complex systems on organizations-libre

8/9/2019 Complex Systems on Organizations-libre

1/257


2/257

COMPLEX SYSTEMS AND EVOLUTIONARY

PERSPECTIVES ON ORGANISATIONSTHE APPLICATION OF COMPLEXITY THEORY TOORGANISATIONS


3/257

ADVANCED SERIES IN MANAGEMENT

Series Editor:Professor Ron SanchezIMD, Lausanne, Switzerland

The intent of the Advanced Series in Management is to produce foundational books for a nemanagement theory that will be long-lived in serving future generations of management researc

practitioners. To this end, the Advanced Series in Management has three goals:(1) publishing volumes that develop new conceptual foundations for management theory;(2) countering the trend towards increasing fragmentation in management theory and research by d

a new theory base for management that is interconnected and integrative;(3) developing new management theory that has clear, direct usefulness for the practice of manage

The volumes in the Advanced Series in Management are intended collectively to elaborate a broadmore integrated theory base for understanding and addressing the challenges facing contemporary mVolumes in the Advanced Series in Management therefore seek to stimulate and shape the develomanagement thought in ways and directions that reach beyond the content and perspectives of esmanagement theory. The Advanced Series in Management intends to be, in effect, the series of management books that is willing to break away from the pack and to publish titles that advaframeworks for management thinking


4/257


5/257

ELSEVIER SCIENCE Ltd

The Boulevard, Langford LaneKidlington, Oxford OX5 1GB, UK

© 2003 Elsevier Science Ltd. All rights reserved.

This work is protected under copyright by Elsevier Science, and the following terms and conditions ause:

PhotocopyingSingle photocopies of single chapters may be made for personal use as allowed by national copyrPermission of the Publisher and payment of a fee is required for all other photocopying, including msystematic copying, copying for advertising or promotional purposes, resale, and all forms of documenSpecial rates are available for educational institutions that wish to make photocopies for non-prot eclassroom use.

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxphone: ( + 44) 1865 843830, fax: ( + 44) 1865 853333, e-mail: [email protected]. You complete your request on-line via the Elsevier Science homepage (http://www.elsevier.com), by ‘Customer Support’ and then ‘Obtaining Permissions’.

In the USA, users may clear permissions and make payments through the Copyright Clearance CenterRosewood Drive, Danvers, MA 01923, USA; phone: (+ 1) (978) 7508400, fax: (+ 1) (978) 7504744, UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham CoLondon W1P 0LP UK; phone: ( + 44) 207 631 5555; fax: ( + 44) 207 631 5500 Other countries may h


6/257

Contents

Contributors vi

Foreword ixAuthor Biographies x

Part I: Introduction 1


7/257


8/257

Contributors

Pierpaolo Andriani Durham University, Durham, UK

Max Boisot Universitat Oberta de Catalunya, Spain

Petruska Clarkson PHYSIS, London, UK

Raul Espejo Syncho Ltd., UK

Jane Gillies Craneld School of Management, Craneld Universi


9/257

This page intentionally left blank


10/257

Foreword

In January 1995, the rst Complexity Seminar was held at the London ScEconomics, in the U.K. This was quite a momentous occasion as it proved to

turning point for the series of seminars, which had started in December 199Strategy Seminar, focusing on the relationship between Information Systebusiness strategies. That seminar, and those that followed it, had a profound efferesearch interests of Eve Mitleton-Kelly, the initiator and organiser of the series laid the foundation for what became the LSE Complexity Research Programm


11/257

Network of Excellence in Complex Systems, known asExystence (http: //www.complexityscience.org) There are 21 founding academic institutions througEurope that together set up the Network. From these are drawn the memberSteering Committee and the Coordinators of the various Work-Packages. Eve MKelly is the Coordinator of Links with Industry and Government and thus a methe Steering Committee, and is responsible for organising 3–4 seminars p.a. oof Exystence. The rst fourExystenceseminars were held at the LSE between Januand May 2003.

Eve Mitleton-Kell Director

Complexity Research Programm London School of Economi

U.K.July 2003

x Foreword


12/257

Author Biographies

Pierpaolo Andriani is a physicist and has been Project Manager for various Resand Development European projects with several years experience in the laser

and laser research. In 1997 he decided to shift from coherent laser light to inclife and moved from Florence, Italy, into academia at the University of DBusiness School, where he currently teaches in innovation and managemtechnology. Current research focuses on the application of complexity theindustrial clusters and on knowledge management.


13/257

a founder of the Applied Philosophy for Business International Association wKaterina Nicolopoulou

Raul Espejo is Managing Director of Syncho Ltd. and Visiting Professor at UniCollege Worcester, U.K. He has published extensively in books and journalsauthor of over 50 academic papers, co-author of the booksOrganization for Programm Management (Wiley, 1979) andOrganizational Transformation and Learning: Cybernetic Approach to Management (Wiley, 1996) and co-editor of the booksTheViable System Model(Wiley, 1989),Organizational Fitness: Corporate EffectiveneThrough Management Cybernetic(Campus Verlag, 1993) andTo be and not to be, thais the System(Auer Verlag, 1997). He has lectured and run seminars worldwide. Ihe created Syncho Ltd., a management consultancy in the eld of organicybernetics, in the Science Park of Aston University and has been consulorganisations like Hoechst AG in Germany, Hydro Aluminium in Norway, Europe, EdF in France and the Nuclear Inspectorate in Sweden and the NationOfce and the Ministry of Education in Colombia. His current research is focuseinformation society, organisational learning and transparency in decision-maparticular he has been working in transparency issues related to the manage

xii Author Biographies


14/257

ethics, autopoiesis and self organisation in social systems, and virtual organisatholds degrees in Management, Information Systems and Philosophy.

Roger Lewin Ph.D. is a prize-winning author of 17 science books, includinacclaimedComplexity: life at the edge of chaos, which was named as one of the 1most important science books of the twentieth century; he was the recipieninaugural Lewis Thomas Award for excellence in the communication of life scie

1989; and the 1992 Annual award for contribution to issues in conservationSociety of Conservation Biology. Between 1990 and 1993 he was a visiting pin biology at Wayne State University, and an Associate of the Peabody MHarvard University from 1993 to 1998. He is a member of the Complexity RGroup at the London School of Economics, and speaks frequently at nconferences on complexity science and business. In 1998, Lewin and his partne

Regine, founded Harvest Associates, a business consultancy that brings the princomplexity science to businesses that are struggling with transformation and chJanuary 2000 Lewin and Regine publishedThe Soul at Work: embracing complexscience for business success. which focuses on complexity in the human domain.

R b M I h received his Ph D in engineering from the University of Strath

Author Biographies xiii


15/257


16/257

crisis at the Harvard Project on the Development of Girls and the PsychoWomen. She was also a teaching fellow for Erik Erikson and trained in gestalt with Michael Miller. In 1996 and 1997 she was a visiting scholar at the CeResearch on Women and afliated with the Stone Center at Wellesley CMassachusetts, where she developed a narrative approach to organisational chprize-winning writer, she also speaks frequently at national conferences on comscience, soul and business. She is currently a member of the Complexity RGroup, at the London School of Economics.

Author Biographies xv


17/257



18/257

Part IIntroduction


19/257



20/257


21/257

The eleven chapters have been grouped into ve Parts. Part I introduces eachand provides an overview. In some cases it also offers a simpler version of the athat may help those not familiar with the concepts. It also highlights the contribution that each chapter makes to our understanding of organisations complexity perspective.

Part II sets the context by outlining the essentials of Complexity Theoorganisation studies, in terms of ten principles. It uses the ‘logic’ implicit iprinciples to argue for a different way of organising, using an ‘enabling infrastof social, cultural and technical conditions to create an organisational environmmay facilitate organisational renewal, co-evolution and sustainability.

