. This whole picture looks quite different from the systemic point of view. Instead of Oneself battling more or less successfully with other surrounding entities, we see Oneself as a complex high-variety system interpenetrating other high-variety systems and finding a balance (as it were, a thermodynamic equilibrium) thanks to an intrinsic regulator competent to match variety with

variety, and to sponge it up (as a cooler body absorbs heat from a hotter body). Social anthropology attests to the intricate, high-variety interactions whereby societary stability is maintained in what we please to call primitive societies. The fact is that in disrupting these elaborate patterns, and by treating Oneself as a classifiable entity and a so-many-bit computer entry, the Western culture has denatured the regulators that are intrinsic to interpenetrative social relationships, and must perforce use other means to re-establish the variety equations that themselves determine social stability. Rather than pumping variety back into the system of Oneselves, so that equilibrium may be newly established, our managerial techniques systematically remove variety from each Oneself. The tacit expectation is that balances will be more easily found if there is less scope for individual manoeuvre – but this is not the case. The system of Oneselves needs individualistic high variety in order to move smoothly towards calm following perturbation: to foster the ability to learn as a societary unit, and to generate experiment – without which social adaptation is impossible. A tripartite selfhood I have made this discursive start in order to provoke disquiet with received ideas, and in order to give a preliminary indication of the new approach. Let me now try to make the core of the argument quite explicit. Let us think of a human being – not just as flesh and bone, not just as a psyche, not just as a social unit, but as an entire and interactive system. This is the registered person that I have called Oneself. According to received wisdom, as I reflected it, Oneself is a kind of billiard ball, that follows a path from birth to death through time; s/he continually learns from this experience, stores memories, and makes plans for the route into the near and distant future. Oneself is in control of oneself, therefore responsible and accountable for the whole journey (although it is conceded that good or bad luck may be encountered on the way). If abnormal behavioural states occur, these can be classified as (for instance) illegal, or pathological, or perverse – but the person may be restored to normality by penology, or by psychiatry, or by some therapy of ethics. Replace this whole picture, if you will, with another. The sphere of the billiard ball grows shimmering and indistinct: it is a high-variety, complex, probabilistic system that just maintains stability – and therefore its integrity – by virtue of an organizing principle. This certainly exists, in that we can talk about identity; but it is not an item susceptible to dissection. The organizing principle is that set of relationships which invariably determines that this is Oneself, and not some other self. Evidently the nucleic acids are implicated in this, evidently the neural nets and endocrines . . . What matters is not the recipe, but the recognition of an intrinsic regulating mechanism that holds everything together. Self-consciousness may be defined, though not explained,

Man in a garrulous silence 815

K 33,3/4

816

as the convergence of the functions that this mechanism performs upon themselves. I need a name for this intrinsic regulator that holds invariant the set of internal relationships which maintain the identity of Oneself. I shall call it cybercyte, from the Greek word for steersman (that is already in wide use to name the science of cybernetics), and the suffix meaning “cell” (that is used in many scientific combinatorial terms). Let us think of the cybercyte as a glowing presence inside the shimmering sphere that is Oneself. The concept of variety, which I have already introduced as the measure of the number of possible states of a system, is central to cybernetic theory. So is Ashby’s Law, which (I remind you) is the law of requisite variety. Now in the present case we are faced with the fact that the stability and integrity and adaptive conduct of Oneself requires that the variety of the cybercyte is at least equal to the number of states that the more commonplace Oneself can possibly generate in its interaction with the totality of existence. In some ways it is truistically evident that the equation holds: for the whole physical body is innervated to facilitate sensation to this end, there are chemical responders, and homeostats of every sort abound. In other ways, the variety equation is always in question. We do not know if the cybercyte commands requisite variety to enable Oneself to cope with some particular threat – if we do not know what the threat is, or even if there is one. But supposing a threat to exist that is of higher variety than whatever Oneself does exist, then an encounter with it would necessarily generate a response variety greater than the limit of the cybercyte to mediate. Moreover, we have to take account of the likelihood that this threat may interact with Oneself in ways for which physical sensors have not developed by natural selection. Let us note that if all this goes for threats, it goes for opportunities too. In contemplating the first cybernetic consequences of our newly evolving picture of Oneself, we may note with excitement that incipient instability is natural and not pathological – that it constitutes quite simply a variety overload relative to the variety of the cybercyte. This means (it is not an implication but a descriptive fact) that the intrinsic regulator breaks down. Then this is in turn a completely functional response, whereas (as I mentioned earlier) received models treat it as dysfunctional. So education should direct itself to enhancing regulatory variety in the cybercyte, rather than to delimiting given world-variety by reductionist taxonomy as it mostly does. As to psychological medicine, “give me lithium or give me death” prescriptions are active inhibitors of variety in the cybercyte, and must therefore be cybernetically counterproductive in the long run. Attempts to understand higher states of consciousness are doomed to failure by research practices that monitor physical epiphenomena of the cybercyte that must be variety-transcended if such states exist. And so on: these are simply briefly identified topics for which established approaches, therapies and

developmental thrusts seem to be not only mistaken, but diametrically opposed to their objective treatment. It is natural to suspect the validity of such scandalous conclusions, reached so soon. Why are they so novel; why diametrically opposed? They are not novel at all to Eastern culture, in which the handling of Oneself is all about the enhancement of variety in the cybercyte, and the explicit need to match variety with variety in maintaining calm. The first is the fruit of satori, the flash of transcendent awareness, and the second describes the contemplative regimen in cybernetic terms. As to diametric opposition, rather than the expression of mere difference, it is a manifestation of the validity of our first results. For the models used in Western culture do not pump variety into the disequation of regulatory variety haphazardly – they pump it into the wrong side, raising variety that is already intractable (as in computerized management systems) and lowering cybercytic variety already in overload (by drugs, soporific television, and simple-minded nostrums of control). But there is more to come. The Oneself so far described is differently described, but it remains so far a billiard ball with a future trajectory marked out upon the field of play. In our new picture, this will not be so. Let us think of the shimmering spheres, in which the glowing cybercyte is embedded, as embedded in its turn in a new sphere – and one of possibly limitless diameter. You will correctly think this stands for the environment of Oneself as depicted up to now. But it is not the physical environment of human potential that surrounds Oneself. Its substance is immanent variety, the richness of the states one might adopt; it is the nutrient medium of growth. If we enhance our variety, we do so at the boundaries of the shimmering sphere; and obviously, by Ashby’s Law again, the variety that the cybercyte disposes must equivalently rise. So, if a non-golfer takes up golf, he has realized some of his human potential. The extent to which this is done successfully is measured by the matching variety engendered in the cybercyte, enabling new regulatory routines to develop. If not, there is an absence of control – and this is dangerous. The golfer may give himself a hernia, break a window with an ill-directed drive, or hit his opponent with his ungoverned club. In the realms of quietude, it is familiar but quite real that people should realize potential in terms of scholarship, the arts, and hobbies: in every case we see how the cybercyte’s variety must parallel developmental growth – otherwise more is undertaken or experienced than can be mastered and the danger at the least is depleting self-indulgence. But if we talk instead of realizing the potential of psychological states themselves, there are more serious dangers: emotions may boil over, or the release of inhibition (which was of course a variety attenuator) may not have matched regulatory variety in the cybercyte. We may see this in T-group training, or in the cult of “transcendental meditation”, since little attention is given in either protocol to the subsequent containment of

Man in a garrulous silence 817

K 33,3/4

818

expanded variety – which is seen as good-in-itself, the regulatory component being an implicit part of the package. But it is not; and I have seen the cybercyte overwhelmed by such variety enhancement. The realization of human potential in the growth of psychic awareness, to move to a final and yet more contentious example, carries with it well-nigh lethal dangers: the cybercytic variety must be grown almost ahead of development, in my opinion. This is why monasticism, in the West as well as the East, nominates a personal teacher to the novice or the neophyte: it is an abbott’s application of the law of requisite variety, and carries its own moral for readers of the glossy manuals of mysticism. How shall we rectify this largest and perhaps limitless sphere of Oneself, and why should we? It seems to me that – vast as it may be – the potential for Oneself is an individual and private universe of immanent variety. It belongs to Oneself, because it stands for the completion of Oneself – and not to someone else. For this, the encompassing sphere, I borrow Aristotle’s word entelechy: “a condition in which a potentiality has become an actuality”, or “the realization or complete expression of some function” (according to the Oxford English Dictionary). For Oneself to fill the sphere of the entelechy, by metabolizing all possible states, is, to speak in Eastern terms again, nirvana – which really means extinction for the want of further fuel. This embedment of three spheres to constitute a picture of Oneself is subject, as I have tried to show, to basic cybernetic laws: in the matching of variety in the cybercyte, a matching that must ultimately extend to embrace the variety of the entelechy itself. This development, and no other, could be taken as the purpose of human existence by many cultures, especially in the East. For others, more materially inclined, the well-being of this whole Oneself as feeling comfortable, as being in general happy rather than either miserable or euphoric, might suffice as a purpose. For this condition, I use another of Aristotle’s words: eudemony. The ambition to maintain eudemony at the level of satisfying needs and simple desires untamed by unscrupulous avarice, might express a middle-class liberal attitude in the West. The community consequence So far I have been speaking quite specifically about the human individual, that viable system, whereas in my introduction, I deliberately used the term Oneself to refer to any viable system at all. And this is the moment when the cybernetic formulation yields a methodological dividend. For the laws that govern viability are invariant; and I have elsewhere shown that they apply to cohesive social systems that survive as distinguishable identities. They are mathematical recursions of each other, connected by a quantitative Law of Cohesion. It is impossible to enter into the demonstrations here; and I hasten to emphasize that the existence of structural invariances across all viable systems in no way minimizes the uniqueness of individuals (all of whom, however, have

two arms and two legs) or of cultures (all of whom, however, exhibit (say) sibling relationships and the regulation of variety proliferation by taboos). But if there are invariant laws, then the descriptive language that has just been invented in which to talk about Oneself the person may be used to illuminate Oneself the social aggregate. And so it proves. If we consider a community, which might range in size and culture from the family, to the village, through cities to nations, and include the species humankind itself, we find cohesive systems that are large, complex and probabilistic – yet have a recognizable and surviving identity. This is a shimmering sphere, its boundaries uncertain, its equilibrium unstable. The cybercyte of this community is its management system. Again, I make obeisance to every anthropological distinction. I do not speak individually about hundreds of regulatory devices – of electronic data processing systems, of charters of rights, of taxation laws, or of population control by cannibalism – but I do speak of them all. Because I know that if there should be no cybercyte deploying requisite variety to match the variety proliferated by the social environment of the societary Oneself, and if this regulatory variety should not be deployed in sympathy with the laws of viability, then the community concerned cannot indefinitely survive. Indeed, if the social environment is highly volatile, survival will be alarmingly short-lived. Suppose you complain that this same description cannot possibly hold for both liberal and totalitarian societies, I reply that it can and does. The societary distinctions that we rightly hold in awe, distinctions for which many people have been and are prepared to risk their liberty and lives, are unique to the myself-Oneself of the community. They represent the methods by which variety is proliferated, channelled, amplified, attenuated, and constrained; but variety has to be engineered with in all these ways by some method in any community, any viable system whatsoever – simply to survive. Second, we observe the community embedded in its own entelechy. And again the cultural distinctions emerge. This time they surely generate a demand for the recognition of community-ness, or of nationhood, or of humanity itself – depending on the level of recursion that you choose. Suddenly our patience with the preparatory thinking and vocabulary building is able to produce rapid results. Consider: . imperialism, or colonialism, or (as we have it today) the hegemony of the superpowers, involves planning in the context of the wrong entelechy from the standpoint of those planned for; . assumptions characteristically made by the dominant nations as to appropriate criteria of progress not only tend (as many have realized) to ignore the indigenous culture: they are geared to the wrong recursion; . the cybercyte of the subordinate community is soon conditioned to rely on surplus regulatory variety from the superior power, and is thereby robbed of requisite variety (this is the condition of dependency);

Man in a garrulous silence 819

K 33,3/4

820

.

the superior cybercyte is out of context, and may therefore be seen to be pumping variety into the wrong side of the regulatory equation (creating bureaucracy).

These are the reasons why the good intentions of millions of overprivileged people towards their underprivileged brethren are not translated into effective action, even in those relatively rare circumstances where the political will exists to help (as in the case of famine) and is not thwarted by covert political action undertaken from ideological premisses. Moreover, we can now see why it is that we can do no more than foster the planning potential in other people or in disadvantaged nations and communities. To offer them methodologies of planning that do not work is cynical and insulting – but these are the only planning methodologies that we have discovered. To offer them a plan content that defines progress in our terms uses a false measure of eudemony expressed within an alien entelechy. Thus are we easily involved in a case of mistaken identity – or at least an irrelevant one. As Marlowe wrote: “But that was in another country: and besides, the wench is dead”. Reflections: the schematic . . . the mandala The schematic diagram of Selfhood that I have provided (Figure 1) is, I suggest, an improvement on the rectangular box provided by the draughtsmen of organization charts. It depicts the shimmering system and its three embedded components; it suggests the tenuous equilibrium of the human condition; it owes more to the East than to the West in its conception.

Figure 1. Schematic framework for discussion of the social cybernetics of the human condition

The social self (depicted by the heavy circle) contains the cybercyte, or organizing principle, which must exert the grip that can hold its explosive variety together. To do this, the grip must contain a model of the self, that maps its creative systems, and matches its number of potential states. There exist cybernetic theorems that demonstrate these and other obligatory requirements of a regulatory system; but it is in this context I think self-evident that if states arise in the self which cannot be modelled or cannot be contained, then the grip is not certain – and might easily be altogether lost. There are two basic approaches, which may be combined in various proportions to form a mixed strategy, to the self-regulatory problem posed by these risks. Either artificial limits must be placed on the proliferation of variety that Oneself entertains, or the quality and range of the grip must be increased. The culture heavily favours the former choice, as being more prudential. Thus by “avoiding the occasions of sin”, variety proliferation of the life-experience is restricted, and the grip is maintained by embracing simple rules of conduct. These rapidly break down if the life-experience becomes too rich. What is easily understood in this ethical example applies to other regulatory modes as well. In general, less knowledge, less profound and less highly organized, is required to handle eventualities circumscribed by a hum-drum life-style, a common place perception, and an unwakened or a blunted sensibility. Now we may note that it is usual for social norms to prescribe both the acceptable range of behaviour (and therefore of experience) and the variety reducers embodied in the grip that normal people are considered able to exert. Thus the regulatory system is self-referential, and information-closed. The entropy of self may be expected to increase, less free-energy being available for conversion to creative ends as time goes on. Thus do we observe the facts. In order to enhance this convenient societary phenomenon, the culture takes good care to reinforce the mechnisms involved. It offers prescriptions as to the use of the body in exercise and diet, as to the use of the mind in what it is seemly to think about or to know, as to social mores in behaviour, with massive technological amplification, and massive regulatory feedback. Millions of people in advanced societies are glued to transistor radio sets that are mobile, to television programmes that are ubiquitous in homes, hotels and bars alike. And of course the feedback constantly reinforces low-variety norms, through the convergence of conservative response as monitored by producers and presenters whose personal interest lies in fostering that readily-manipulated consensus of the lowest common denominator. I think that we should note that this process already bids fair to operate as tyrannically in so-called democracies as does regulation by fear and coercion in openly totalitarian regimes – where many people do at least realize that their freedom has been circumscribed. In the West we have traded freedom for the thought that we are free; we behave with sufficient conformity that we pose no threat to those who hold power, then boast that there are no guns to be seen on

Man in a garrulous silence 821

K 33,3/4

822

street-corners. But the human condition is delimited all right; and even loving concern is squeezed through the variety mangle, so that people in the United States use up their rations in constant admonitions to “have a nice day” and to “enjoy your life” – gestures as empty as the vacuum of feeling from which they spring. As to the fullness of the entelechy, betokened by the larger constructions in the diagram, we have to realize that whatever is done to metabolize potential must remain within the modelling capacity and variety-generating potency of the internal regulator, the grip. That is why the picture shows the grip so tentatively exploring the space of the fullness that potentially informs the surrounding void. There is very little argument as to the direction in which the variety equation is balanced now. The richness of the entelechy is cut back, stifled or impoverished in order to match the grasp of the grip; and no mechanism that would enlarge the grip has much societary support. In science, phenomena that do not conform to the current model of the universe that science underwrites are held to be impossible or illusory. (Lord Kelvin, as president of the Royal Society (1890-95), declared that X-rays would prove to be a hoax, and that heavier-than-air flight was impossible.) In religion of whatever adherence, whatever goes beyond approved dogma is unorthodox (a word which actually means “wrong”), and people are anathematized for heresy (a word which simply means “choice”). And so we could continue. One consequence is evident. Society ill tolerates anyone to whom comes an unnerving experience, generated from his own potential but beyond his present grip; s/he may well be locked up, shunned at the best. Let us notice too that institutions actually dedicated to variety engineering in favour of enlargement and enrichment of the grip, often house zealots who have only one recipe for personal advancement. The arrogance with which they preach humility gives away the poverty of variety in the approach. The self with which our picture deals is a tenuous systemic network of existing and immanent probabilities continuously generated by a high-variety self-referential framework of identity. Thus when we elaborate the picture to that of an organization chart or a community, we do not find a set of discrete boxes inter-connected by arrowed lines. What we do find I cannot even draw. Think of the figure as revolving on a vertical axis to make a spinning or shimmering sphere, and encourage several of these selfhoods to approach each other. Despite the spinning, the peripheral constructions representing the entelechy are not impermeable – because they stand for probabilistic states; thus the pictures interpenetrate. Then the effects upon variety in terms of opening and closing possible state for each member of the group are vast in number and largely unpredictable in effect. It is just such a concatenation of selves that constitutes a human group; and it is the miniature picture of this group that I ask you to insert in the main

(heavy-lined) circle of the basic diagram in order to reach a perception of societary regulation at the next higher level of recursion. The regulatory properties of this new picture are the same as before. There is the shimmering sphere that we call the community, consisting of interpenetrative selves. Embedded in this is its grip; the whole is embedded in the fullness of its social entelechy. This is the language and the picture of the concept of human identity that is offered here. Part II: A public statement A basic stance The human being looks like a fixed, closed and compartmentalized entity. In the Western culture s/he is increasingly treated as such: as entity permitted a limited number of characteristics, a so-many-bit entry on a magnetic disc. But a person, from any standpoint, is neccessarily a system: one that maintains identity through a set of relationships between its internal elements, and has self-consciousness of that identity through the self-references of its internal activity. Moreover, this person-system interacts with others to form social units (and ultimately the totality of humankind) which are themselves systems, similarly defined. In their societary roles, people do not so much bounce off each other like billiard balls, as they may appear to do, as interpenetrate systemically – sharing aspects of their personalities in less or greater richness. The dominant culture typifies the problems that humanity faces with adjectives that deny this systemic wholeness of the human condition. It calls them political or social or economic, environmental or educational or medical, psychological or physical or moral, and so on. But our problems all share one underlying characteristic which all dynamic systems exhibit: it is the problem of regulation in favour of a flexible stability. The billiard-ball approach to regulation means containing free expression, circumscribing free interaction, and in general constraining the human potential. By such means it is hoped to make people behave themselves properly in an orderly society on a planet with a future. It does not work, as we may observe; and this is because there is no agreement as to what behaviour counts as “proper”, no agreement about the point where “orderliness” constitutes straightforward tyranny, no agreement as to the destiny of humankind. The reduction in complexity which all this involves is counter-productive, because it interferes with the natural laws that determine stability in self-regulating systems. In purblind recognition of the mistake, complexity is pumped back into the situation at the wrong level of aggregation (namely at the top instead of the bottom), in the wrong forms (namely the legalistic and the bureaucratic) and with the wrong objectives (namely to forecast the future and to stop it happening). Moreover, if the classification system within which all of this happens were availing to the plight of humankind, then by now some of our problems would surely be within sight of solution – but they are not.

Man in a garrulous silence 823

K 33,3/4

824

The problems become more massive and the world more unstable through whatever pair of spectacles one uses as a filter. The rules applied Now the rules to which I refer have been detected by cybernetic scholarship as subsisting within the natural domain. They are not theories or prescriptions, but statements about how viable systems work. The massive political questions that face humankind are not about whether alternative rules should be adopted – for that is impossible – but about the manipulation of the regulatory rules that exist. By determining a world outlook through education, control of movement and opportunity, and governance of the communications media, the variety of the human ambience is currently controlled, and the variety of response attenuated. This reduces the variety required in maintaining the grip to the level at which it can be imposed upon people, communities and nations from outside, instead of being autonomously self-generating. This is the present day fate of humanity, and perhaps it always was. By this token, it does not seem to matter very much by what name oppression goes. Specifically, it is high time that the opulent nations recognized that the conditioning of society towards the ever-increasing consumption of scarce resources, the amassing of consumer durables, and the explosive though fictitious need for services, does not constitute freedom. It is a mode of existence as oppressive to the human condition as is penury; it is another kind of serfdom. Most certainly its selfishness condemns most of the world to a life of physical squalor – but its own spiritual life is rendered squalid too. Wise men have long taught these truths. But in our lifetime we have put faith in science and technology to satisfy economic equations that literally have no solutions, because they are not founded in planetary homeostasis. Twenty years ago it became clear that the universal spread of Western-style consumption is impossible, because the Earth’s resources are finite. This has affected the sense of security and even of decency of some thinking folk, but it has influenced policy and strategy not one whit. Today it is clear that our policies and strategies involve massive interference with the regulatory systems of nature on a planetary scale, so that every aspect of our affairs is incipiently unstable. This means to say that the grip exerted by the species on itself and on its habitat is almost lost – in disobedience to the regulatory rules that have been listed. Inducing change To do anything about this, we shall need to change our perceptions and our policies. It is not a matter of modifying these, nor of pursuing reckless courses more relentlessly in the hope that this will make them effective and safe. New initiatives are needed that embody principles learned from cultures other than the dominant culture of the West, from philosophies other than materialism,

from methodologies other than reductionism. Moreover, we have loved our technology not wisely but too well . . . The strategy for change advocated here is based on a specially constructed concept of selfhood. This says first that the recognized self – whether a person, a community, a nation or the species – has an inward component constituting the autonomous regulatory machinery that allows oneself to exert a grip on itself. The over-riding rules that govern the effectiveness of this grip are that the regulatory process must be able to generate as many states as the self disposes in responding to its ambience, and to minimize this variety by detection patterns which can be stored for recognition later. These principles are fundamental to coherent behaviour, to learning, and to adaptation. The concept of selfhood secondly says that the recognized self exists within a potential self, the realization of which constitutes its fulfilment. The rules that govern the extension of the recognized self in this way include the exercise of perceptual powers to encompass greater variety, training in thinking itself as a divergent process, and above and beyond all of that the extension of the grip as already defined to maintain cohesiveness of the self under the pressure of its variety enhancement. The strategy for change is thus an internal affair, because it is concerned with volitional change in the self. This is true at every recursive level of selfhood: the only effective route to the survival, never mind the enhancement, of life comes with enlightenment for the individual, autonomous development for the community, and collaborative surplus for the species. In each case the tripartite concept of selfhood points to the rules that govern the grip, the rules that release potential, and the stability of the recognized self as a cohesive entity that conforms to the rules that govern its own identity. It is an irony indeed that the very technology that was (and in many quarters still is) supposed to handle the limitations on resources should be so largely responsible for inducing this threatening instability – because of its ignorance of basic cybernetic processes and its inherent propensity, when wielded from economic greed, to amplify hugely the positive feedback that is so destabilizing. Hence planning addressed simply to technological transfer is likely to prove counter-availing in the extreme. Yet this is the basis of current endeavours, ostensibly to rectify inequalities among the peoples of the Earth, but actually to submit them to the same materialistic ethic and the same fate that has overtaken the rich nations already. If it appears that this must lead us to advocate the withholding of technology from developing nations, that would be absurd. The discoveries of the human mind are the birthright of us all, as they would very quickly tell us. They would tell us, unfortunately, because their leaders, too, have been trained in the same error, and in the same style, and with the same expectations as leaders of the West, and because these people have themselves started to undergo the blandishments of consumerism.

Man in a garrulous silence 825

K 33,3/4

826

It is a new kind of leader of whom the world stands in need. And if there be any hope remaining for the human race it probably resides in the fact that half the planet’s population is a bare 20 years old. Here is the source of a new leadership, not yet perverted by greed nor contaminated by idiotic ideals that have manifestly poised our species trembling on the brink of an extinction that is possible and even likely within the 18 years that remain before the magical year 2000 is attained. We have first, and imminently, to get through George Orwell’s year of 1984 . . . New dimensions for planning This argument points with urgency to the only plan that has a chance to work. It would be founded in new communities of an experimental kind which really need to be as large as whole nations, since time may be so short. We have had enough of talk, of “research”, of lugubrious reports – and of the crass inaction that all of this promotes. If nothing is to be achieved, then create an International Institute that will enshrine its high hopes in an expensive mausoleum and glossy publications. If nothing is to be achieved, then convene the world’s statesmen to determine what to do. If nothing is to be achieved, then avidly train young people to pursue our feckless ways with greater zeal – just as if we knew how society should be organized and run. All of these things we have done and do, and the results we know. The mark of a fresh deparure will surely be the negation of these tried and untrue approaches. We need to devise communities in which the problems are tackled by those individuals who demonstrably succeed at the personal level of recursion in grip and in growth. This is the human dimension of advance, and its emphasis will be on the rounded individual and the balanced community. Wider rather than increasingly specialized knowledge and skills will be sought; a more integrated experience of life than the divisions of work and play, science and art, and the role-playing of different ages and sexes currently allowed will be fostered. Man’s way of life, however, is very much a function of his tools – those extensions of the human physiology that amplify all its attributes. Then we should distinguish the technological dimension of advance. All the emphasis here is on problem solving. What are the best tools to use on current problems and in current conditions? That is the core question; and have done with constituting the folk with the problem of a captive market for exploitation, or a dependent recipient of outdated or unwanted machinery and expertise. Third, it is vital to pick out the ethical dimension of advance. Nothing can be salvaged from the current mess of corruption among officialdom, both public and private, that is worldwide without a committed stand. We cannot legislate against the frailty of human nature, nor should we try to make the joyful dour. It is better to work against corruption as a set of acts which denatures our chosen regulators, and induces instability.