Part III offers four different perspectives on organisational processes. Raul(Chapter 3) uses the perspective and discourse of autopoiesis to make an incontribution to the study of complex social systems. McCarthy and Gillies (Ch

introduce cladistics as a method of classication and help us to see the evolutwhole industry as a complex adaptive system. Bill McKelvey (Chapter 5) addreof the key issues in complexity, the creation of new order and explains cleaelegantly some difcult concepts in complexity science, such as entanglemenmay have some practical management applications. Pierpaolo Andriani (Chapt

4 Eve Mitleton-Kelly


22/257

ts well into a complexity context as it starts the next iteration by inviting the rgo back to earlier chapters and re-read them using the 7 Domain framework.

The ten chapters that follow make a unique contribution to our understancomplexity in an organisational context. Each chapter provides a different perof a complex organisational world and builds understanding by providing dpieces of a multi-dimensional jigsaw puzzle.

Part II: Essentials of Complexity Theory for Organisation Studies

Chapter 2 sets the theme, denes terms and explores ten ‘generic’ character

complex adaptive systems. It explains how complexity arises through inteof individual elements; it provides some of the scientic background to the deveof this new theory and introduces the following generic characteristics of cadaptive systems: self-organisation, emergence, connectivity and interdepefeedback, far-from-equilibrium, exploration-of-the-space-of-possibilities, co-ev

Introduction 5


23/257

Part III: Complexity Perspectives on Organisational Processes

Espejo in Chapter 3 takes quite a different standpoint and focuses on what individual and social complexity. He introduces the term ‘bodyhood ’, which is deneas ‘the embodied knowledge or complexity of an organisation, which is constiits resources (human and others) and their relations’. This is associated with vathe number of the possible states of a situation, which could be extraordinarilHowever, the number of actual states we are aware of is much smaller and is deon our history and situation — Espejo therefore calls thissituational complexity. “Mycomplexity in this situationis likely to evolve from the distinctions I can make. relate to theactions required from me in that situation over time”. An individhistory is the series of selected distinctions and decisions made over time, frmany possibilities open in time and space, to that individual. They thus becorepertoire of incorporated practices, which make up that individual’s bodyhoodthen relates language, which becomes the articulation of the actions throughohistory, to the space of possibilities.

The importance of bodyhood or the incorporated distinctions and practicesdene personal complexity becomes clear when it is linked with learning B



24/257

the creation of meaning and the doing are separated and there is no communication of those creating the meaning with the environment. Thus, theiris less effective. This is one of the main conclusions of the chapter — that system creates the capacity for learning and change and that not all institutinecessarily social systems, only collectives. It follows from this that certain insare dysfunctional and the question raised is ‘how can desirable social systproduced?’ The chapter explores a possible answer — it is based on learnreection and an effective organisation structure.Furthermore, it is the alignment of purpose in the information domain with in the operational domain that is critical in the emergence of an effective organwith the coherence of a social system. The process of meaning-creation is thcritical and Espejo makes the strong claim that “failure, tension and unfairness systems are the result of poor understanding of meaning-creation-processes

unreasonable external impositions”.The internal processes of meaning creation are predicated on closing the gap individuals’ meanings and the meaning they generate through the total orgaproduced by their interactions, which may be achieved by enabling effectiorganising processes and through reexivity and recursive leaning. A system’s

Introduction 7


25/257

A point to note is that McCarthy and Gillies advocate the use of metaphors anwhich are able to capture characteristics such as self-organisation, emeinnovation, learning and adaptation. They therefore open the debate in the bookuse of metaphor and of tools, which have their origin in the biological sciences

They also raise some very interesting and currently relevant questions, on thand application of cladistics, in terms of: (a) Benchmarking — what is our competitive position and how do we compare? (b) Diversity and conguratwhere do we want to be? (c) Change, parsimony and strategy — how do we ge(d) Population Ecology — what promotes an increase in a certain tyconguration?

Chapter 4 is a learned journey through the different terms of classicaticlaries the difference between, for example, typologies and taxonomies. Otional classications do not yet have a governing body for approving a s

classication system and the Chapter suggests some guidelines for a classiframework. McCarthy and Gillies take into account all the guidelines and cladistics as a method of classication. Cladistics studies the evolutionary relatbetween entities with reference to the common ancestry of the group and the mused by the authors as a way of classifying entities according to how they ha



26/257

efcaciousand not maladaptive emergence is to occur — that is emergent strfostering adaptation that enhances survival.Entanglement can be seen as the interdependence of two entities (electrons or asuch that neither can behave nor be understood independently; each entity has historyof effects from all the other entities it has come in contact with and cannot be from its interactions with these entities. Entanglement occurs when a pair of entcorrelated histories. The negation of the entanglement effect isdecohence.The rst par

of the argument is that rms need to create and re-create entanglement pintroducing and maintainingvariety and by discouraging the retention of obsostructures based on strong cliques, advance specialisations, narrow funboundaries, etc. Decohence or de-ordering of the old structures needs to take p‘uncorrupt’ the entanglement pool if it has become corrupted by legacy structu

The second part of the argument is based on the Bénard process and what M

callsadaptive tension. The Bénard process is explained in Chapter 2, briey, it is on two horizontal plates with a thin layer of liquid between them. At rest the temof the plates and the liquid is uniform; however, when heat is applied to theplate, a temperature difference is created between the bottom and top placonduction followed by convection set up a motion in the liquid. After a certain

Introduction 9


27/257

incoming stimuli” that focus on creating entanglement ties rather than on emstructure. ‘Fields’ are described as culture, specic organisational and power st(e.g. command and control), markets, technology, etc.; and (e) to destroy or dobsolete structures, so as to recreate viable entanglement pools. What McKtrying to identify are the conditions that will facilitate the regeneration of the riof entanglement pool and points out that producing entanglement at the microindependent of producing emergent macrostructure. He also acknowledg

activities aiming to create entanglement and efcacious emergence could work purposes. Nevertheless the two tasks of regenerating the entanglement poolcreating adaptive tension are necessary and sufcient for the creation of efemergent macrostructures.

The chapter goes into some depth to explain the background to entangincluding the difference between ne and coarse graining. Briey, according th

Mann the quantum world is the ne-grained structure, whereas the world of cphysics is the coarse-grained structure — one way of thinking about the diffebetween the micro and macro levels of individual agent interaction and emmacrostructures and the question McKelvey raises is “how does coarse-structure emerge from ne-grained — entangled — structure?” Those who need



28/257

uses complexity and economic geography to develop an evolutionary modedynamics of industrial clusters, using the Italian town of Prato as a paradigmaof an industrial cluster, but also citing research on Hollywood and Silicon Vallliterature on clusters uses economics and sociology and focuses on the networkorganisation. Andriani uses complexity to reinterpret the cluster phenomenolreaches the following conclusions:• A cluster is an emergent property arising from the interaction of a netw

organisations within a geographic locality;• It is dened by a particular type of environment, which features pa

technologies and production methods; a distributed system of knowledgecommunity;

• The social transactions set up an internal dynamic, which sustains the cluster• Continuous innovation and adaptability to extreme uctuations in the

characterise the cluster;• Knowledge in a cluster is highly distributed, tacit and dynamic. Individual ag

on local knowledge on the basis of micro motives; yet their constant intecreates the macro behaviour of the cluster and its collective knowledge.

Introduction 11


29/257

academically both the science and business studies domains. This shows in tway in which they describe the complexity principles that they use, such as disstructures. On the other hand this background may also have contributed to direct application of principles from the natural sciences to human systems. Inhonest and disarming account of their intellectual journey, and after trying to apmodel in practice they came to the conclusion that “complexity theory mdeveloped further to embrace many of the idiosyncrasies of social systems and

elements . . . such as reexivity, intentionality, emotion and intuition . . .”Being aware of the criticisms against ‘simple rules’, they nevertheless perwith the assumptions that identication of existing rules and new order generatiwould help the organisation achieve a “rapid switch from one organisational arto another”. They achieved the rst part of the approach, which was to idenexisting rules. The problem came when the new rules were to be applie

organisation refused to have such rules imposed from the top. This is a very imnding as it indicates that such ‘rules’ tend to emerge rather than be designedindicates that the organisation as a whole, in its manifold interactions, creates of working and relating and the ‘rules’ then emerge from those interactions aways of working.



30/257

promote the tri-partite relationship — but not in order to abstract, generalise and(which are the usual requirements of funded research), as this can only be ofapplication; instead they should establish, nurture and recycle a plethora of trknowledge-producing micro-systems. This conclusion arises from the experieeach human system is quite unique, with its own history, culture and set of relatiIt therefore appears to be very difcult to generalise ndings across differentorganisations. Perhaps what may be required are some general frameworks

would then need to be specically tailored to each organisation.The debate on the use of ‘simple rules’ (i.e. whether organisations operatebasis of a set of simple rules) in academia, consultancy and business would benthe lessons so clearly and honestly described by MacLean and MacIntosh. Onkey ndings was that ‘rules’ emerge and cannot be designed and imposed toAnother was that understanding complexity theory did help the MD of the comp

was instrumental in facilitating the creation and sharing of new knowledge.Intuition and emotion also feature in the Lewin and Regine chapter (Chawhich is based on a study of a dozen companies in the USA and U.K. rangingfrom 35 to 22,000 people. Lewin is the well-known author of ‘Complexity: life at thedge of chaos’ and Regine is a developmental psychologist who specialises

Introduction 13


31/257

‘employees’ and one manager described it as: “. . . without seeing who the pewanting them to be something for you rather than recognising who they are, iof imposition, not engagement. To be blunt, it’s dehumanising. And people wwhen they’re not included in the process and have things imposed on them”.

The study found three common behaviours between these new leaders: theyallowednew processes to emerge rather than be imposed; they were genuinelyaccessible; andthey wereattuned to their organisations, both at the macro level of the whole sy

and at the micro level of interaction between individuals.The process ofallowing, means that they encouraged experimentation, and crthe conditions whereby mistakes, contradictions, uncertainty and paradoaccepted, so that the organisation could learn and evolve. Beingaccessibleand beingattunedare linked, as both require an ability to listen to what individuals, teams organisation as a whole are saying. To this I would add, that a good leader is also

to the subtle changes in the external environment — to what the market is saywhat their customers need.In one case study, the vice president of Patient Care at a USA Medical Centre

Rusch, describes theripple effect of small changes, which could propagate througorganisation building a critical mass for change. Those small introductions of



32/257

they adopted with their clients. The clients became involved from the very bewith the St. Luke’s team, and together they co-created the advertising campaigact of co-creation and direct involvement is again a common element in organusing complexity thinking. In a totally different environment, in a DuPont cprocess plant, the operators and mechanics directly affected were involved indesign of a new control system, instead of the usual practice of bringing in engineers. The operators, mechanics and engineers worked together to co-create

system that was up and running in “half the time and half the cost that it ntakes”.In all the cases cited a common element was trust. Individuals were trusted t

with the job. Another element was the language used to describe these changesvariously called it ‘an experiment’ or ‘to experiment’ or ‘experiment in prog‘grand experiment’ — and in each case the leader created an environment

encouraged exploration of the new and cultivated conditions where people coorganise and could create new structures. The emphasis throughout is on relatiAs the authors point out, human-relations management is not new, what is newcomplexity provides insights intowhysuch practices are usually successful.

In Chapter 9, Boisot asks whether complexity can be reduced or whether it

Introduction 15


33/257

complex organisational environment need to co-exist, to complement and to enrother.Boisot shows the notion of multiple cultures both at a micro level, depicting fuwithin a rm, and at a macro level showing different industry structures (monocompetitive, emergent and oligopolistic), each with its distinct culture. But thnew work or strictly speaking related to complexity. It is the next step development of the framework that brings it into the complexity arena, by intr

Kauffman’s NK network to explore the diffusion dimension. In the I-Space framN represents the number of agents and thus corresponds to the length of the ddimension; K represents the degree of agent interconnectedness; while theparameter P is used as a rough measure of data-processing complexity. By vaand N and appropriately tuning P, phase transitions are created in the I-Space thordered, complex and chaotic social processes.

The term culture has many denitions, but nearly all of them involve the struand sharing of data within or across groups. How effectively it is done, assertsis a function of the volume of data, the size of the group(s) and the density ointeraction within and between the groups. Using these three variables he locfour institutional cultures in the ordered, complex and chaotic regimes. The ou



34/257

and qualitatively different . Therefore, there can be no direct mapping between domains of reality. To support and substantiate this argument he asserts thasystems are situatedsocially constructed and historically emerging phenomena, anduses two notions central to our understanding of the social,historicityandreexivity, toexplain the distinctive difference of social phenomena.

Introna calls ‘complexity theory’ a general term, which includes chaos dissipative structures and autopoiesis. Chapter 2 includes all these and some ad

theories under the umbrella term. Espejo in Chapter 3 uses autopoiesis extensdevelop his argument for ‘desirable social systems’.Furthermore, Chapter 10 is very outspoken and critical of the popular and c

practice of using complexity as metaphor or analogy. Introna explicates both ‘aand ‘metaphor’, and helps clarify what these two devices mean and why tlimiting and often inappropriate. The conclusion is that we need to develop

vocabulary of concepts and a newsocial complexity theory, which draws on what commensurate, but recognises that which is incommensurate. This could innovative developments in our understanding of social systems and ultimappropriate ways to intervene in organisational development. But such develorequire extensive scholarship and detailed empirical work. There are no short

Introduction 17


35/257

may be inappropriate and/or irrelevant. A third epistemological error iscross-leveldisplacement , when a condition cannot be expressed satisfactorily at its own levtries to manifest itself on another level in symbolic form. The confusion bdescribing complexity using a factual/scientic discourse and expressing ihypothetical/theoretical discourse is an example of cross-level displacemeauthors point out that there are different kinds of realities requiring different kknowing, which need different criteria and different modes of discourse to expre

The denial of different kinds of co-existing realities leads todomain conation and tothe simplistic monistic error of “one-truth must be true for everybody all the timThe authors propose that we consider at least a heptuality (seven-sidedness

existing epistemological discourses appropriate to the following seven domhuman experience:Level 1: Thephysiological/perceptual epistemological domain is the realm of senexperience. The sources of knowledge are the senses, including observation threyes.Level 2: Theaffective/emotional epistemological domain, deals with emotionsubjective feelings, which arise in response to stimulus events. These are ind



36/257

The seven domains of discourse illustrate the interwoven multiplicity of comp

terms of different realities, or different aspects of the same reality or diinterpretations of phenomena and cultural constructions. By distinguishing the of discourse it brings a much-needed conceptual clarity to this new emdiscipline.

EMK 8 January 2003

Introduction 19


37/257


38/257


39/257



40/257

Chapter 2

Ten Principles of Complexity and EnablingInfrastructures

Eve Mitleton-Kelly

Introduction

If organisations are seen as complex evolving systems, co-evolving within ‘ecosystem’, then our thinking about strategy and management changes. Wchanged perspective comes a different way of acting and relating which coulda different way of working In turn the new types of relationship and approa


41/257

(Prigogine & Stengers 1985; Nicolis & Prigogine 1989; Prigogine 1990),

Stengers (Prigogine & Stengers 1985), Gregoire Nicolis (Nicolis & PrigoginNicolis 1994) ondissipative structures; work by Humberto Maturana, Francisco Va(Varela & Maturana 1992) and Niklaus Luhman (1990) onautopoiesis(Mingers 1995)as well as the work onchaos theory(Gleick 1987) and that on economics andincreasingreturns by Brian Arthur (1990, 1995, 2002).