Here are three chosen dimensions of advance, the human, the technological and the ethical, which are vectors always passing through the moment now at right-angles to each other. Each is a condition of existence for humankind, but none is defined in terms of the other two. Here is the necessary framework within which our problems are both posed and to be solved. It needs to be established wherever people are, at each level of recursion beginning with their own and extending to the limits of their responsibility. Note that this can be done, everywhere and within whatever ideological constraints may happen to obtain. There is no time left in which to resolve the ideological conflicts that debar effective international planning. There is no sense left in talking about activities such as education, health, welfare, and the transfer of technology as if they existed in a vacuum, and could be pontificated about in their own right. All these things should be discussed in terms of the problems that people have, in their own here-and-now, according to their own wishes. I have proposed the recognition of three dimensions of advance within three odes of selfhood: planning then becomes a matter of obtaining recognition for these frameworks and then facilitating choice. Choice is the stuff of freedom. Choice is the commodity that ultimately the rich and powerful will not share. In Chile, I worked for Salvador Allende – a President of democratic choice. The Chileans were setting out to enhance the lives of the majority, and to provide those folk with further choice – beginning with the availability of food and shelter and welfare, and the purchasing power to choose. We mobilized technology in the solution of problems in the social economy, seeking once again to offer choice, under strong moral leadership that wished to outlaw corruption. It is a matter of record now as to how the Chileans were robbed of choice in each regard at the instance of another country. Such manipulations continue, worldwide. The challenge to humankind is this: . will some nation, somewhere, be permitted to choose a new course, and have its developments facilitated and not aborted by the outside world? No more than this is needed. Many insights and discoveries and abilities await their turn to be used; the enthusiasm and hopefulness of the young listen for the call to action; and for those who see the cosmos as bonded together with love, a bold new experiment would provide a focus for their yearning. Author’s note: The same schematic mandala depicting Selfhood was used in the keynote address to the Seventh European Meeting on Cybernetics and Systems Research in Vienna, April 1984. The address is recorded in the “Recursions of power” chapter in: Trappl, R. (Ed.), Power, Autonomy, Utopia, Plenum Press, New York, 1986.

Man in a garrulous silence 827

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

K 33,3/4

828

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

Ten pints of Beer The rationale of Stafford Beer’s cybernetic books (1959-94) Discussion with Stafford Beer Keywords Cybernetics, Organizational analysis, Management science Abstract Details a discussion with Stafford Beer about the rationale of his cybernetic books published in the twentieth century (1959-94). Includes a dialogue concerning his contributed works: Cybernetics and Management; Decision and Control; Management Science; Designing Freedom; Platform for Change; The Heart of Enterprise; Diagnosing the System for Organisations; and Beyond Dispute.Concludes with a perspective of some of his important early work and some unpublished contributions which have been collected into one book.

Preface A discussion with Stafford Beer about the rationale of his cybernetic books published in the twentieth century (Beer, 1959, 1966, 1968, 1972, 1974, 1975, 1979, 1985, 1994b) proved to be both illuminating and valuable. Since Professor Beer is the undisputed founder of Management Cybernetics it provided an opportunity to discuss his contributed works, to date, as a continuous whole. Indeed, it provided an essential background to the understanding of his cybernetic writings. This is a significant moment in history when we are able to examine the contribution made by the cybernetics community and look forward to a new and fruitful millennium. Consequently, we are indeed fortunate that such a distinguished author is able to help us understand his work and give us such a valuable insight into his thoughts. Nine books by Stafford Beer, published between 1959-94, are discussed with the author. Finally, we are able to put into perspective some of his important early work and some which had not been printed before, which has been collected into one book (Beer, 1994a). 1. Cybernetics and Management Cybernetics and Management was published (1959) by English Universities Press, London and John Wiley, New York. It was then reprinted in 1960 and 1965. A new edition, complete with a new chapter was published in 1967, to be reprinted in 1970, 1971, 1973 etc. It was translated into Spanish, Russian, German, Czech, Polish, Dutch, Italian, Portuguese, French and Japanese (see Beer, 1959).

Kybernetes Vol. 33 No. 3/4, 2004 pp. 828-842 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523724

The foundation of cybernetics as a science by Norbert Wiener, and characterized by him as “control and communication in the animal and the machine”, was marked by his publication of Cybernetics in 1948. Stafford Beer This discussion with the Kybernetes Editorial Team took place in January 2000, and was first published in: Kybernetes, Vol. 29 No. 5/6, 2000, pp. 558-572. q MCB University Press, 0368-492X.

encountered the book in 1950, and at once recognized its direct relevance to Ten pints of Beer his own work in management science in the UK steel industry. Having written and lectured extensively about the new field, he presented three plenary addresses to the International Association for Cybernetics at the founding assembly in Namur, Belgium. As an outcome, he was invited in 1957 to give a series of lectures and seminars by the Royal Swedish Academy of 829 Engineering Sciences in Stockholm and Uppsala. He was urged by the sponsors, who presented him with the Silver Medal of the Academy, to incorporate these lectures into a book. Cybernetics and Management was the result, and it was first published in 1959. The context of “system” was set against the reductive processes that have dominated our culture. From the managerial point of view, systems were seen as being essentially in control or not in control – these turning out to be functions of perception, conditioned by the ubiquity of feedback. The description of complex situations as black boxes, and the notion that a purposive system is defined by the black-box output itself (rather than by wishes or intentions), was the origin of Stafford Beer’s later celebrated dictum “the purpose of a system is what it does”. The role of theoretical logic as conditioning the world, and leading to a cybernetic account of management itself was revolutionary. Above all, the book was the first to argue the possibility of designing an organization scientifically to be a learning, adaptive and evolving system. The book was selected as the management book of the year by the prestigious reviewers of The Sunday Times, and Norbert Wiener himself came to refer to Stafford Beer as “the father of managerial cybernetics”. Despite adding a new chapter in 1967, to comment on the progress of automation, the basic book has never been amended – because it offered a new perspective on the history and philosophy of science, which are themselves matters of record. The book was constantly reprinted over a period of more than 20 years, after which it was lost to view in a series of publishers’ takeovers and reorganizations. That new perspective was essentially an appeal for holism, to be achieved through the study of total systems, as distinct from the reductionism which already pervaded the scientific method. Forty years later, reductionism is triumphant everywhere, from its base in academia, through science and medicine, to the social sciences. Those seeking to understand the work of Stafford Beer as a whole would do well to begin with his own first book. In a preface to Cybernetics and Management, the author pointed out that the ultimate source of energy in the hydrogen-helium fusion had been discovered and put to work. It needed to be balanced by equivalent advances in the ultimate source of control – that is, in the cybernetics of natural processes, especially in the evolution of the nervous system and brain itself. Forty years later that ambition remains compelling.

K 33,3/4

2. Decision and Control

830

Intended as a definitive statement about holistic science and problem solving, this book has been continuously in print for more than 30 years. It was awarded the Lanchester Prize by the Operations Research Society of America as “the outstanding contribution to the literature of management science in the world” for 1966. From the early days in World War II, the objective of Operational Research was to address intractable problems, and to increase the probability that decisions taken under conditions of extreme uncertainty would be effective. The approach was formalized by Stafford Beer at the invitation of the Operational Research Society when he first joined the Council, and was officially adopted by the Society. The definition was printed in the front of every issue of the Society’s journal for a great many years. It reads:

Decision and Control – The Meaning of Operational Research and Management Cybernetics was published by John Wiley in 1966 and then reprinted in 1967, 1970, 1971, 1974, 1978, 1988 etc. It was translated into Spanish. In 1994 a paperback version was published (see Beer, 1966).

Operational Research is the attack of modern science on complex problems arising in the direction and management of large systems of men, machines, materials and money in industry, business, government and defence. Its distinctive approach is to develop a scientific model of the system, incorporating measurements of factors such as chance and risk, with which to predict and compare the outcomes of alternative decisions, strategies or controls. The purpose is to help management to determine its policy and actions scientifically.

The purpose of the book was to elucidate and illustrate these propositions in the management context. The preface points out that Stafford Beer had already spent more than 20 years working in the field – all of it in active leadership of operational research in the biggest steel company in Europe. The multidisciplinary professional staff reached 70, and by 1956 had installed the first computer ever dedicated to management science. The outcome was not a product of academic research. Indeed, his first academic appointment followed publication of Decision and Control, and was as a full professor at the Manchester Business School, where he remained for 24 years. He has held the same title in many universities around the world. All of his teaching work has always been on a part-time basis, however, because he has remained a senior manager or consultant to this day. Stafford Beer still believes that his problem-orientation is vital to society, and that his advocacy of cybernetics offers the most powerful approach to multidisciplinary model building available. That is why the book remains fresh. In the meantime, academic demands for curriculum, for a list of techniques that lend themselves to ritual examination, have turned OR into something approximating to applied mathematics, and removed from the creative solution of problems. Decision and Control does not regard all models as mathematical, although mathematics may serve as a powerful language in

discussing them. He has used models ranging from the A of Apiary to the Z of Ten pints of Beer Zoology. In the book’s index, under the entry listing types of model, there are 31 entries. That leaves the problem of the appropriate methodology for selecting and handling models. In Chapter 6 “About models” (pp. 95-119) the foundation is laid for a distinctive approach to the epistemology of science, based on the 831 holistic nature of systems. This is the book’s major contribution to scholarship. The more rigorous account had already appeared in the prestigious journal. Nature (No. 4968, January 16, 1965) presented as the Stephenson Lecture for 1964, under the title “The world, the flesh and the metal”. 3. Management Science Management Science: The Business Use of Operational Research, was published by Aldous Books UK and USA, in hard cover and as a paperback (1968). It was translated into German, French, Swedish, Japanese, Finnish, Dutch and Norwegian (see Beer, 1968).

Decision and Control was a weighty tome, although every effort had been made to make the work digestible. Other publishers thought that there was a serious need to convey the underlying message in a more readily accessible form – in a book limited to 200 pages, and decorated with plenty of diagrams and pictures. The book was to be published simultaneously in many languages, based on a text of uniform layout in them all. The standpoint of Management Science is expressed like this. To say that there can be a science of management is not to deny that management is an art. At the time, controversies raged in the business press over this issue. It is still true today that heavily intuitive entrepreneurs tend to disregard the scientific possibilities of systemic synthesis. But the man with the genius for designing buildings is no less an artist because he is a competent architect. It is a good thing for us all that he is. Second, we should not confuse the contention that there can be a systematic knowledge of management, based on systems theory, with the notion that there are particular techniques used by managers which can be taught and applied to situations like a poultice. This is an argument already familiar to readers of Decision and Control, and annoyance with the way in which problem solving was already giving way to an obsession with pre-packaged technique is evident. “This is a mistake which is sometimes made in very high quarters”, declares the dust-jacket of Management Science. Once again, as we saw in Decision and Control, there is an emphasis on the nature of models, and the idea that models can be made larger or smaller – by analogy with optical resolution – does not invalidate their utility. Another preoccupation of Management Science was the virtually total ignorance of the theory of probability at the time the book was written. Managers considered that figures were either right or wrong, influenced no doubt by accountancy and especially bookkeeping. Estimates were at best

K 33,3/4

832

inadequate figures that ought to be “put right”, and probabilities were treated as equivalent to guesses. The whole scientific domain of likelihood was (and unhappily largely remains) a closed book to managers. Therefore an attempt was made to introduce some basic statistical theory in an interesting way. For example, over many years the author mystified audiences by considering the probabilities that two people in the room shared the same birthday. In fact, the odds are even when only 23 people are present. In a sizable audience of say 90 people, the chance of a shared birthday is not one-in-four, as most people think, but 99.99 percent – virtual certainty. We know that managers ought to be able to read a balance-sheet, to understand costings, to consider personnel factors, to peruse market research, to offer leadership and so forth, and obviously these are particular skills which can be communicated about all such matters. But the claim that there is a management science cannot be justified as if it were simply a collection of topics in that way. The issue is about complex, probabilistic systems, considered as a whole. The science of a subject is always about its very nature, not about virtuosity. 4. Brain of the Firm Brain of the Firm. The first edition was published (1972) by Allan Lane, the Penguin Press, London; Herder and Herder, USA. Translations were made into German, Italian, Swedish and French. A much extended second edition was published by John Wiley, London and New York in 1981 and subsequently reprinted in 1986, 1988, etc. It was translated into Russian and in 1994 appeared as a paperback (see Beer, 1972).

This book is an account of the firm – or any other cohesive organization – based on insights derived from the study of the human nervous system. Such understanding of neurophysiology as is required is explained in the text. Brain and management structures are elucidated and continuously compared, and a theory of effective organization is thereby evolved. This process should not be thought of as providing an “analogy”, but as the pursuit of fundamental principles whereby self-regulatory viable systems are necessarily constructed. The model distinguishes five subsystems that have particular systemic roles, and can be measured according to a particular protocol that assesses numbers of relative potential states. The five subsystems continuously interact to maintain an overall balance, known to biology as homeostasis, whereby we customarily say that the total organization is “in control”. This means that the flux of significant variables achieving the balance are all maintained within limits that do not stress any one of them. The entire regulatory process is intended to ensure survival in both the short and the long term, through processes of learning, adaptation and evolution. The five subsystems cannot be isolated from each other, and attempts in the literature to identify them separately with managerial names are ill-conceived. Brain of the Firm is the first of a trilogy of books devoted to explaining what is now known as the Viable System Model – or VSM. Using the epistemology

described in Decision and Control, the VSM was newly investigated on the Ten pints of Beer basis of the necessary and sufficient conditions that would account for viability in any system. The second book, then, entitled The Heart of Enterprise, presents an enriched version of the model shorn of neurophysiology. In the meantime, the second edition of Brain included five new chapters that constitute a case study of truly massive importance. The first edition was still 833 with the printers when the author was invited by President Salvador Allende to apply cybernetic thinking to the whole social economy of Chile. He took with him the manuscript of the first edition, which became the project’s “Bible”. As the Project Director, Stafford Beer was directly responsible to Dr Allende, until the coup of 1973 and the murder of the president. The story, called The Course of History, recounts how managerial cybernetics transformed some two-thirds of the social economy in a few short years. The VSM was of course deployed as the framework for the project. The whole premise of this work is novel, and nothing should be taken for granted. Orthodox information systems were totally supplanted by a real-time computerized system, based on Kalman filters and Bayesian probability theory. It used microwave links (covering the 4,000 miles extent of the country), automatic statistical filtration of information, and operation rooms as ergonomically designed environments for decision. Economic information, at every level of organization from the plant through industrial sectors to the cabinet and president himself, was filtered to highlight matters of importance – and none of this information was more than 24 hours late. Even so, the computing system was organized to decentralize authority. Despite the evident complexity of all this, Brain of the Firm in its extended edition, is written to be understood by managers. There is little mathematical content, only descriptions and diagrams. The book should be read in its entirety, and in the order presented, or the argument will not be followed, and the Chilean story at the end will make little sense. 5. Designing Freedom Designing Freedom was published by CBC Learning Systems, Toronto, and by John Wiley, London and New York in 1974. In 1993 it was transferred to House of Anansi, Toronto. In 1994 John Wiley published it in paperback form. It is also available in audio tape. It was translated into Spanish, Italian and Japanese (see Beer, 1974).