The above can be summarised as ve main areas of research on: (a) complex

systems at SFI and Europe; (b) dissipative structures by Ilya Prigogine and authors; (c) autopoiesis based on the work of Maturana in biology and its applicsocial systems by Luhman; (d) chaos theory; and (e) increasing returns andependence by Brian Arthur and other economists (e.g. Hodgson 1993, 2001). shows the ve main areas or research that form the background to this chapterten generic principles of complexity that will be discussed. Since the ten pr

incorporate more than the work on complex adaptive systems (CAS), the termcomplexevolving systems(CES) will be used (Allen) as more appropriate to this discussioBy comparison with the natural sciences there was relatively little wo

developing atheory of complexsocial systems despite the inux of books complexity and its application to management in the past 6–7 years (an extensiv



42/257

etc.) Such a theory may provide a different way of thinking about organisatio

could change strategic thinking and our approach to the creation of new organforms — that is, the structure, culture, and technology infrastructure of an orgaIt may also facilitate, in a more modest way, the emergence of differentways oforganisingwithin a limited context such as a single department within a rm. Thstudy at the end of this chapter describes how a different way of organising emthe Information Technology Department in the London ofce of an internationa

The chapter will discuss each principle in turn, providing some of the scbackground and describing in what way each principle may berelevant andappropriateto a human system. Regarding the ve areas of research listed on the left handFigure 1, dissipative structures are discussed at length as part of the ‘faequilibrium’ and ‘historicity’ principles; complex adaptive systems research umost of the other principles and the work of Kauffman is referred to exten

autopoiesis is not discussed in this chapter but it has played an important rolthinking underlying the current work (for the implications and applicatiautopoiesis see Mingers 1995); chaos theory is given a separate section, discussion is not extensive; and Arthur’s work on increasing returns is discussethe ‘path-dependence’ principle.

Ten Principles of Complexity and Enabling Infrastructures 25


43/257

biological, physical or chemical entities; (b) a number of researchers consi

principles of complexity only as metaphors or analogies when applied to systems. But metaphors and analogies are both limiting and limited and do notunderstand the fundamental nature of a system under study. This does not mneither metaphor nor analogy may be used. We use them as ‘transitional objecttime in the sense that they help the transition in our thinking when faced withdifcult ideas or concepts. The point being emphasised, is that using metap

analogy is not theonly avenue available to us in understanding complexity organisational or broader social context. Since organisations are, by their verycomplex evolving systems, they need to be considered as complex systems in thright.

Another way of looking at complexity is that suggested by Nicolis & Pr(1989: 8) “It is more natural, or at least less ambiguous, to speak ofcomplex behaviou

rather than complex systems. The study of such behaviour will reveal certain ccharacteristics among different classes of systems and will allow us to arrive at understanding of complexity”. This approach both honours the Principle of Conand avoids the metaphor debate. It may however upset some sociologists whond ‘arguments from science’ convincing. But this is to miss Nicolis’s and Prig



44/257

with the ‘state’ of each related individual and system, at the time. The ‘state

individual or a system will include its history and its constitution, which in tinclude its organisation and structure. Connectivity applies to the inter-relatedindividualswithina system, as well as to the relatednessbetweenhuman social systemwhich include systems of artefacts such as information technology (IT) systeintellectual systems of ideas.

Complexity theory, however, does not argue for ever-increasing inter-connecti

high connectivity implies a high degree of interdependence. This means that ththe interdependence between related systems or entities the wider the ‘rippperturbation or disturbance of a move or action by any one entity on all the otheentities. Such high degree of dependence may not always have benecial throughout the ecosystem. When one entity tries to improve its tness or positmay result in a worsening condition for others. Each ‘improvement’ in on

therefore may impose associated ‘costs’ on other entities, either within the sameor on other related systems.Connectivityand interdependenceis one aspect of how complex behaviour ar

Another important and closely related aspect is that complex systems aremulti-dimensional, and all the dimensions interact and inuence each other. In a h


l ll


45/257

1.1. Degrees of Connectivity

Propagation of inuence through an ecosystem depends on thedegree of connectivityand interdependence. Biological “ecosystems are not totally connected. Typicaspecies interacts with a subset of the total number of other species, hence the sysome extended web structure” (Kauffman 1993: 255). In human social ecosystsame is true. There are networks of relationships with different degrees of conn

Degree of connectivitymeans strength of coupling and the dependencies knowepistatic interactions— i.e. the extent to which the tness contribution made bindividual depends on related individuals. In biological co-evolutionary procetness of one organism or species depends upon the characteristics of thorganisms or species with which it interacts, while all simultaneously adapt and(Kauffman 1993: 33). In other words a single entity (allele, gene, organism or does not contribute to overall tness independently of all other like entities. Thcontribution of an individual may depend on all the other individuals in that This is a contextual measure of dependency, of direct or indirect inuence thentity has on those it is coupled with.

In a social context, each individual belongs to many groups and different cont


T P i i l f C l i d E bli I f


46/257

turn inuenced by all other related elements in an ecosystem is part of the pro

co-evolution which Kauffman describes as “a process of coupled, deforming lanwhere the adaptive moves of each entity alter the landscapes of its neigh(Kauffman & Macready 1995).

Another way of describing co-evolution is thatthe evolution of one domain or entiis partially dependent on the evolution of other related domains or entities(Ehrlich &Raven 1964; Pianka 1994; Kauffman 1993, 1995; McKelvey 1999a, 1999b;

Lewin 1998); orthat one domain or entity changes in the context of the other(s).Thenotion of co-evolution places the emphasis on theevolution of interactionsand onreciprocal evolution(Futuyama 1979). In human systems, co-evolution in the sethe evolution of interactionsplaces emphasis on the relationship between theevolving entities.

A point emphasised by Kauffman is thatco-evolution takes place within anecosystem , and cannot happen in isolation. In a human context a social ecoincludes the social, cultural, technical, geographic and economic dimensions evolution may affect both the form ofinstitutionsand therelationshipsand interactionbetween the co-evolving entities (the termentity is used as a generic term which capply to individuals, teams, organisations, industries, economies, etc.).


E Mitl t K ll


47/257


T P i i l f C l it d E bli g I f t t 31


48/257

rst example was given by Maturana at an Open University workshop (Maturan

When I buy a pair of shoes, both the new shoes and my feet will change to accomeach other. They co-evolve. What I observe at a macro-level after wearing thseveral times and suffering from sore feet, may be co-evolution happeningsame time, as both my feet and shoes change to accommodate each other. Bmicro short-term level of minute-to-minute walking, there could well havshort-term adaptation of the one to the other. This reciprocal movement is ill

more clearly by the second example given by a senior Marks & Spencer executLSE Seminar. Weavers and knitters have inuenced each other and producmaterials, which are knitted but look woven, and materials that are woven bknitted. They have co-evolved over time, with short-term adaptation to each othe market. Through the process of co-evolution they have produced somethinnew order or coherence; which is, as has been pointed out earlier, the key distin

feature of CES.Co-evolution also happens between entitieswithin a system, and therate of their co-evolution(McKelvey 1999b) is worth considering. For example, howthe rate of co-evolution within and between teams be facilitated and impCo-evolution in this context is associated with learning and the transfer of info


32 Eve Mitleton Kelly


49/257

3. Dissipative Structures, Far-from-Equilibrium and History

Another key concept in complexity is dissipative structures, which are ways iopen systems exchange energy, matter, or information with their environment anwhen pushed ‘far-from-equilibrium’ create new structures and order.

The Bénard cell is an example of a physico-chemical dissipative structure. It up of two parallel plates and a horizontal liquid layer, such as water. The dimen

the plates are much larger than the width of the layer of water. When the tempof the liquid is the same as that of the environment, the cell is at equilibrium uid will tend to a homogeneous state in which all its parts are identical (NiPrigogine 1989; Prigogine & Stengers 1985). If heat is applied to the bottom pthe temperature of the water is greater at the bottom than at the upper surfathreshold temperature the uid becomes unstable. “By applying anexternal constraintwe do not permit the system to remain at equilibrium” (Nicolis & Prigogine 19If we remove the system farther and farther from equilibrium by increastemperature differential, suddenly at a critical temperature the liquid performmovement which is far from random: the uid is structured in a series oconvection ‘cells’ known as Bénard cells.