Following the author’s extensive applications of cybernetics across Canada in major organizations and government departments during the 1960s and onward, the Canadian Broadcasting Corporation invited him to deliver the Massey Lectures for 1973. This was the 13th in the annual series, which had included Noam Chomsky, Northrop Frye, Carlos Fuentes, R.D. Laing, Doris Lessing and George Steiner. The lectures consisted of a series of six broadcasts transmitted coast-to-coast on consecutive Wednesdays. The objective was to explain

K 33,3/4

834

some key concepts of cybernetics to the general public – although the audience was likely to have attended a university. The Lectures were set in a context of discussions and seminars at York University near Toronto, promoted and broadcast by the Ideas programme of CBC under the direction of James Anderson. The programme was transmitted and repeated in many countries of the world, and the Lectures themselves are available in audio tape. The published book was embellished with the author’s own sketches – intended to add to understanding rather than merely to decorate. Canada has a special copyright arrangement for the Massey Lectures, which remain in print at a cheap price for a book of only 100 pages. The Wiley publications, both in hardcover and later in paperback, were published in keeping with its overall pricing policy for this author, and the international retail cost was about five times the Canadian version – it was unsuccessful on each occasion, but sales in the Canadian version continue to this day. The Lectures had to be prepared during 1973. This was during the final year of Project Cybersyn in Chile, while Stafford Beer was reporting to President Salvador Allende. The project ended with the assassination of the President in September. For sometime during the summer the political situation was highly volatile. Beer continued as scientific director, but left Santiago to live incognito in the small village of Las Cruces on the Pacific Coast. Political associates visited there, or he went to the capital, covertly during the night. That was how the first of the Massey Lectures came to choose the ocean location to make its first points, and the third Lecture entitled “A Liberty Machine in Prototype” draws obliquely on Project Cybersyn itself. By the time of writing the final Lecture, Beer was back in London having said goodbye to the President – and three weeks after that Allende was dead. The use of the oxymoron in the title of Designing Freedom was the more poignant given that the despot General Pinochet seized power, and plunged Chile into long years of torture, murder and repression. Producer James Anderson responded to the stress he detected in Stafford Beer’s situation, and discovered that the cybernetician is also a poet. He therefore commissioned a further programme to follow the Massey Lectures – on the night of the seventh week – for a full hour’s programme of his poetry. The occasion, following a dinner party at his home in the UK, involved Stafford Beer and a close friend – the prominent British actor Barrie Ingham – who had a penchant for his poetry. These two men had an informal interchange of readings: the programme itself was unplanned and unscripted. Although the Lectures were greeted with critical acclaim, far more letters were received at CBC from the public about the poetry. . . 6. Platform for Change Platform for Change was published by John Wiley, London and New York in 1975 and reprinted in 1978, etc. In 1994 it was published as a paperback (see Beer, 1975).

The year 1970 saw a commitment to make an unprecedented number of major Ten pints of Beer speeches to specialized audiences at a very high level. There was to be a keynote address for an international assembly of the Pierre Teilhard de Chardin Association in London. Following the author’s election as President of the Operational Research Society, there had to be an inaugural address. There was to be a major address in London on the management of health, a report for 835 Unesco, a plenary address to the International Cybernetics Congress, and a presidential address for the Institute of Management Science. An annual memorial lecture for the police would soon have to be followed by an appearance in Washington DC to talk about complexity to a US House committee. Obviously, immense significance attached to that. Meanwhile, there was an important article to fit in, and a radio broadcast, before major addresses began again. A dinner address for the British Institute of Management would be followed by a second visit to Washington to talk about ecological systems. Following election as President of the Society for General Systems Research in the USA, there would have to be another presidential address, this time in Philadelphia. There was a big statement to celebrate the 500th public lecture, and the original idea was to conclude Platform for Change with that address in that same year of 1970. However, the SGSR meeting moved into another year. Finally, just edging into the following year of 1973, another memorial lecture was chosen to make the first public announcement in the UK about the Chilean work, while President Allende was making his own announcement in Santiago simultaneously. These two writings were so germane to the whole thesis of the book that they were included in the final plan. The commitments of 1970 were daunting indeed. It was on option to write one basic piece, and then tailor it for 15 different audiences. . . Instead, the whole year was devoted to designing different statements, specific to their audiences, which between them would be a vehicle of more encompassing intention. That was to express the relevance of the new science of cybernetics to holism, and to propose it as a new worldview. Insofar as key cybernetic insights were concerned, to illustrate them in all those different contexts should buttress the thinking. The work in Chile pre-empted the preparation of the book, which therefore took until 1975 to achieve its consolidated basis, although the over-arching plan had to be in place from the start. This consists of a thesis called the total system, which represents a philosophy about the world written as a single systems diagram. The integral thesis comprises six statements, each with an appropriate explanatory systems diagram, which are developed during the course of the book, and printed on blue pages. Meantime, the 15 arguments of change, as the individual public statements were called, proceed in ordinary printed format on white paper.

K 33,3/4

836

Each argument of change is prefaced by a personal statement, presented on yellow pages, explaining what the author was thinking about the occasion. This is followed by a more formal explanation of the intention underlying the statement – these are the gold pages. The yellow and gold pages, like the thesis in blue, are presented in a format that breaks sentences down, and arranges the words in groups – thereby replacing most punctuation. This is an experiment in communication, which clearly distinguishes the text of the arguments from the metasystems in which they are embedded. The result gives the appearance of free verse, but it is by no means intended as poetry (“I do write poetry and this isn’t it”, says Stafford Beer). 7. The Heart of Enterprise The Heart of Enterprise was published by John Wiley, London and New York in 1979. It was reprinted with corrections in 1988 and in 1994 was published as a paperback (see Beer, 1979).

This is the second in the trilogy dealing with the VSM – The Viable System Model. After the first book, Brain of the Firm, had become well-known, it became clear that people could not handle the theory of models for which it called – an epistemology elucidated in Decision and Control and many papers. It was often said that the book was a treatise on neurophysiology, to which other kinds of organization might be extended “by analogy”. According to this position, the “analogy” might prove to be useful or not depending on circumstances. In fact, the book postulated cybernetic principles that were supposed to be independent of context. This was a much stronger position altogether. In the 1970s when Brain was written, Stafford Beer had thought that – since people knew about their own bodies – they could be helped to perceive underlying cybernetic principles by referring to them. However, several universities and a leading teaching hospital took a reverse position. They said that, because people could understand about regulation well enough, Brain of the Firm would help them to learn about the central nervous system – and they prescribed course work accordingly! The second book in the trilogy was written to assert the stronger position. It does not mention neurophysiology. It proceeds from a basic discussion as to the very meaning of the word system, and its essentially subjective and relativistic nature, through modes of measurement, to the discovery of the five subsystems deemed essential to support any viable system. Two regulatory aphorisms, four principles of organization, and three axioms of management, are put forward, together with a theorem of recursion and an associated law of cohesion. All of these are inferred from first principles. Part 4 of the book contains notes on implementation, and includes a full-scale case study of the evolution of a management cybernetic process, based on a mutual life insurance company. The reference to heart in the title was meant to deal with another criticism – that the cybernetic approach, because it invokes scientific rigour, is basically

positivistic, and fails altogether to deal with human factors. Some contend that Ten pints of Beer it is in fact inhuman. The Heart of Enterprise tried to deal with this second major issue too. It set out to discover the necessary and sufficient conditions that must apply to any viable system; and this was explicitly to include human systems. This did not satisfy those who feel that nothing human can be said that does not involve sententiously pious sentiments expressed in gouts of 837 emotive language. On the other hand, the fact that each chapter of Heart ends with a conversation in a bar was widely tut-tutted as “inappropriate” in a serious work. . . The real meaning about the interchanges called “Later in the Bar. . .” has everything to do with humanity after all. On page 562 a systems diagram is described specifying the layout for the whole book. The episodes in the Bar are shown as attempts to foresee the misunderstandings that might predictably arise from the previous chapter, and to correct them through the medium of provocative conversation – in which humour is not disdained. No review has yet commented on this structural innovation. In the expectation that none of this would be understood, Appendix 3 draws attention to the “bar problem” by providing an alleged review of the book. The language of this is completely meaningless, but no one has ever commented about that fact either. The review is signed by the initials CWRW, writing from Wales. Someone may be interested to know that cwrw (pronounced cooroo in Welsh) is the Welsh word for beer. Between Appendix 3 and the spoof review is a wholly serious select bibliography, innovatively organized as a model of a reading list – arranged in a circle. The only comment on the bibliography, written by a distinguished reviewer, complained that there was none. It seems that neither novelty nor irony is much appreciated. 8. Diagnosing the System for Organizations Diagnosing the System for Organizations was published by John Wiley, Chichester and New York, 1985 and reprinted in 1988. It was translated into Italian and Japanese. In 1994 it appeared as a paperback (see Beer, 1985).

The third in the VSM trilogy is a kind of workbook. By working through these 150 pages systematically, a particular application of any viable system may be created. This was meant to be done on the understanding that the underlying theory in Brain and Heart had already been studied, which would mean that the crucial cybernetic philosophy of science entailed would already be deeply embedded. The workbook, then, provides a checklist of actions, and a recapitulation of key terms, to facilitate the development of a given application. Sadly, some people have naı¨vely plunged into Diagnosing the System without reading the first two books. Perhaps because it is the more recent of the three, and despite or because of its brevity, and ignoring its opening warning, it has been taken as an inclusive compendium on the viable system. Many writers have presented applications of VSM models which reference only the

K 33,3/4

838

workbook. If they have followed the instructions, they have rarely made monumental mistakes, but they offer little of substance beyond the pretty pictures. Most of all, they fail to make the advertised diagnosis – because they have not sufficient understanding to know where to look for dangerous symptoms, or how to quantify them. In June and July 1982 a group of nine mature post graduate students collected in Concordia University in Montreal to study the cybernetics of organization with Stafford Beer. Teams were formed to apply the VSM to a range of viable systems. The whole group would be involved throughout, so that the applicability of the same model to the diversity of topics would be appreciated. Seven systems were chosen for study. The university itself, and a manufacturing industry, were obvious starters. The health service of Quebec Province came next: we were meeting in Montreal after all. Next came a third world broadcasting corporation, covering no fewer than 13 levels of recursion. An organization for planetary protection was conceived and designed. The family as a viable system turned out to be an elaborate affair. Finally, the recursive embodiments of language led to some surprises when submitted to the VSM. To give a full-scale account of these studies, with all its multiple recursions, would probably have taken a full year and a very large book to write. There was no time for any of this, and no particular interest in its publication anyway. The value was in testing the presentation of the VSM in highly diverse applications. The workbook Diagnosing the System was written following this experience. There was a significant reason for taking so much time over a small book. It concerned the core diagrams of the VSM. The Viable System Model had its origins in a mathematical model of the brain, first published in 1960. This set-theoretic model (reprinted in How Many Grapes Went into the Wine? (Beer, 1994a)), was not accessible to the general reader. Moreover, the notion of recursivity which became central to the VSM is also a difficult mathematical concept. Stafford Beer took the decision to represent the mathematics in terms of rigorous diagrams. The diagrammatic rigour meant that the mathematics itself would maintain its validity. This is the reason why the author has made so many attempts to preserve the accuracy of the diagrams – with little success, as so many writers have felt free to “improve” the drawings to make them “more comprehensible” – and have destroyed the theoretical validity in the process. The first attempt to do this, in Brain, correctly depicted the structure of a single recursion. Every attempt to build into that structure the full recursivity of the VSM failed. In Brain, therefore, icons were used to indicate that a diagram repeated itself. In Heart, considerable diagrammatic advances were made, and the rigour was achieved – but without much artistic elegance. In Diagnosing, however, the problem was finally solved by depicting a level of recursion below the VSM in focus at an angle of 45 degrees. Since then, the

formal depiction has been adopted and insisted upon – not out of obstinacy, Ten pints of Beer but because no alternative diagram has ever been presented that preserved the mathematical basis correctly. 9. Beyond Dispute Beyond Dispute: The Invention of Team Syntegrity was published by John Wiley, Chichester and New York in 1994 (see Beer, 1994b).