Ten Principles of Complexity and Enabling Infrastructures33


50/257

Ilya Prigogine was awarded the 1977 Nobel Prize for chemistry for his w

dissipative structures and his contributions to nonequilibrium thermodyPrigogine has reinterpreted the Second Law of Thermodynamics. Dissolutientropy is not an absolute condition, but “under certain conditions, entropbecomes the progenitor of order”. To be more specic, “. . . under non-equconditions, at least, entropy may produce, rather than degrade, order (and) orga. . . If this is so, then entropy, too, loses its either/or character. While certain syst

down, other systems simultaneously evolve and grow more coherent” (PrigoStengers 1985: xxi).Symmetry breakingin complexity means that the homogeneity of a current or

broken and new patterns emerge. Symmetry breaking may be understood as a gof information, in the sense that when a pattern of homogeneous data is brodifferentiated patterns, the new patterns can be read as ‘information’. This phenapplies to and can be interpreted at different levels, from undifferentiate




51/257

(homogeneous data) to exception reporting, when different or unexpected

appear to deviate from the expected norms.In dissipative structures the tendency to split into alternative solutions isbifurcation, but the term is misleading in that it means a separation intotwopaths, whenthere may be several possible solutions. However, as it is easier to explain the of possibilities into two alternative paths, this simplied meaning will be used, proviso that multiple solutions are also possible. In the Bénard cell, a unique

is present until the heat differential reaches a critical value. At that point the mself-organise themselves and become right- or left-handed cells. The two possare present simultaneously. Figure 4 is borrowed from Nicolis & Prigogine (19and illustrates bifurcation.

3.1. History

An observer could not predict which state will emerge; “only chance will through the dynamics of uctuations. The system will in effect scan the territwill make a few attempts, perhaps unsuccessful at rst, to stabilize. Then a pa




52/257

type, a newdissipative structurethat is characterised by symmetry breaking

multiple choices. In chemistry,autocatalysis(the presence of a substance may increthe rate of its own production) shows similar behaviours, and the Belousov-Zha(BZ) reaction, under certain non-equilibrium conditions shows symmetry breakiorganisation, multiple possible solutions, and hysteresis (the specic path of stcan be followed depends on the system’s pasthistory) (Nicolis & Prigogine 198Kauffman 1993, 1995). Furthermore,self-reproduction, a fundamental property

biological life, is “the result of an autocatalytic cycle in which the genetic mareplicated by the intervention of specic proteins, themselves synthesized throinstructions contained in the genetic material” (Nicolis & Prigogine 1989: 18)sense, complexity is concerned with systems in which evolution — and hence— plays or has played an important role, whether biological, physical, or csystems.

Similarly in a social context, when an organisation moves away from equilibrfrom established patterns of work and behaviour) new ways of working are crenew forms of organisation may emerge. These may be quite innovative if challowed and the symmetry of established homogeneous patterns is broken. Thowever a fundamental difference between natural and social human systems. T




53/257

ecosystem. In essence, unstable environments and rapidly changing markets

exible approaches based on requisite variety (Ashby 1969).Flexible adaptation also requires new connections or new ways ofseeing things.Seeinga novel function for a part of an existing entity is called ‘exaptation’.4 A smallexample might help explain the concept. While on holiday, I was using mycomputer in the garden. The computer was on a garden table, with a hole in thefor an umbrella. The laptop was connected to a mobile telephone, which enable

send and receive emails and faxes. Both the computer and the mobile were attpower leads, which were passed through a window into the house. The plethorawas both ugly and fragile, as people passing by could trip over them. They alsoa lot of space on the table. My son Daniel then used the hole in the middle of tto keep the leads tidy and out of sight. The umbrella hole therefore gained function, in keeping the leads tidy and safe. That simple solution was an examp

exaptation. Daniel ‘saw’ the different function for the umbrella hole, while no had even considered it.When searching the space of possibilities, whether for a new product or a d

way of doing things, it is not possible to explore all possibilities. It may, howibl t id h t f h t l d i t I thi




54/257

realised in the current adjacent possible, a new adjacent possible, accessible f

enlarged actual that includes the novel discoveries from the former adjacent pbecomes available. The constant opening up of niche markets in areas and prodonly a few years earlier had not even been thought of, is an example of texpanding possibilities of the adjacent possible.

5. FeedbackFeedback is traditionally seen in terms of positive and negative feedback mechwhich are also described as “reinforcing (i.e. amplifying) and balancing” (KLehman, http://www-dse.doc.ic.ac.uk/ ~ mml/). Putting it another way, positivforcing) feedback drives change, and negative (balancing, moderating, or damfeedback maintains stability in a system. A familiar example of negative feedprovided in a central heating system. A thermostat monitors the temperaturroom, and when the temperature drops below a specied level, an adjusting meis set in motion, which turns the heating on until the desired temperature is aSimilarly, when the temperature rises above a set norm, the heating is switched




55/257

balancing feedback processes that once were able to adjust or inuence the be

of the organisation can no longer produce the desired outcome. When efforts to behaviour in order to improve performance and market position continually when incremental changes are no longer effective, then managers of organisatiresort to major interventions in an effort to produce radical change. These intervmay also fail, however, and an organisation may become locked in a constant ineffective restructuring. One reason for such failures is over-reliance on ‘adjmechanisms’ based on negative feedback loops that have worked in the past. turbulent environment, the entire ecosystem may be changing, and we cannoextrapolate successfully from past experience. New patterns of behaviour astructures may need to emerge, and these may depend on or become establishednew positive feedback processes.

In human systems, the degree of connectivity (dependency or epistatic inteoften determines the strength of feedback. Feedback when applied to interactions means inuence that changes potential action and behaviour. Furthin human interactions feedback is rarely a straightforward input-processprocedure with perfectly predictable and determined outputs. Actions and behmay vary according to the degree of connectivity between different individuals




56/257

effects of small economic shifts”, andincreasing returnsfrom positive feedback make

for many possible equilibrium points, depending on the negative feedback loops may also operate in a system (Arthur 1990).The possibility that a system may have more than one possible equilibrium p

also been described in section 3 under dissipative structures. In physico-csystems “two (or sometimes several) simultaneously stable states could coexithe same boundary conditions”. Nicolis and Prigogine call this pheno‘bistability’ and describe it as “the possibility to evolve, for given parameter valmore than one stable state” (Nicolis & Prigogine 1989: 24). Furthermore, the paths that a system may follow depend on its past history. The point here is thistory affects future development, and there may be several possible paths or that a system may follow. This explains why the precisebehaviour of a complex systemay be very difcult to predict, even while keeping the system within certain b

The classic example illustrating Arthur’s argument of increasing returns 1990, 1995) resulting from a virtuous circle of self-reinforcing growth videocassette recorder. “The VCR market started out with two competing selling at about the same price: VHS and Beta. Each format could realise increturns as its market share increased: large numbers of VHS recorders would en

p p y g 39



57/257

Arthur in later studies (Arthur 2002) looks closely at the development of tec

clusters (e.g. with electrication come dynamos, generators, transformers, swipower distribution systems; with mass production and the automobile come prolines, modern assembly methods, ‘scientic management’ road systems, oil retrafc control), which have dened “an era, an epoch, a revolution” (Arthur 20shows how they eventually change the way business is done, and that they mchange the way society is conducted. The process starts with one or more techthat ‘enable’ the new cluster (Perez 2002). The new technology cluster mayattract little notice, but then starts to achieve successes in early demonstratismall companies may be set up based on the new ideas. These compete intenseearly turbulent phase and as successes increase, and Government regulation isabsent, the promise of large prots becomes apparent and the public may speculate. In certain cases this rst exuberant phase is marked by a crash, andcites three examples, the railway industry crash in the U.K. in 1847; the Canaof the 1790s with the shares crashing in 1793; and the recent Internet crash. In the crash was followed by a sustained build-out or golden age of the technologinuenced growth in the economy and the period was one of condence and prlike the period after 1850 in the U.K. when the railways became “the engine

y



58/257

In systems theory, emergence is related to the concept of the ‘whole’ — i.e

system may need to be studied as a complete andinteracting wholerather than as anassembly of distinct and separate elements. Checkland denes emergent propthose exhibited by a human activity system “as a whole entity, which derives component activities and their structure, but cannot be reduced to them” (Ch1981: 314). The emphasis is on theinteracting wholeand thenon-reductionof thoseproperties to individual parts.