A reader who studies the VSM diagrams in any one of the trilogy Brain, Heart, or Diagnosing, that discusses the Viable System Model, will have noticed the “fat arrows”. These are uniquely used to designate one feature of the VSM – the homeostat connecting System Three and System Four. The fat arrows are an attempt to emphasize the importance of the balancing device that provides homeostasis between (Three) energies devoted to maintaining the profitability of the status quo, and (Four) energies devoted to creating the future. Team Syntegrity was an invention designed to create a process that would facilitate the homeostasis marked by the fat arrows. It is directed to normative, directional, and strategic planning – and other creative decision processes. The underlying model is a regular icosahedron. This can be thought of as a sphere with 20 flat faces. The resulting solid object has 20 equal triangular sides. It also has 30 equivalent edges, each of which represents a person. An internal network of interactions is created by a set of protocols – “rules of the game”. It is so arranged that each person selects a unique “ownership” in two of the group’s 12 eventual topics, and a responsibility to act as a critic towards two different groups. A group organized like this is an ultimate statement of participative democracy, since each role is structurally indistinguishable from any other. There is no hierarchy, no top, no bottom, no sideways. Stafford Beer illustrates how continued dynamic interaction between persons causes ideas and resolutions to hum around the sphere, which reverberates into a kind of group consciousness. Mathematical analysis of the structure shows how the process creates synergy by spreading information uniformly throughout. Team syntegrity offers managers and their advisers a new planning method that captures the native genius of the organization in a non-political and non-hierarchical way. It has been found to generate great enthusiasm and commitment among those involved in the team – who after all are the sole designers and generators of the outcome. No agenda are put before them at the start, and they settle their own priorities as they go along. Although the invention was founded on the urgent need to make three-four homeostasis in the VSM effective, the technique can be used in other contexts – for example, to explore conflict resolution, or to generate new products. Part 1 of Beyond Dispute explains the origins of these ideas – many of which antedate the VSM – and the path-finding experiments that led to its eventual formulation as a model and technique. Part 2 elaborates further, particularly in relation to the VSM and to governance in general. The form of the model and its

839

K 33,3/4

840

icosahedral space are addressed in Part 3, and many deeply significant mathematical issues are discussed at the high-school level. Part 4 consists of a single Chapter 14 which explains what Stafford Beer means by the notion of “group consciousness”. He traces the cybernetic development of his theory that consciousness begins with a single nerve cell, or neuron, and by a process of increasing complexification reaches the level of self-awareness. The part concludes with a few speculative paragraphs on the notion of transcendence. The final part contains supporting work by a number of collaborators. One returns to the mathematics involved – not this time at high-school level, but using graph-theory and spectral analysis. The design of the algorithm required to choose between the allocation of individuals in the way most satisfying to the participants comes later. Two chapters deal with problems of facilitation in supporting the team’s activity. There are two chapters about practical concerns – one in organizing and one in implementing the outcomes. Finally, there is a very brief note to introduce the VSM to those who know nothing about it. 10 How Many Grapes Went into the Wine? How Many Grapes Went into the Wine? – Stafford Beer on the art and science of holistic management – was edited by Roger Harden and Allenna Leonard. It was published by John Wiley, Chichester and New York in 1994 (see Beer, 1994b).

This collection of Stafford Beer’s writings covers important early work which is difficult to find, and publishes some which has never been printed before. The editors included a few poems and cartoons as light relief in tracing the historical and intellectual progress of someone who considers that rigorous methods and goal alignment may coexist with artistic achievements and a concern for human beings. The title itself is a quotation from a poem that mocks the technique of cost benefit analysis. After an introduction by the editors, Stafford Beer’s own writing opens with a relatively recent (1987) address given as President to the triennial conference of the World Organization of Systems and Cybernetics, which summarizes the holistic position. This is immediately followed by way of juxtaposition with an account of empirical research undertaken in the 1950s, seeking systems that could be inherently self-organizing. Only then does Part 1 of Grapes begin with selections on interdisciplinary applications of operational research to tangible problems. These writings all date from the 1950s, and all are concerned with industrial production and its organization. Throughout this decade, Stafford Beer was head of management science for the largest steel company in the UK. For much of the time he also held the managerial position of production controller. With a scientific staff that grew to 70, and the world’s first computer (1956) dedicated to management science (a Ferranti Pegasus), major increases in productivity were recorded. The department’s holistic and problem-solving approaches dominated the

purely analytic enquiries of orthodox OR. Before the computer arrived, Ten pints of Beer completely innovative techniques were invented: massed batteries of controlled charts were designed by nomography, and simulations of stochastic processes were undertaken on a mechanical machine called SAM. Both are described here. Part 2 consists of a number of addresses introducing cybernetics – 841 considered as a new world view. These date from the late 1950s to the early 1960s, and feature a number of seminal writings. They include a plenary address at the 2nd World Conference on Cybernetics in Belgium, and the Stevenson Lecture in Newcastle which laid the foundations for the author’s epistemology. This was followed in 1960 by the inaugural address for the new Institute of Systems founded at the Case Institute in Cleveland. The speech made here was regarded as revolutionary. Many other events of a six-week intensive visit were recorded in a long forgotten diary, which is printed here for the first time. It adds retrospective comments noted in 1992. These point to the fact that much of the advanced and well-funded work on cybernetics, welcomed with such enthusiasm in 1960, has been virtually forgotten since. It is slowly and hesitantly being rediscovered today. The editors entitle Part 3 as dealing with complexity in human and social affairs. It contains four major writings. In India, the government founded an Institute in honour of the country’s illustrious scientist, Dr Zaheer. Stafford Beer was asked to give the inaugural address. He was also invited to write a critical essay as the Preface to Maturana’s seminal work on autopoiesis. Finally, there are two Memorial Lectures – strangely enough commissioned on the same day. The first was in honour of the founder of General Systems Theory, Ludwig von Bertalanffy. The second honoured Teilhard de Chardin, the Jesuit palaeontologist and mystic. Stafford Beer took a full year to write these two papers which, otherwise apparently unconnected, both ended in the recital of the 82nd Psalm. . . Perhaps the most influential paper in the whole book, which is also the most difficult to read and understand, is Chapter 10 in Part 2. This presents a model of the brain, couched in Bourbakian set theory, which Stafford Beer was working on throughout the 1950s. This may well have been the earliest formal model to take the position that the brain is a closed system. It also argued that the vast complexity of the brain must necessarily be attenuated by a conditional probability process – and postulated a possible neuro-mathematical mechanism that would make learning possible. The contentious theorem on the indeterminacy of configuration structure belongs here. The chapter concludes by displaying a picture of the operation of a cybernetic factory which was actually being implemented in a steel rolling mill. This includes a printout of an “electroencephalograph” of the company’s “brain”. The occasion of this presentation was a small closed conference on

K 33,3/4

842

self-organizing systems called by Heinz von Foerster, and attended by both Warren McCulloch and Ross Ashby who were Stafford Beer’s mentors. Their confirmation of his work was crucial to his support, as was that of Joel Elkes, a leading authority on the brain, when he came across the 1960 paper in 1982 (pp. 226-7). From these origins were developed the Viable System Model. Chapter 10 ends: “A self-organizing system must always be alive and incomplete. For completion is another name for death.” References and further reading A chronological list of cited books by Stafford Beer (including a book, edited by Roger Harnden and Allenna Leonard). Beer, S. (1959), Cybernetics and Management, English Universities Press, London; John Wiley, New York, NY, reprinted 1960, 1965, new ed. (with new chapter) 1967, reprinted 1970, 1971, 1973, etc., translated into Spanish, Russian, German, Czech, Polish, Dutch, Italian, Portuguese, French and Japanese. Beer, S. (1966), Decision and Control – The Meaning of Operational Research and Management Cybernetics, John Wiley, 1966, reprinted 1967, 1970, 1971, 1974, 1978, 1979, 1988, etc., translated into Spanish, 1994, paperback, ISBN-0-471-94838-1. Beer, S. (1968), Management Science: The Business Use of Operational Research, Aldous Books, UK and USA, hardcover and paperback, translated into German, French, Swedish, Japanese, Finnish, Dutch and Norwegian. Beer, S. (1972), Brain of the Firm, 1st ed., Allan Lane; the Penguin Press, London; Herder and Herder, USA, translated into German, Italian, Swedish and French, 2nd ed. – much extended, John Wiley, London and New York, 1981, reprinted 1986, 1988, etc., translated into Russian 1994, paperback, ISBN 0-471-94839-X. Beer, S. (1974), Designing Freedom, CBC Learning Systems, Toronto, 1974, and John Wiley, London and New York, NY, 1975; 1993 transferred to House of Anansi, Toronto; 1994, Wiley, paperback, available in audio tape, translated into Spanish, Italian and Japanese, ISBN 0-471-95165-X. Beer, S. (1975), Platform for Change, John Wiley, London and New York, NY, reprinted 1978, etc., 1994, paperback. Beer, S. (1979), The Heart of Enterprise, John Wiley, London and New York, reprinted with corrections 1988, 1994, paperback, ISBN 0-471-94837-3. Beer, S. (1985), Diagnosing the System for Organizations, John Wiley, Chichester and New York, NY, reprinted 1988, translated into Italian and Japanese, 1994, paperback, ISBN 0-471-95136-6. Beer, S. (1994a), “Stafford Beer on the art and science of holistic management”, in Harnden, R. and Leonard, A. (Eds), How Many Grapes Went into the Wine?, John Wiley, Chichester and New York, NY. Beer, S. (1994b), Beyond Dispute: The Invention of Team Syntegrity, John Wiley, Chichester and New York, NY.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

A filigree friendship

A filigree friendship

Stafford Beer Toronto, Ontario, Canada Keywords Cybernetics, Computers

843

Abstract This paper concerns the author’s long-standing friendship with Gordon Pask. It describes several occasions where their lives overlapped significantly, namely working with chemical computers, the founding of the Department of Cybernetics at Brunel University and working together on the design of Joan Littlewood’s Fun Palace project.

Introduction Jonathan Swift, that naughty dean, defined a genius as somebody against whom dunces formed a confederacy. By that criterion, and it is suitable in his case, Gordon Pask was certainly a genius. People often found him difficult to comprehend. They also recognized his powerful intellect and amazing versatility, and found it daunting. The mixture is threatening, after all; and I often saw people respond to it in the way noted by Swift. But Gordon could be naughty too. He quite often seemed to relish the effect he was having, and would nourish the confederacy of dunces rather than seek to dismantle it. To those many people who took the trouble to grapple seriously with his work, however, he was attentive, charming and infinitely patient. People started telling me colourful stories about Gordon when he was still at Cambridge and rather precocious to be a legend. Maybe he was still there when we first met. At any rate, our truly collaborative friendship lasted through the 1950s. We remained close for the rest of his life, but I shall be recalling here mainly the early years and work before his eventually huge circle of friends knew him and the later work for which he is more famous. We were both extremely conscious of the pioneering work being done in the USA in the emerging topic that Norbert Weiner had named cybernetics, and knew of everyone in the UK who was interested as well. Two such people, A.M. Uttley at the National Physical Laboratory, Teddington, and F.H. George at Bristol University, had been working on machines that could classify nervous impulses and thereby recognize pattern, not by comparison with a previously stored pattern, but as an immanent property of a neural network. So whereas it had been made clear by Heisenberg in physics decades earlier that the observer affected the system observed, it was now obvious that the observer was not only a part of a cybernetic system as being “in the loop”, but intrinsic to it. First published in: Kybernetes, Vol. 30 No. 5/6, 2001, pp. 551-559. q MCB University Press, 0368-492X

Kybernetes Vol. 33 No. 3/4, 2004 pp. 843-852 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523733

K 33,3/4

844

Connections Pask’s prototype learning/teaching machine was called Eucrates. The fact that this (eventual) family of machines was dealing with predetermined content, has come to overshadow the key point that the machine works by building a continuously adaptive model of the observer (called the student) who is building a continuously adaptive model of it. Nothing in that process is predetermined. So my very first exposure to Gordon’s science was when he sat me in a room with a monitor, in the capacity of meta-observer, and invited me to determine which screen was being driven by the human and which by Eucrates. It was impossible. The behaviour of the two elements was converging, and each was moving towards the other. And of course the mode and speed of convergence was quite unpredictable. But the content issue, the fact that we know in advance what the content-objective is, grabs all the attention. Cybernetics and Management ( Beer, 1959, pp. 124-5), which I wrote in 1958, records my experience with Gordon in learning how to operate a 12-key punch used for preparing punched cards. I had never touched such a device before, nor could I type, and yet, “forty-five minutes later I was punching at the rate of eight keys a second: as fast as an experienced punching girl [sic]”. People rightly saw such results as sensational, but failed to understand the cybernetics of how it was done. They saw it as working through conditional probability – it does – and thought that was the secret – which it is not. When Gordon took the machines to market, in the form of SAKI, a machine that could teach any keyboard skill, the engineers somehow took the cybernetic invention away. I suspect that they saw themselves as designing a machine to achieve the content-objective (learn to type), instead of building a Paskian machine to achieve the cybernetic objective itself – to integrate the observer and the machine into a homeostatic whole. Machines such as these are not available to this day, because they are contra-paradigmatic to engineers and psychologists alike. Meanwhile, the machines that I was building suffered the same fate, and Gordon and I grew close in intellectual adversity. One day we set off from London to drive to Dorking, where the engineering for SAKI was being undertaken, and had yet another frustrating visit. What I best remember about the occasion was the trip itself. Gordon was driving. And he was conversing in his usual lively fashion. This meant that he was looking intently at me, and waving his expressive hands under my nose. It follows that he was steering the car with his elbows, intuition and a large slice of luck. It was, I confess, the last time that I drove with him. Hair-raising stories about his driving followed his reputation around for years, until he finally gave up. But I stand by what I wrote 40 years ago about his Eucrates machines: “Such is the first tremendous contribution of Gordon Pask to the theory and practice of cybernetics”.

The next personal story can be properly dated to 1956. My company, United Steel, built something of a public relations exercise – filling the City Hall in Sheffield – around my delivery of the George Bray Memorial Lecture for the Institute of Production Engineers. Gordon Pask and the renowned UK pioneer of cybernetics Ross Ashby were both invited, to show professional support. The fact was that these two close friends of mine had never met, and each was most curious about the other. So the promised introduction was a large part of the draw to Sheffield for which I got the credit. They stayed at the old Grand Hotel, and the three of us met in the bar. Perhaps because Gordon’s subsequent. Conversation Theory was not yet current, these two men absolutely could not talk to each other. Gordon was speaking in his familiar style – evocative, mercurial, allusory. He would wave his arms about and try to capture some fleeting insight or to give expression to some half-formed thought. I was used to this – as I was to Ross’s rather punctilious manner. So Ashby would constantly interrupt Gordon’s stream of consciousness to say, “Excuse me, what exactly do you mean by that?” or “Would you define that term?”. Both were somewhat frustrated, and the evening was close to a disaster. But I had discharged a debt: it had been Gordon who had introduced me to a man whom I instantly adopted as a mentor. This was when he handed over Warren McCulloch to me on Euston Station, so that Warren could make the first of his visits to Sheffield. Baker Street recollection I come now to the most important and indeed exciting of my personal recollections of working with Gordon. My best estimate of the date is 1956 or 1957. He was now living in Baker Street in the heart of London. I used to call on him whenever I possibly could. When I was living in Sheffield, we used to exchange visits, funded by a small retaining contract from my department in United Steel. But whenever I was in London for other reasons I would spend most of the night working with him. On this particular evening, I was coming in just as his wife was leaving – I remember distinctly that she was going to Newcastle. I said “Hello Liz, goodbye Liz”, and Gordon and I sat down with drinks and smokes as usual and talked. We had just got into the conversation when there was a phone call. Gordon picked up the phone. It was Liz reporting that she had arrived in Newcastle! Neither of us could believe this – but that was the kind of intensity that our relationship generated. We went on from there that very night to experiment with one of Gordon’s colloidal machines. The essential component of a colloidal machine was a little dish with an open top which contained a colloid (in this case it was ferrous sulphate), a number (perhaps ten) of platinum electrodes dipped into the colloid, and an EMF was generated across the electrodes. Fine threads of metallic iron were deposited on the floor of the cell. Thus if you think of the different inputs to the set of electrodes as the coefficients of an equation, the cell “grows” the solution

A filigree friendship

845

K 33,3/4

846

to the equation while you watch. Now suppose that the electrical inputs change, then the solution to the new equation will gradually appear as an adaptation of the filigree pattern – the tree – laid down on the cell floor. This quotation describing the effect comes from Beer (1959, pp. 163-4) and is reprinted because the account was almost contemporaneous with the experiments, and because Gordon himself accepted it at the time: [The cell] is also capable of a rudimentary form of learning. Suppose that the set of input data is held constant for some time. The tree which represents the solution to these input equations grows steadily and becomes well formed. At a given moment the input set is transformed to another. This causes the tree to reorient in new directions, which it does fairly slowly. If a further transformation of the inputs is now effected, so that their values all return to the original set, the cell now tends to take up the tree formation that it had in the first place. Because it has only been slowly breaking down, it quite readily reforms: the result is that stability is reached on the second occasion very much more quickly than on the first. The strength of this conditioning is proportionate to the length of conditioning time, since this determines the robustness of the tree, and is also a function of the time taken by the cell to dissipate that growth. This can in turn be specified by the actual composition of the colloid, the size of the cell relative to the electric potential, and so on.

I describe the precise state of the thinking on that night, partly because it exactly matched – mutatis mutandis – the status of my biological experiments in Sheffield. I had been working with living systems, with daphnia, euglena, with tanks of live organisms, whereas Gordon was working with dishes of inorganic substances. The devices, however, were formally equivalent (see brief account in a 1950s paper reproduced in Harnden and Leonard (1994, pp. 25-32)), and we had been comparing notes all the way through. The first topic of discussion concerned the flexibility and self-repairing characteristics of the fabric. I had been drawn to the organic cell largely because the components were versatile and effectively self-repairing, which struck me as a huge advantage. But Gordon wanted to show me something. He had a ready-grown cell in which he had erected barriers – vertical dividers of various heights rising from the cell floor. It was immediately clear that the device could deal with this intervention: the filigree trees had simply climbed over the dividers. The talk was therefore of robustness – in particular, the ability to withstand damage. Each of us was impressed (though rather differently) by Ashby’s concept of ultrastability, whereby systems could adapt to circumstances not envisaged by the designer of the system. The colloidal cell had not been designed to climb over barriers. It was a start, but more was to come. We discussed first what would happen if the cell were massively damaged, not simply affected by a minor defect which we knew could be handled by component redundancy. Suppose, for instance, that a chunk of filigree were cut clean out. Self-repair was not a prior design characteristic, but we each reckoned that the device would succeed in closing the gap, as we put it. There was much laughter as we attempted to take out half an inch of delicate

“wire” – as if it were fuse wire which, indeed, it resembled – with a small pen-knife. The molecular bonds were very strong, and a hammer and a cold chisel had to be fetched. Even then, the operation was not at all easy to effect. But the gap was duly created. We switched on, and watched for the gap to close. It did not, but the thread repaired itself all the same. The gap simply moved along the metallic thread from anode to cathode until it was gone. Obvious. Neither of us had thought of it in advance. But we then remembered that if a nerve is cut then, if it ever regenerates, the same thing happens that we had just observed. A critical experiment So far so good. And yet these demonstrations, though exciting at the time, were somehow recognized to be trivial. “Adaptation to the unexpected” should mean more than this, and yet there must be limits. I was already developing my theory of viable systems, and often used myself as an example. But what if someone pulled out a gun and shot me. Would that be proof that I am not after all a viable system? Surely not: the system itself would have been annihilated. We fell to discussing the limiting framework of ultrastability. Suddenly Gordon said something like, “Suppose that it were a survival requirement that this thing should learn to respond to sound? If there were no way in which this ‘meant’ anything, it would be equivalent to your being shot. But this cell is liquid, and in principle sound waves could affect it. It’s like your being able to accommodate to a slap, rather than a bullet. We need to see whether the cell can learn to reinforce successful behaviour by responding to the volume of sound.” It sounded like an ideal critical experiment. I cannot remember what exactly the reinforcement arrangements were, but the cell already had them in place in order to study the rate of adaptation to input changes, and we had created various gaps in the filigree by now. And so it was that two very tired young men trailed a microphone down into Baker Street from the upstairs window, and picked up the random noise of dawn traffic in the street. I was leaning out of the window, while Gordon studied the cell. “It’s growing an ear”, he said solemnly (ipsissima verba). A few years later Gordon was to write (Pask, 1960, p. 261): We have made an ear and we have made a magnetic receptor. The ear can discriminate two frequencies, one of the order of fifty cycles per second and the other on the order of one hundred cycles per second. The “training” procedure takes approximately half a day and once having got the ability to recognize sound at all, the ability to recognize and discriminate two sounds comes more rapidly. . . The ear, incidentally, looks rather like an ear. It is a gap in the thread structure in which you have fibrils which resonate at the excitation frequency.

This was the first demonstration either of us had seen of an artificial system’s potential to recognize a filter which would be conducive to its own survival, and to incorporate that filter into its own organization. It could well have been the first device ever to do this, and no-one has ever mentioned another in my

A filigree friendship

847

K 33,3/4

848

hearing. Moreover, this facility would transform the world of information technology, if it could ever forget and transcend its origins in mere data processing. But that would require the overthrow of yet another paradigm. Meanwhile, back in the pre-computer 1950s, our critical experiment had succeeded. It was a huge encouragement to us both. The pioneers of cybernetics, especially McCulloch, had always argued on neurophysiological grounds that adaptation involved being open to information which could then be filtered and used to enrich the performance of a system that was organizationally closed. The brain seemed to rely on this fact, and now we had seen how the trick might be done in artefacts. But Gordon and I disagreed about the implications. For several years, I had been calling this behaviour, identified in managerial and social systems, “intrinsic control”. It was now becoming clear that the observer of a brain-like system had not only to be inside it, as mentioned earlier, but capable of operating on its own internal structure. That was what McCulloch had meant when he talked of the system designing its own filters according to immanent pattern. The mathematical model of the brain ( Beer, 1962) on which I was working at this time, and first presented in 1960 in the presence of Gordon (and Ashby and McCulloch), therefore proceeded on the basis that the job of the motor cortex was to find a state that matched the state of the sensory cortex – action being the by-product of that cortical homeostasis. People had (and still have) great difficulty with this epistemology; but to me it was the emergent scientific version of Leibniz’s monadology – which had made a permanent impact on me as a student of philosophy in 1943. I had no difficulty in referring to a property of self-organization within the brain/monad, therefore, since although the system had to be organisationally closed, its structure was perennially immanent. I do not think that Gordon (or Ross Ashby) ever agreed with that nomenclature. The matter hinged as far as I was concerned on the set theoretic proof I had offered in the brain model (Beer, 1962, p. 178, proposition 1.43) that was called the Law of the Indeterminacy of Configuration Structure. According to this, insofar as an organizational configuration retains its identity through time, it cannot be denoted by a general algebraic function, whereas if it can be so denoted, it cannot retain its structural identity. In other words the structure is perennially immanent, and Gordon’s objection to using the term self-organizing in this context collapses. The distinction is really important in my opinion. But we were good at agreeing to differ – even as to what actually matters. A critical chairman Let us jump a decade. In the second half of the 1960s I was development director of the International Publishing Corporation – then by far the largest publishing concern in the world – and attempting to drag an essentially

paper-and-ink organization into the electronic age. The boss of this empire (and I do mean boss) was the last of the great press barons, Cecil King. I was answerable to the managing director, F.R. (now Sir Frank) Rogers. I had been in close touch with all my cybernetic friends as I recruited a large R&D unit, and also founded the new enterprise division. This was a “container” organization for new and taken-over companies, such as Butterworths, the legal and medical publishers, which Mr King allowed me to bid for – successfully – against the competition of one Robert Maxwell. As to new companies, I was able to found International Data Highways (yes, 30 years ago) as an IPC company within the new division. Two eminent cyberneticians, Frank George and Gordon Pask, shared in my own disappointment in the performance of Harold Wilson’s “white heat of technology” government. This was operating at a barely perceptible glow, and the ministers with whom I had been trying to design a whole new strategy for national computing failed to exert any real clout. There were five ministers involved – the Postmaster General himself (John Stonehouse) “did a runner” and was discovered much later in Australia. But that whole story has yet to be told. The immediate point is that Frank and Gordon were talking seriously about emigrating. If the government could not mobilize itself in terms of cybernetics, I would try private enterprise. With the blessing of Frank Rogers I went to see the newly appointed Vice Chancellor of the new Brunel University, Professor James Topping. We devised a plan to create a National Institute of Cybernetics to be centred on Brunel. Next, I enrolled the chairman himself – Cecil King was enthusiastic indeed. He thoroughly disapproved of the brain drain. He agreed to put up a substantial sum from IPC and, of course, it was very much in the company’s interest to demonstrate leadership in electronic publishing. He had the vision to see that. Where were the other tranches of cash to come from? He instructed me to call together leading financiers and industrialists for a dinner at the Savoy at which donations would be solicited. It was a real challenge to devise a strategy that would persuade such elevated personages as were needed to come – how to lift the occasion above the inevitable suspicion of commercial skullduggery? I went to see Lord Louis Mountbatten, with whom I had had some dealings in his capacity as chairman of the National Electronic Development Council. He was very knowledgeable, and proud of his status as a Fellow of the Royal Society. He strongly approved of the plan, and agreed to be guest of honour at the dinner. He was of course also the Queen’s uncle; and it was soon possible to recruit Mr Angus Ogilvy as well – the husband of Princess Alexandria. So everybody who was anybody in the context duly came, chequebooks at the ready. Of course, Frank and Gordon were there, with the vice-chancellor, having already agreed to the plan. One day I shall publish the guest list.