Francisco Varela (Varela & Maturana 1992; Varela 1995) in his study of the brain sees emergence as thetransition from local rules or principles of interactibetween individual components or agents, toglobal principles or states encompassithe entire collection of agents. Varela sees the transition from local to global interaction occurring as a result of explicit principles such ascoherenceandresonance,which provide the local and global levels of analysis (Varela 1995), but addsunderstand emergence fully, we also need to understand the process that enables atransition. The emergence of mental states for example, such as pattern recogfeelings and thoughts may be explained by the evolution of (macroscopic) (Vglobal principles or states) “order parameters of cerebral assemblies which are by non-linear (microscopic”) (Varela’slocal rules or principles) “interactions of neu

p p y g



59/257

decides what to do, how and when to do it; and no one outside the group direc

activities. An example is what happened in an Integrated Project Team (IPTAerospace industry. The team was brought together to create a new projemembers of the team represented rms, which outside the IPT were competitwithin the team had to cooperate and to create an environment of trust to enssensitive information, necessary for the creation of the new product, could bexchanged. The team had to prepare a six-monthly report for its various stakeThis report was on hard copy and was usually several inches thick. Some mwithin the team decided that they would try an alternative presentation. They fothey had the requisite skills among them and they put in extra time to produce report on a CD. The coming together of the sub-team to create the new formareport illustrates the principle of self-organisation. No one told them to do it suggested it. They decided what to do, how and when to do it.

Emergence in a human system tends to create irreversible structures orrelationships and organisational forms, which become part of the history of indand institutions and in turn affect the evolution of those entities: e.g. the generknowledge and of innovative ideas when a team is working together could be das an emergent property in the sense that it arises from the interaction of individ

y



60/257

blocked or restricted even in what are considered to be liberal organisational cu

complicated authorisation procedures. It is not however the case that all emproperties and all self-organisation are necessarily desirable or efcacious. M(Chapter 10, current volume) eloquently argues that under certain conditions emcould be “compromised, biased, fragile, sterile or maladaptive”. A negative sapplies to connectivity. Again complexity theory does not argue for ever-incconnectivity, as there are limits to the viable connections that can be sustained ainformation that any individual can handle, that arises from these connections.

To summarise, the main points are: (a) if we see organisations as complex esystems and if we understand their characteristics as CES, we can work witcharacteristics rather than block them; (b) those characteristics are closely relawe need to understand their interrelationship to gain maximum benet frapplication of the theory; for example, looking at emergence or self-organisisolation does not provide that deeper understanding; (c) to introduce the enabling environmentsbased on socio-cultural and technical conditions that facrather than inhibit learning and the generation and sharing of knowledge; ansound a warning that connectivity cannot be increased indenitely without breand that emergence is not always efcacious but can also become maladaptive.



61/257

7.1. Self-Similarity

One of the features of complex systems is that similar characteristics may adifferent levels and scales. In an organisational context, the generic charactercomplex systems may applywithin a rm at different levels (individual, tecorporate), as well asbetweenrelated businesses and institutions, including direcindirect competitors, suppliers, and customers, as well as legal and economic sFractal is the term often used to describe the repetition ofself-similar patterns acroslevels or scale.

The concept of fractals is related to but distinct from the notion of ‘hierarchy’ insystems theory. Hierarchy in the systems context does not refer to vertical relatof organisational structure or power, but rather to the notion ofnested subsystems. It isthe interpretation of ‘subsystem’ that differs between the two theories. A fractalreects and represents the characteristics of the whole, in the sense that similar of behaviour are found at different levels, while in systems theory, a subsystem part of the whole, as well as being a whole in its own right. It is “equivalent to systcontained within a larger system” (Checkland 1981: 317). As Checkland (1981hierarchy is “the principle according to which entities meaningfully treated as

b l f ll h h h l h l d



62/257

condition, it was not sufcient for success. Many other conditions needed to beinternally to provide asocio-technical enabling infrastructure.

The project introduced new technologies, and because of its high prole impinternational team of technical experts. What facilitated technical success weresocial conditions initiated by the project manager in charge of the project. Onmost important aspects was creating a closer working relationship between busiinformation systems professionals than had been the norm in that paorganisation. Previously, the system developers, business managers, and oppersonnel simply did not talk to each other unless absolutely necessary.

The project manager initiated a series of monthly meetings at which alconstituencies had to be present and had to discuss their part of the project in a lthat was accessible to the others. The monthly meetings, supported by information updates, enabled the three managers of technology, business, and opto talk together regularly. Initially the meetings were not welcomed, but in tivarious stakeholders involved in the projects began to identify cross-dependencibusiness project relationships, which led to new insights and ideas for new wworking. Once conditions for new forms of communication were providindividuals involved were able to self-organise, to make necessary decisions a



63/257

Another important element in this project was the articulation of business requias an iterative process through regular face-to-face meetings. The business ments meetings in the Bank were at a senior management level with: (a) a vice pwho owned the product, was responsible for the P&L (prot & loss) and determbusiness requirements; (b) a senior and experienced business project manager wa seasoned banker, with a good knowledge of the bank; and (c) a senior tecproject manager who dened the IS platform(s) and the technical developmenproject. This constant dialogue created a willingness to communicate and a glevel of trust, both of which were essential enablers of co-evolution. Thesprocesses can also be seen as positive feedback or reinforcing processes. For exampletrust facilitates better communication, which in turn enables the building of IT that facilitate both better communication and the evolution of the business.

What was achieved in this case involved a project manager, supported by himanager, who created conditions that enabled dialogue, understanding, and articulation of requirements. He created the initial conditions that improvrelationships between the domains, but he could not exactly foresee how the would work, or indeed whether it would work. As it happened, it did wosubstantialnetwork rapport was established between the domains based on tru



64/257

whose interactions create emergent properties, qualities, and patterns of behavithe actions of individual agents and the immense variety of those actions that coinuence and create emergent macro patterns or structures. In turn the macro sof a complex ecosystem inuences individual entities, and the evolutionary moves constantly between micro behaviours and emergent structures, each inand recreating the other.

The complexity approach to managing is one of fostering, of creating econditions, of recognising that excessive control and intervention can be cproductive. When enabling conditions permit an organisation to explore its spossibilities, the organisation can take risks and try new ideas. Risk taking is mhelp nd new solutions, alternative ways to do business, to keep evolving established connectivities while establishing new ways of connecting (Mitleto2000).

This approach implies that all involved take responsibility for the decisioactions they carry out on behalf of the organisation. They should not take unnrisks, nor are they blamed if the exploration of possibilities does not work. It nature of exploration that some solutions will work and some will not.

Thus, another aspect of an enabling infrastructure is the provision of space,



65/257

collaborative project with Warwick University, Craneld Ecotechnology CentreAerospace Industry (Rolls-Royce, BAe Systems, GEC, Hunting Engineering,Industries, DERA, IMI Marston, Pilatus Britten, Lucas Aerospace).

A fourth major award, also by the EPSRC is funding a 3-year collaborativeunder the Systems Integration Initiative entitled ‘Enabling the Integration of DiversSocio-cultural and Technical Systems within a Turbulent Social Ecosystem’ (GR/R37753). The industrial collaborators are British Telecommunications, NorwicLife, Rolls Royce Marine and Shell Internet Works. The project started on 1 Se2001.

The Complexity Programme industrial collaborators are both funding and rpartners. They have included BT, Citibank (New York), GlaxoSmithKlinHumberside TEC, Legal & General, Mondragon Cooperative Corporation Country), Norwich Union, Rolls-Royce Marine, Shell (International, Finance aInternet Works), the World Bank (Washington, D.C.), AstraZeneca and companies in the Aerospace industry including BAe Systems, DERA, GEC/MHS Marston, Hunting Engineering, Lucas, Rolls-Royce and Smith Industries.

The research has also been enhanced by the Strategy & Complexity Seminathe Study Groups on Complexity and Organisational Learning, the Complexity



66/257

Gell-Mann, M. (1994).The quark and the jaguar: Adventures in the simple and the comple. W.H. Freeman.

Gell-Mann, M. (1995/1996).Complexity Journal, 1 (5).Gleick, J. (1987).Chaos: Making a new science. Cardinal, McDonald & Co.Goodwin, B. (1995). How the leopard changed its spots. Phoenix.Goodwin, B. (1997). LSE strategy and complexity seminar, on 23/4/97, report o

/www.lse.ac.uk/complexityHodgson, G. M. (1993).Economics and evolution: Bringing life back into economics. Polity

Press.Hodgson, G. M. (2001). Is social evolution Lamarckian or Darwinian? In: J. LauNightingale (Eds), Darwinism and evolutionary economics(pp. 87–118). Cheltenham: EdwaElgar.

Holland, J. (1995). Hidden order: How adaptation builds complexity. Addison Wesley.Holland, J. (1998).Emergence: From chaos to order . Addison Wesley.Kauffman, S. (1993).The origins of order: Self-organisation and selection in evolution. Oxford

University Press.Kauffman, S. (1995). At home in the universe. Viking.Kauffman, S. (2000). Investigations. Oxford University Press.Kauffman, S., & Macready, W. (1995). Technological evolution and adaptive organ

l ( )


67/257


68/257

Part IIIComplexity Perspectives on Organisational Processes


69/257



70/257

Chapter 3

Social Systems and the Embodiment of Organisational Learning

Raul Espejo

1. Introduction

People’s interactions may producesocial systems, which, if and when they emerdepend upon theirorganisation in order to learn. This paper relates the concepcollective, social system and organisation. First, I introducecomplexityas the concepunderpinning our discussion of both effective organisation and learning Second

54 Raul Espejo


71/257

2. Complexity: Language, Conversations and Grounding

2.1. What is Complexity?

Something complex is not the same as something complicated. In a particular sunderstanding the total behaviour of many dynamically interrelated componentvery complicated, but dealing with the situation may not be necessarily comgood model of the situation may help the viewpoint to deal with it with only a rsmall number of alternative actions or responses. A Prime Minister dealinghugely important policy issue is likely to recognise only a limited number of opthave a very limited number of alternative actions. Forhim/her the situation icomplicated but not complex.2 On the other hand the situation is likely to be comfor those in the organisation implementing the policy. They have to deal witdetail of the policy; therefore a situation is complex if dealing with it requires a very large number of distinctions and producing a very large number of reseven if each situational state is very uncomplicated.

Complexity as detail rather than as complication can be related to Ashby’s co

Social Systems and the Embodiment of Organisational Learning 55


72/257

in any situation, and my languaging3 limitation restricts the distinctions I can makany point in time (i.e. there is a limited number of distinct futures I can langucreate at any moment). While history relates to my incorporated practices —embodied knowledge — language (grounded in my history) relates to my spossibilities. Indeed, my possibilities, however creative I might be, are restrictedistinctions (i.e. states) I can invent, appreciate and act upon and not by the nulogically possible states, which are beyond me. Implicit in language is the pinterplay between deconstructing and reconstructing the meanings of my incopractices, hence the possibility of reconguring my complexity.

Therefore,my incorporated distinctions and practices dene my personal complat a given moment in time. This is part of a learning process. As this happens my practices may become transparent to me, they are already part of my emknowledge. These distinctions and practices dene mydetailed complexityor mycomplexity in theoperational domain. For instance, if I were a musician I would hstarted incorporating very simple distinctions and practices, like notes and scaonly after these distinctions and practices became transparent to me, that isproduced them without effort, that mastering more complicated scores woubecame possible, and so forth. That is, this learning provides me with the platf


73/257



74/257

Moment-to-moment interactions among people happen through conversatshared interactive spaces or structural contexts. If these conversations and cproduce stable linguistic structures, norms and values, a social system is emergaspects conserved give form to their interactions, that is, to their relationships. say that a social system emerges from people’s recurrent interactions, which cothese people as itsroles, that is, as restricted human beings producing the interaproducing the system. Therefore, the social system emerges from people’s inteand not from the individuals themselves. The social system will exist for as lonspecic forms of their interactions are conserved. This kind of self-productionthe interacting roles are constituted by the social system emerging frominteractions, is a form of social autopoiesis4 (Luhmann 1985). It is only when trecursion happens that we have an autonomous social system, otherwise it argued that there is only a collective of people.

It is apparent that without people constituted as roles, that is, without resources, there is no social system. There is no energy producing the social These resources are necessary to produce the social system, though specic indare not essential to its emergence. The social system is produced byroles in interactionand not by specic individuals. They can be any, as long as they conserve the

58 Raul Espejo


75/257

l h l

Figure 2: Complexity in social systems.



76/257

into new practices, thus developing the system’s complexity. Of course, it mhappen, as they respond to breaks, that already learned practices are lost. Indeeyear, several world languages disappear!

The trend in modern societies is towards increasing complexity. We are conwitnessing an increased functional specialisation, which is creating more anspecialised conversations and related linguistic structures. This means that in societies we have an increased number of embedded social systems, each oconstituting roles from which particular social systems, with their own idemerge. This trend may be a strength but also, as we will see below, may be daif these emerging systems are not aligned with people’s primary concerpurposes.

3.2. Social Systems, Institutions and Organisation

An institution, that is, a collective with a normative constitution, formally creatpurpose, may support the development of a social system. Institutions can embodiment to social systems. They may make possible functional differentiatiit is also possible that institutions may never produce desirable social systems

60 Raul Espejo


77/257

are perhaps the clearest example of this; a ministry may wish to create a meathe sector it represents (e.g. Health) but often fails to produce an organisation ointeractions with those other institutions responsible for its regulation and impltion. Other institutions, like for instance enterprises, may have themselves resocreate, regulate and implement their tasks, but still are unable to produce their isocial meanings. Their structure may be inadequate. The challenge is establisscope for designingeffective organisationsable to produce desirable social systems

It is important to deal with the problem of meaning creation, that is, of pavoiding the reication of social systems. As people make sense of their interactshare meanings they create the platform for co-ordinating their actions. These purposes-in-use. People may also espouse purposes for their institutions. Hwhether there is an explicit or tacit declaration of purposes, it is important to untheir generation. Social systems, in our context, are by denition self-construcis, their meanings are created by themselves. In this sense they are purposefuactivities. On the other hand, in general, for institutions this generation of comes from without, that is they do not create their own meanings. Others produmeanings for them. They are purposive rather than purposeful. The implicatiothe thinking of those creating meanings and the doing of those producing

Social Systems and the Embodiment of Organisational Learning61


78/257

An issue to consider when awareness becomes signicant is the alignmenoperational and informational domains. For a system without self-awarenalignment may not be a problem, as its operational domain evolves naturallinformational domain distinctions are grounded. But, for a self-aware system,the clear possibility that people’s constructions in their information domain, ascribe and agree purpose (that is, meaning), are inconsistent with the syembodied knowledge. As people ascribe consciously purpose they are impparticular embodied knowledge, which is unlikely to be naturally in place. Otional diagnosis and design may help to bridge these gaps.