A filigree friendship

849

K 33,3/4

850

The bossiness of the boss has been remarked already. Unfortunately for this enterprise, the senior management chose this moment to engineer a palace coup – an undertaking in which, ironically, I was also involved. Cecil King was instantly written out of the script. Hugh (now Lord) Cudlipp was designated the new Chairman, and he insisted on going ahead with the dinner armed with the speech that I had written for Mr King. Mr Cudlipp was a very great journalist, but I had never managed to explain even the rudiments of cybernetics to him. Moreover, it is probably fair to say that he was not one of my greatest fans – there was a paucity of sound-bites in my conversation perhaps. At any rate the dinner broke up in some disorder, without a single donation forthcoming. Dr Topping went ahead with the plan insofar as he was able, based on the solitary commitment that Cecil King had made which the new Chairman was too late to withdraw. Gordon was greatly disappointed, and he could not bring his own operation (as he had intended) into the ambit of the diminished Institute which soon became a simple department at Brunel. The funding was just not there. However, both he and Frank George used their Chairs on the diminished scale. Incidentally, and contrary to what has sometimes been said, there was never any prospect that I would take an appointment at Brunel. It was suggested early on, but I had immediately dismissed the suggestion as unseemly. The essential fun As for Gordon and our friendship, well, we became like loving brothers who kept a deeply affectionate eye on each other. Our intellectual paths diverged, our physical paths often crossed by accident or design, and our emotional paths were embedded in a relationship too complex to name lightly. I think of it as a filigree friendship, because of the central experience with the colloidal machine, and because the threads of our lives were seemingly randomly wandering but strongly bonded in a common matrix. The paper called “Easter” (Beer, 1993) was an attempt to express this obliquely. It was based on a real, if rather mysterious, experience. Gordon told me that it was the nicest communication he had ever had. I do not want to stop on so solemn a note. After all, we had a tremendous amount of fun together over 40 years. In the 1960s I had joined in with Gordon’s initiative in mounting an exhibition of cybernetic art called “Cybernetic serendipity” at the Institute for Contemporary Arts in London at their centre in The Mall. Of course, Gordon the cybernetician was also Gordon the impresario, and the ICA event had been preceded by the designing of a Fun Palace that would have metabolised high cybernetic science for people’s pure enjoyment. I at least enjoyed myself on the sidelines of that project, in which Joan Littlewood was notably involved, but it did not reach maturity. There were no Lottery funds available then, of course. Now funding is potentially available, spotlights fall on derelict sites that advertise the absence of insight and

adventure. Fun palaces are not the only projects where financial deficits are used to conceal deficits of imagination and courage. Fun itself, however, cannot be suppressed, and turns up in unlikely places. Even recently Gordon and I could be found in some corner of our shared and sedate London Club, where we were well capable of raising the eyebrows of distinguished fellow members. But I shall position my final couple of anecdotes in the city of Montreal. This is now the early 1980s. Gordon and I left our hotel in a frightful blizzard, sub-zero temperatures, and inadequate clothing, to attend a rather small meeting in an unfamiliar building. It was Sunday, the streets were deserted, and we could not gain entrance – all the doors appeared to be locked. One entrance of course proved to be open, but we did not discover it. Distraught organizers found us huddled together in a doorway, freezing and covered in snow. We were fussed over and thawed out amid much genuine concern. But it was obviously a matter for giggling – I think that we perceived it as some kind of metaphor. The next and final scene takes place in Concordia University, where we were at the same time on quite different assignments, with David Mitchell as the coenetic variable. The idea was promoted that we should engage in a public debate. We agreed, and were installed at a pair of lecterns, each with a place for notes, a glass, and a fresh bottle of Perrier water, on opposite sides of a stage. A vast blackboard covered the wall behind us. Since Gordon was in residence as a Research Professor, we agreed that he would explain what he was doing and I would attempt to pick holes in the exposition. I had no idea what to expect, but that meant that I had no need to waste time in preparation. Fair enough. The occasion was well advertised, and the place was packed. Gordon launched into a statement of profound obscurity, and covered the blackboard at high speed with his idiosyncratic mathematics and complex diagrams. There was no hope that the audience would understand this – he had chosen to address me instead of the audience, in the vein of our long-standing intimacy. Finally he tossed me the chalk crying triumphantly, “There!” There was little choice but to reply in kind. I rushed to the blackboard and amended the whole thing. “Not so!”, Gordon cried. And so we went on, the chalk flying between us and the leaping up and down gathering momentum. The audience was looking stunned, their heads moving as one like people following the progress of a tennis match. In the end, and to change the metaphor, I reckoned I had Gordon pinned down for good; I hurled the chalk at him, and it hit him quite hard on the forehead. There was a dreadful silence. Then Gordon unscrewed the cap of his unopened bottle of Perrier, walked slowly across the stage and poured the entire contents over my head. He had shifted the domain of the conversation. We put our arms round each other and left. Vale et Salve

A filigree friendship

851

K 33,3/4

852

References Beer, S. (1959), Cybernetics and Management, English Universities Press, London. Beer, S. (1962), “Towards the cybernetic factory”, in von Foerster and Zopf (Eds), Principles of Self-organization, Pergamon Press, New York, NY, (also published by Harnden and Leonard, 1994). Beer, S. (1993), “Easter: for Gordon with love”, Systems Research, Vol. 10 No. 3, pp. 13-17. Harnden, R. and Leonard, A. (Eds) (1994), “A progress note on research into a cybernetic analogue of fabric”, How Many Grapes Went into the Wine: Stafford Beer on the Art and Science of Holistic Management, John Wiley, Chichester and New York, NY. Pask, G. (1960), “The natural history of networks”, in Yovits, M.C. and Cameron, S. (Eds), Self-organizing Systems, Pergamon Press, New York, NY.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0368-492X.htm

What is cybernetics?

What is cybernetics?

Stafford Beer Keywords Cybernetics, Systems analysis, Management science Abstract An address delivered at the University of Valladolid, Spain. Asks the Question-What is Cybernetics?. Discusses popular notions and genuine difficulties. Looks at the origins, derivations and definitions of cybernetics. Considers intrinsic control and Socio-Economic Governance in real-time. Relates cybernetics to the current world situation.

853

Salutation Magnificent and most excellent Rector of the University of Valladolid Most excellent and most illustrious Authorities Members of the University Senate Ladies and gentlemen I wish that I could speak to you in Castellian, but it is almost thirty years since I last had occasion to use your beautiful language – and that was in circumstances of such stress that I still find it difficult to speak. On the 11th September 2001 a crime was committed in New York that the world will long commemorate. I hope that the Spanish-speaking world at least commemorates that earlier 11th September in 1973, the day of the Chilean golpe, when my Companero Presidente Salvador Allende died, and the people of Chile fell subject to decades of brutal suppression (Plate 1). Today, however, is a happier occasion. I wish to thank the University of Valladolid which, following the proposal made by the Escuela Universitaria Polytechnica, has bestowed on me the high honour of Doctor Honoris Causa (Plate 2). Popular notions and genuine difficulties The subject to which I have devoted my professional life is cybernetics. I am all too well aware that most people have no more than a hazy idea of what it is all about. I have often been assured that it is about freezing people – but they were thinking of cryogenics. The more informed realize that it is concerned with systems and their regulation. But even then, there are so many ways in which that notion can be approached. Some people seem to think that cybernetics is part of engineering – they reckon that feedback loops are involved, and therefore focus on specific devices described by servo-mechanics. Others have heard of systems of training through behavioural psychology, and think of pigeons, or rats running mazes. Editor’s Note: Professor Stafford Beer’s address to the University of Valladolid, Spain in October 2001. The University bestowed on him the high honour of Doctor Honoris Causa, and was first published in: Kybernetes, Vol. 31 No. 2, 2002, pp. 209 – 219. q MCB UP Limited, 0368-492X, DOI 10.1108/03684920210417283

Kybernetes Vol. 33 No. 3/4, 2004 pp. 853-863 q Emerald Group Publishing Limited 0368-492X DOI 10.1108/03684920410523742

K 33,3/4

854

Plate 1. Professor Stafford Beer addresses the University of Valladolid

Plate 2. Stafford Beer – Doctor Honoris Causa of the University of Valladolid

Even so, the most advanced type of control machinery we know about is surely found in the central nervous system of the human body. So they may go on to speculate that cybernetics is probably a branch of neurophysiology. All of these intuitions are true enough – but you might think rather localized. Let’s face it, the modem world revolves around money and global markets – we might conclude that economics should therefore have the clue to its regulation in econometrics. But wait a minute. There might be broader issues. Maeterlinck talked about the spirit of the hive, and Canetti talked about crowds. These anthropological concepts offer a more general and cohesive notion of system. Behind that idea, in turn, is arrayed a whole range of biological facts that certainly deal with regulatory phenomena. They encompass wide-ranging kinds – from the specifics of the potassium-sodium pump, to the generalities of prey-predator balances. They discuss the whole epigenetic landscape, which leads in turn to unfolding the course of evolution itself. Mind you, this discussion is still humano-centric; and maybe we should start with ecology – which one can only suppose ought to be a thoroughly systemic science. . . . It is time to stop. The shocking thing is that there is truth in every one of these notions, and the reason is because cybernetics is an interdisciplinary subject. It must be complicated. If I may be allowed one joke in a dignified discourse, it concerns three men who are about to be executed. The prison governor calls them to his office, and explains that each will be granted a last request. The first one confesses that he has led a sinful life, and would like to see a priest. The governor says he thinks he can arrange that. And the second man? The second man explains that he is a professor of cybernetics. His last request is to deliver a final and definitive answer to the question: what is cybernetics? The governor accedes to this request also. And the third man? Well, he is a doctoral student of the professor – his request is to be executed second. The great schools of a great university are valued – even revered – for their scholarship. But they do tend to suffer from a Hardening of the Faculties. This makes it particularly difficult to deal with interdisciplinary science. So that is why I feel doubly honoured to be standing here today. You have heard that I am already a doctor of science, and have received honorary doctorates in both law and in economics. But you are according me something I regard as more special. The doctorate of the University, rather than a particular faculty of the University, somehow recognizes that knowledge is a unity. This is something that I have believed all my life, but a fact difficult to bring home in a basically reductionist world. Origins – and an anecdote Cybernetics had its origins in the early 1940s, when a group of distinguished scientists was gathered together in Mexico to deal with various assignments

What is cybernetics?

855

K 33,3/4

856

associated with the second world war. It is well-documented how they discovered that – precisely because of their eminence in different fields – they found it difficult to talk to each other about anything serious. So they decided to choose a topic that was nobody’s speciality, but of interest to everyone. And their eminence was really important for another reason: they had nothing to prove. They decided to discuss the nature of control. It took a long time to get anywhere, because each specialist had an idiosyncratic view of the matter. We have already seen how popular misconceptions arise 60 years later; at the beginning, there was nothing to go on. What did control actually mean? An astrophysicist might think about laws of gravitation and cosmic repulsion as controllers, whereas a neurophysiologist might think about brains. The concepts, and certainly the vocabulary, do not coincide. I shall focus on one famous and fully authenticated incident, because it seems to embody both the form of interdisciplinary work, and also the excitement it can generate. Two members of the group had been designing a machine which would enable the blind to read with their ears. A bank of photocells would scan a line of print. As each letter passed, it would sound an audible group of notes. It is not difficult to imagine that a common word, such as the definite article, would sound a short chord that would soon be recognized as such. The main difficulty would be to cope with different sizes of print. After all, the snag already encountered with Braille, whereby every book has to be reprinted in a special format, needs to be avoided if at all possible. What these two scientists were discussing was the prospect of having the machine adjust itself automatically to the appropriate print size. They developed their idea by arguing through a schematic diagram – not an electrical circuit – which they left on the common room table when they went to bed. The next man to come into the room, who was a famous neurophysiologist, picked up the diagram. He asked: “Who is trying to draw a diagram of the fourth layer of the visual cortex of the brain? If you do not already find this story exciting, then consider the sequel. Any scanning process will have a characteristic cycle time for its periodic sweep. That will depend on its input rates. The great mathematician Norbert Wiener asked if anyone knew the rates at which the occipital lobe of the brain registers visual information from the retina. It is a complicated question, because several minicomputers operate between the two as the optic nerve is transversed. But the brain people knew the answers, and the mathematician was able to calculate the scanning rhythm. So the question was: if the human brain actually worked like the schematic diagram, what would its rhythm be? The answer was ten cycles per second – which is of course the resting rhythm of the brain. If the whole story leaves you unmoved, it’s possible that you may never acquire an interest in the subject of cybernetics.

Derivations and definitions Of the many distinguished scientists involved in those early days, the one I have mentioned by name was the ranking mathematician Norbert Wiener. That is because he actually named the subject. Why then did he call it cybernetics? Probably the first clear insight into the deep nature of control reached by the group, was that it is not about pulling levers to produce intended and inexorable results. This notion of control applies only to trivial machines. It never applies to a total system that includes any kind of probabilistic element – from the weather, to people; from markets, to the political economy. No: the. characteristic of a non-trivial system that is under control, is that despite dealing with variables too many to count, too uncertain to express, and too difficult even to understand, something can be done to generate a predictable goal. Wiener found just the word he wanted in the operation of the long ships of ancient Greece. At sea, the long ships battled with rain, wind and tides – matters in no way predictable. However, if the man operating the rudder kept his eye on a distant lighthouse, he could manipulate the tiller, adjusting continuously in real-time towards the light. This is the function of steersmanship. As far back as Homer, the Greek word for steersman was kubernetes, which transliterates into English as cybernetes. Note that on the way, via Rome, the same word in latin transformed into gubernator, which in English is governor. We should also acknowledge that long ago the French scientist Ampere, in his general classification of all knowledge, chose the word la cybernetique to describe government. I like the word cybernetics in the managerial context, because it shifts the emphasis from the biblical “do this and he doeth it”, and places it instead on the word governance. What a pity that popular usage has hijacked the root to cover everything from cybermen and cyberdogs to cybercafes, and in general to cyberspace. Anyway, not only did Wiener title his book cybernetics, he provided an illuminating subtitle in two parts. The first added the descriptor: “ communication and control ”. The pointed and direct linkage of these two words was astonishing at the time of publication, 1948. It was in no one’s minds. Today, the connection is more obvious – but I do not find that it is properly understood nor exploited. For example, it is nowadays clear that the control of crime is greatly dependent on good communication. It is also clear that the global economy depends on the fast reactions of money and markets. But forgive my saying that to observe these things is merely truistic, in the absence of deep insight into systemic purposes, and their quantification. The problem lies in the belief that massive data bases are the clue to success. But mere data are useless in themselves. They have the same relationship to information as gossip has to sustained argument. What is missing in both cases is the construction, or model, that underwrites intention – the purposes

What is cybernetics?

857

K 33,3/4

858

for which data are collected. Next, we need to quantify the complexity involved, to check that the model we are using has requisite variety. Ashby’s Law of Requisite Variety states that only variety can absorb variety. I see this law of cybernetics as of equal importance to the law of gravity in physics. Much reflection is needed to understand the law, and some people see it as tautological. By the same token, however, the whole of mathematics is either tautological or wrong! However that may be, most people would agree that mathematics turns out to be quite useful. . . The second pointed and direct linkage included in Wiener’s original definition states: “in the animal and the machine”. I fear that this continues to be astonishing to many people, and it is not surprising. For hundreds of years the categorization of knowledge drew a fundamental distinction between the living and the inanimate, and we still live with that. Well, some progress is apparent. People are fairly comfortable with such a term as biophysics, although we might wonder how many biologists think that physics has much to say to them, or vice versa. We need a lingua franca in which to talk cybernetics – and unfortunately that involves mastering new terms and new usages. In my own work, I have been careful to minimize the number of neologisms, and to define them with copious examples. Then, in appealing to the new kind of conversation generated by cybernetics, I am accustomed to appeal across the boundaries of specialties by proposing a rhetorical question to the audience. It is this: Do you think God knows the difference between physics and chemistry? It has always seemed to me a provoking thought, as a way of questioning our reductionist ways of thinking. But perhaps this piece of intellectual seduction does not work in Spanish. When I tossed the question to a large audience in Columbia five years ago, they seemed to get the point. But a leading periodical awarded me this headline: Cybernetic Guru challenges God. [Guru de la cibernetica reta a Dios]. I do hope that you did not imagine that I had any such effrontery. Intrinsic control Anyone who examines an alleged control system, rapidly realizes that a great deal depends on the speed of response. The most successful kind of control is one built into the very process of going out of control. In that case the response time is instantaneous. Because of the impressive age of this university, I was looking for an ancient example of what I call intrinsic control. I found it in the weather vane as depicted by the Dutch artist Escher. His woodcut of the ancient example includes a dedication in old Dutch, which my friends in Holland were able to translate for me. The fact that the weather vane’ must veer endlessly’ is pointed out. But the weather vane asks us whether we scorn its capriciousness – and gives the answer that it is the vane’s loyal task. There is a cybernetic sermon to preach on this text. We have here an intrinsic control system that necessarily measures wind direction – entirely its

value. It is the wind that is capricious; the measurement does not make mistakes. Compare this with the most finely honed instrument we have for measuring the capriciousness of the economy in both the United States and Britain. The best complete record of economic performance is nine months out of date. Suppose that such a delay applied to the weather vane! Because such an historical record is quite useless, treasuries are accustomed to make rather wild guesses at current figures, and they often prove to be embarrassingly and dangerously wrong. The next question to ask is this: are we interested only in maintaining and measuring the status quo, and in foreseeable issues? Not so, of course. We are interested in recognizing whatever may concern us – regardless of whether we can foresee it or not. This is precisely the issue that confronts any future scenario. Unless we have a crystal ball, we do not know what to do. The cybernetic answer to this is the detection of ultrastability. Allow me to explain. Suppose that there is a very big computer, in which we detect the risk that it may overheat. In that case, we shall install a sensitive thermostat, and deal with its overheating. Suppose that we detect a risk that the computer will be stolen. Then we shall appoint guards. These moves protect against instability. How do we deal with the unknown? We set up a programme of critical factors, and continuously test the computers effectiveness against those factors. If the test fails, we switch on the motors that will run the computer out of the building. We do not know why the building has been evacuated – but we do know that the computer is safe. In this ultrastability is the key to viable performance. I have defined a viable system as a system that is self-sustaining, or survival worthy, in just the way that a human being is viable when it can survive outside the womb. It is not totally independent – nothing in this world ever is – but it is autonomous within limits that are defined in terms of its own physiology. The model of any viable system is the basis for my major work in industry, in transportation, in education, in health, and in every other system that seeks to survive. Three of my, books are devoted to explaining the detailed theory of viability, and you will hardly expect me to summarize them in this short address. But speed of response is a major clue. In fact, we should be directing such large systems as the economy in real-time. The second clue I offer concerns recursivity. If you can envision a model of viability that is universal, then it will be effective recursively. It will apply to whole industries, and to individual firms. It will apply to large towns, and to small villages. Putting together the issue of real-time and of recursivity, we may conceive of a model in which continuous sampling recognizes incipient change before it occurs – and is therefore subject to modification. So as I said earlier, we do not try to build massive-data bases, but selective and immediate responses. This can in effect break the time barrier, pushing forward into likely futures.

What is cybernetics?

859

K 33,3/4

860

Socio-economic Governance in real-time In the beginning of the 1970s, I was invited by President Salvador Allende to redesign the social economy of Chile. I was scientific director of Project Cybersyn. There were eleven levels of recursion ranging from the State as such down to villages and enterprises. Since every model conformed to the viable system as defined, all the models were structurally identical. This explains why it was possible to have two-thirds of the work completed – two-thirds of the social economy covered – in the two years that were available. In particular, all the measures were made in real-time. No management information could be more than 24 hours out of date at any level, from the President down to the most local places. At each level there was disseminated regulation. All the measurements related to flowlines inside the Viable System Model, at each appropriate level of recursion. So by converting all measurements into uniform indices, and then filtering them continuously using Bayesian statistical theory, massive amounts of data could be processed, and presented daily to the appropriate level of management in the form of information. And I have a special definition of information: that which changes us. We do not want to engulf management with trivia; we want to isolate incipient dangers on which management can instantly act. It was intended that every management, at every level, should be equipped with an operations room. This would facilitate collegial management, and make it independent of paperwork. The prototype of this room was built in the Avenida Santa Maria, and became operational in 1972. A key to collegial management is put forward by the cybernetic concept that we call the Redundancy of Potential Command. According to this, the interaction of key managers in the operations room enables a much stronger managerial action, based upon the availability and recognition of information, than is ever possible by orthodox managerial practices. The most vivid illustration of this in my own experience, happened in Chile in October 1972. A powerful attempt to overthrow the government was made by the political opposition, with the help of the CIA. Small businesses, in the form of the gremios, were financed to mount a blockade. The idea was to take the ordinary requirements of people [food, cigarettes, petrol. . .] out of circulation, and blame the government. We already had a communications centre in working order, although nothing specific had been designed to regulate distribution. But evidently that was required, and evidently a number of ministers and key staffs were involved. We had only one computer for everything we were doing – all other communications had to work through Telex, a network that we had already appropriated. Eight teams were self-organized, and within 24 hours messages were flowing, non-stop, round-the-clock, at the rate of 2000 Telexes a day. Ministers slept on the floor, in the middle of the hubbub.

This demonstration of the redundancy of potential command in action, and in real time, truly convinced many people in the government who had hitherto been merely intellectually, acquiescent in the approach. Something as dramatic as this, perhaps, is needed to break the paradigm. One senior minister said flatly that the government would have collapsed without the cybernetic tools available to it. As it was, President Allende was allowed to live for another year. Sadly, the absence of paradigm shift, not to mention the vested interests of all concerned in operating the standard systems of management, mean that the Chilean operation has never been repeated. Bits and pieces of the holistic approach have been adopted in various other countries, but by definition they lack cohesion. The whole story is crisply explained in the last five chapters of my book Brain of the Firm. The recent horror With the complete range of cybernetic discoveries to hand, it is always possible to analyze a situation from the point of view of its regulatory phenomena. According to the cybernetician, the purpose of a system is what it does. This is a basic dictum. It stands for a bald fact, which makes a better starting point in seeking understanding than the familiar attributions of good intentions, prejudices about expectations, moral judgments, or sheer ignorance of circumstances. Last month, the tragic events in New York, as cybernetically interpreted, look quite different from the interpretation supplied by world leaders – and therefore the strategies now pursued are quite mistaken in cybernetic eyes. In the first place, we heard the usual description from world leaders of an outrage perceived as “mindless”, “senseless” and “cowardly”. We should always react with dismay to these prior judgments – because they simply mean that the speaker has no idea what is really happening or why. The real reasons are not difficult to comprehend on a systemic basis, although they are deeply offensive to the United States. In the Twin Towers, a bastion and symbol of international dominance has been overthrown. This dominance is regarded by millions, especially in the Third World, as wielding an indefensible use of economic, cultural and political power. In many countries, people have seen their compatriots slaughtered with US bombs and starved by US blockades. Their legitimate and democratic governments have often been overthrown, and replaced by US puppets who are also despots, ruling by terror and torture. The West has made little attempt to understand that point of view, because their dogma has no systemic foundation. When it comes to modus operandi, the cybernetician knows that information is power if properly deployed. It seems that 50 or 60 countries are involved in the networks that mobilized this informational power in September. It is not simply a question of “having a database”. The cybernetician knows that resources can be amplified through the process he calls intrinsic control.

What is cybernetics?

861

K 33,3/4

862

A large assailant charging a judo adept, finds himself thrown over the adept’s shoulder, and crashing into a corner – destroyed by his own strength. This amplification of system turned the hijacked planes into guided missiles. I am not approving this heinous crime, merely trying to understand it with cybernetic insight. I see nothing here that is either meaningless or mindless. On the contrary, I see a stroke of strategic brilliance, backed by ingenious tactics, and supported by thoroughgoing logistics. As to the men who sacrificed themselves in the endeavour, in other circumstances they would not be called cowardly, but heroic. The purpose of a system is what it does. That was the cybernetic dictum. What the system inaugurated on September 11th has actually done is to provoke retaliation against yet another poor country, Afghanistan, on the principle that might is right. In doing so, the United States may well seem to have abandoned the principles of justice that it has proclaimed. The outcome, which was predictable, is that the opponents of the United States are in the position to proclaim a jihad, a holy war. It is safest to assume that this was the objective from the beginning. The purpose of the system is what it does. How actually to deal with the crisis provokes further disquiet. Attempts to guard against an infinite number of inexplicit threats do not have Requisite Variety. Deeper solutions must be sought. It is vitally necessary that the United States [and its associated global capitalism] faces its bad image. It could begin by rescinding all the key propositions of the Bush era, beginning by reinstating the Kyoto agreements – feeble enough though they were. It should next begin to treat international debt seriously, starting from a recognition that for every dollar of “aid” extended to the third world, $14 of interest is exacted by way of usury. They could move on to take action on behalf of the 40,000 children who die daily from starvation and the lack of clean water, which could be put right for the cost spent on tobacco advertising. Conclusion But these are deep waters. I set out half an hour ago to answer the question “What Is Cybernetics?”, and part of that undertaking includes trying to show not only how novel the approaches become, but their relevance to what is called real-life. Whether we look back thirty years to the Chilean golpe, or come up to date with the New York disaster, we still face regulatory catastrophes well understood in cybernetic terms. And I am forced to note that in both these instances we did not fail to diagnose correctly. In both cases there is a gross disproportion of wealth and prosperity. There is no more expectation that the regulatory balances will be adjusted today then there were in Chile. President Allende said to me often: “How can we run a socialist economy in a capitalist milieu?” He took his case to the United Nations, and they cheered him to the echo. But nothing happened. Today, we have invented a war which does not exist. There is no foe, and no way of defining victory. Allies are expected to

go into battle against an abstract noun, and to assault any nation unwilling to mobilize in such folly. But still the cheers echo. We shall continue to seek answers, and I would like to emphasize that orthodox solutions are unlikely to be successful. We have tried them over and over again, and they do not work, And so we try to make them work, by spending more money or force to do again those things that do not work. We need not stereotypic solutions but new pathways. I am personally very fond of the poem Cantares, and I commend this couplet to you: caminante, no hay camino se hace camino al andar. References Beer, S. (1959), Cybernetics and Management, English Universities Press, London; John Wiley, New York, NY, reprinted 1960, 1965, new ed. (with new chapter) 1967, reprinted 1970, 1971, 1973, etc., translated into Spanish, Russian, German, Czech, Polish, Dutch, Italian, Portuguese, French and Japanese. Beer, S. (1966), Decision and Control – The Meaning of Operational Research and Management Cybernetics, John Wiley, 1966, reprinted 1967, 1970, 1971, 1974, 1978, 1979, 1988, etc., translated into Spanish, 1994, paperback, ISBN-0-471-94838-1. Beer, S. (1968), Management Science: The Business Use of Operational Research, Aldous Books, UK and USA, hardcover and paperback, translated into German, French, Swedish, Japanese, Finnish, Dutch and Norwegian. Beer, S. (1972), Brain of the Firm, 1st ed., Allan Lane; the Penguin Press, London; Herder and Herder, USA, translated into German, Italian, Swedish and French, 2nd ed. – much extended, John Wiley, London and New York, 1981, reprinted 1986, 1988, etc., translated into Russian 1994, paperback, ISBN 0-471-94839-X. Beer, S. (1974), Designing Freedom, CBC Learning Systems, Toronto, 1974, and John Wiley, London and New York, NY, 1975; 1993 transferred to House of Anansi, Toronto; 1994, Wiley, paperback, available in audio tape, translated into Spanish, Italian and Japanese, ISBN 0-471-95165-X. Beer, S. (1975), Platform for Change, John Wiley, London and New York, NY, reprinted 1978, etc., 1994, paperback. Beer, S. (1979), The Heart of Enterprise, John Wiley, London and New York, reprinted with corrections 1988, 1994, paperback, ISBN 471-94837-3. Beer, S. (1985), Diagnosing the System for Organizations, John Wiley, Chichester and New York, NY, reprinted 1988, translated into Italian and Japanese, 1994, paperback, ISBN 471-95136-6. Beer, S. (1994a), “Stafford Beer on the art and science of holistic management”, in Harden, R. and Leonard, A. (Eds), How Many Grapes Went into the Wine?, John Wiley, Chichester and New York, NY. Beer, S. (1994b), Beyond Dispute: The Invention of Team Syntegrity, John Wiley, Chichester and New York, NY. Beer, S. (2000), “Ten pints beer – the rationale of Stafford Beer’s cybernetic books (1959-94)”, Kybernetes, Vol. 29 No. 5/6, pp. 558-72.

What is cybernetics?

863