4. Producing Desirable Systems

Learning is critical for the effective embodiment of social systems. This is the

underpinning the creation of organisational complexity. Learning is relatively eawe have a clear focus for it, however it is not easy when this focus is hiddemultiple layers of contingency. Social systems are co-evolving in their metherefore in need of learning to maintain stability and conserve their identity. Ye

62 Raul Espejo


79/257

that a weak capacity for learning will add to their inexibility to produce dchange.

Organisational learning relates to conversational processes in which peoplsense of their actions, develop shared meanings and codes and ground practicform of frameworks, rules, routines, procedures and so forth. These practicestheir interactions as they build up the system’s complexity. But, for all this to individuals need to create and develop distinctions and practices. Organilearning relies on individual learning.

No doubt the cohesion of biological systems is far stronger than that of collbut social cohesion, if based on roles and systems of meanings rather than on pindividuals, can be indeed strong, as we witness when people defend their rbeliefs and values. As said above it isroles, and not specic individuals, that gcohesion to social systems. People may defect from systems if they so wish, broles may be constituted again and again. But, on the other hand, to the extent troles are not effectively constituted by these systems, their contributions fragmentedand distant, limiting the scope of their contribution. In such situatiochances of achieving an effective organisation, and therefore learning, are redu

The challenge is to work out those aspects more likely to produce social co



80/257

emerge from fragmented resources, however, wherever there is a stable linstructure there is some form of organisation behind it. We need to understandproduce an effective alignment of resources with desirable social meanings, in enable the development of socially desirable systems. This is an important rprogramme for the future.

It is apparent that learning is necessary for this alignment to take placeindividuals need to learn new distinctions and practices to interact effectively wproducing with them the social system. Second, institutions need to learn to aliresources with those of the system they are tacitly creating. And, third, social need to learn to align their own linguistic structures with those of the social systeaccept to belong to. These alignments and learning processes have profounimplications for individuals, institutions and social systems. The difculty ipossible mismatches between the values and norms emerging in people’s interatheir informational domain and the values and norms emerging in the opedomain of social institutions, based on the resources that society has allocated tthem. When resources are inadequate, institutional norms-in-use are unlikelconsistent with people’s expectations emerging from their interactions iinformational domain. This is where aneffective organisation structureis critical. It

64 Raul Espejo


81/257

enough practices have been incorporated. Based on this idea, we can say multitude of moment-to-moment transactions already incorporated in a trabank’s operational domain may make its complexity much larger than the compa similar size high tech non traditional bank. The latter may be facing in relativmore ‘breaks’ than the traditional bank (that is, may have a much larger variealso may have a much ‘smaller’ operational domain, of already incorporated pThe same is the case for individual managers. Their ‘large variety’, as expressedon-going problem solving, does not imply that their complexities in the enteaction domains are large. On the contrary, it is likely to be very small as they others to do whatever they do. Therefore, while the complexity of a chief exetask (i.e. the total enterprise) is likely to be very high (because of the multipleincorporated practices in the enterprise’s action domain) his/her personal compthis domain is likely to be very low. He/she is not dealing directly with all the incorporated, practices (like paying bills, sending orders to suppliers and the liexecutive’s activities are more likely to be focused on the organisation’s informdomain, where he/she will be dealing with a relatively large number of (compdistinctions and related responses, for which learning is necessary. His varietyinformational domain, may be high, but his action options in each case are few



82/257

66 Raul Espejo


83/257

to this accounting. For this purpose we need to discuss further the social accouorganisational complexity.

5.2. Social Accounting of a System’s Complexity

The issue is that in order to manage social change it is necessary to accountsystem’s complexity in both the operational and informational domains and no

the latter as often happens. Indeed it is a common experience for people to themselves to implement change without recognising the system’s relevant comin its operational domain. At the personal level, if someone wants to do somphysically demanding, where their lives may be at risk, they will account for thup practices and see that they are adequate for the requisite stretching. Organisathis is less easy to account for and often we see that systems get involved in twhich they have not developed the requisite complexity. This is often referred t‘problem of implementation’. We fail accounting for organisational complexity

When we ascribe purpose to a system the emphasis is on its relations wenvironment. Is its organisation t for the purpose?13 Or, in other words, is thorganisation effective? In this respect we are talking about its performance Perf



84/257

external impositions. Also, fragmentation of resources is responsible for an inadevelopment of the organisation’s embodied knowledge. In either case, empiricwill not recognise observational closure, that is, a coherent whole with a large to absorb variations.

These internal processes of meaning creation are hindered when there is a between individuals and the total organisation. This is one of the problemdemocratic processes in modern societies, where people often feel alienated anto contribute to the global processes for meaning creation, leaving them in the ha few politicians. Global purposes are likely to be seen as remote and far findividual’s concerns. People may feel alienated and unable to understand whaton. It is necessary to bridge this mismatch. Bridging the gap between individualglobal social system requires enabling effective self-organising processes. A smeaning is produced by the interactions of its constituted autonomous units components). And, the meanings of these autonomous units are the outcom

interactions of their own constituted (subsumed) autonomous units and so forunfolding/constitution of complexity is at the core of therecursive organisationof social systems (Beer 1979, 1985). This has important implications for the accoucomplexity. A complex task is only possible if functional specialisation take

68 Raul Espejo


85/257

operational domain. However, experiencing breaks is the engine for their olearning. It is in this general framework that we need to think about accouncomplexity. To a signicant degree the embodied operational domain for thosethis system’ is that of its subsumed systems (i.e. primary activities). The manaof breaks in these subsumed systems, to a large extent, is transparent to those global view. The implication of this structural recursion is subsumed autonomoincorporating practices in their operational domains and experiencing breaks own informational domains. This is repeated as many times as necessary to abfull the complexity of their self-constructed tasks. This implies that the opedomain of a system is the outcome ofrecursive learning processes. The transparencexperienced by people is the outcome of their structural position in the organrather than the outcome of already fully grounded practices. The transparenexperience is built on top of the breaks of all those who are experiencing the vtheir realities at different levels of the structural recursion.

The above proposition suggests that accounting for a system’s complexirecursive process demanding performance assessments for all primary activiautonomous (social) systems in a social system need to have capacity tomeanings for themselves and to regulate and implement the changes implied b



86/257

Social systems have been presented as closed networks of interrelatedpurposefully creating meanings for their actions. I have proposed that social rbecome social systems when they develop the capacity to create, regulaimplement their meanings. By this denition they self-construct their tasks, wconstraints of their own linguistic structures. We need to learn how to harneorganisation processes in order to support the evolution of desirable social syst

The challenge is creating new desirable social systems. Autonomy emergeengine for social development.

References

Ashby, W. R. (1964). An introduction to cybernetics. London: Methuen.Argyris, C., & Schön, D. A. (1978).Organizational learning: A theory of action perspecti.

Reading, Mass.: Addison-Wesley.Beer, S. (1979).The heart of enterprise. Chichester: Wiley.Beer, S. (1985). Diagnosing the system for organizations. Chichester: Wiley.Espejo, R. (1987). From machines to people and organizations: A cybernetic ins

g t I M J k & P K (Ed )N di ti i g t i


87/257



88/257

Chapter 4

Organisational Diversity, Congurationsand Evolution

Ian McCarthy and Jane Gillies

1. Introduction

despite the ease with which we may identify meaningfulgroupings of organisations, no commonly accepted

classication scheme has been developed.

72 Ian McCarthy and Jane Gillies


89/257

congurations and the associated knowledge on the processes, events and charathat both shape and dene them.

Although classifying is often a simple and habitual process, it provides a vsystem for storing and communicating knowledge. It facilitates differentiation the similar and dissimilar and has largely contributed to the advancement of knin most academic disciplines. Cladistics, as with all classications, is a metsystematically

complex systems on organizations-libre

Documents