Organisational Change Management

Complexity Theory and Strategic Change: an Empirically Informed Critique*

K. Houchin and D. MacLean w

Curriculum Leader, Faculty of Business and Computing, Stevenson College Edinburgh, Bankhead Avenue,

Edinburgh, EH11 4DE and w Senior Lecturer, Department of Business and Management, University of

Glasgow, West Quadrangle, Gilbert Scott Building, University Avenue, Glasgow G12 8QQ, UK

Email: [email protected] [Houchin]; [email protected] [MacLean]

We present a four-year ethnographic study of a public-sector organization and use

narrative to describe its development in terms of four complexity theory concepts: sensitivity to initial conditions, negative and positive feedback processes, disequilibrium

and emergent order. Our study indicates that order emerges at the boundary between

the organization’s legitimate and shadow systems. We suggest that the underlying

dynamic leading to the emergent order is the need to reduce anxiety. Our findings cause us to question the assertion that organizations are naturally complex adaptive systems

producing novel forms of order. We propose an alternate view that in social systems,

equilibrium-seeking behaviour is the norm; such systems can self-organize into

hierarchy. We draw attention to some of the difficulties we found in applying complexity-theory concepts to a social system and conclude by advocating the

development of complexity theory through the incorporation of insights from

psychology and social theory.

Introduction

This paper uses complexity theory to help our understanding of the development of specific organizational characteristics in a newly formed quango. In the last decade or so complexity theory has been advocated as a way to help understand organizational change and innova- tion. Much of the literature promotes the theory (Anderson, 1999; Colado, 1995; Stacey, 1995). Organizational examples of how it is applied in practice are not so frequent (Brown and Eisen- hardt, 1997; Pascale, 1999; Shaw, 1997; Stacey, 2000). This article helps to fill that gap by describing how key concepts of complexity theory can be used to explain how order develops

in an organization. To contain the research within manageable proportions, we limited our study to four complexity-theory concepts: sensitivity to initial conditions, negative and positive feedback processes, disequilibrium and emergent order. We have seen the development of different approaches to complexity theory; however, these four concepts are common to them all. The research tells a story, the development of AYTAG, a new regulatory quango, during its first four years, and the order that emerged in it. One of the authors worked in a key change-agent role to bring about the unification of a diverse group of people into a fully functioning organization. She gained deep insights into the actions of senior managers and staff as the organization developed its identity through the emergence of patterns of behaviour, which took the AYTAG away from the order it originally set out to establish. In this paper we offer the suggestion that

organizations can be considered complex recursive systems rather than complex adaptive systems.

*The authors are grateful to Dr MacIntosh and Dr Mayer of the Department of Business and Management, University of Glasgow, and to two anonymous re- viewers for their helpful comments on an earlier version of this article.

British Journal of Management, Vol. 16, 149–166 (2005) DOI: 10.1111/j.1467-8551.2005.00427.x

r 2005 British Academy of Management

We highlight the circularity of complexity-theory concepts when applied to human social systems. We follow the lead of Stacey (2000) and argue that when applied to human social systems, complexity theory has to be informed by psy- chology and social theory.

Theoretical approaches to strategy

The study of strategy is multifaceted. There are strongly differing opinions on most key issues. The word is generally used in different ways, suggesting that implicitly we accept various definitions, even though formally we tend to quote only one. In order to compare and contrast the many different perspectives, it is useful to use a structured framework. Mintzberg, Ahlstrand and Lampel (1998) and de Wit and Meyer (1998) both offer structured ways of looking at strategy formulation and implementation. Whittington (1993) gives us a useful model. He describes four conceptions of strategy: classical, evolutionary, processual and systemic, to help us distinguish different strategic approaches. Tensions exist in the literature, for example between strategy content and process, strategies deliberate (An- drews, 1987; Ansoff, 1984; Porter, 1980) or emergent (Mintzberg, 1988), internally ‘driven’ (e.g. the resourced-based view of the firm, core competencies) and externally ‘driven’ (e.g. Por- ter’s five forces). There are calls for a more dynamic view of strategy, which are essentially seeking the reintegration strategy content and strategy process; both strategic decision-taking and strategy processes are emergent phenomena (MacIntosh and MacLean, 1999). Whittington, Pettigrew and Thomas (2002)

have indicated that strategy research has to move away from its modernist heritage and become more creative, whilst Prahalad and Hamel (1994) recognize that the strategy field needs a new paradigm to break from the limitations of existing mindsets. Complexity theory may be the theory to do this, reconciling the essential unpredictability of industries and organizations with the emergence of distinctive patterns (Levy, 1994). It offers the prospect of an holistic framework that pulls together into a coherent whole literature, covering a number of views of strategy, some of which do not currently com- mand much attention from strategy researchers

(Stacey, 1995). Many of the ideas which complex- ity theory brings are not new. It is a reframing that is provoking a second look at many of the ideologically rooted management ideas and sees them emerge from the theoretical foundations of complexity (Anderson, 1999).

Complexity theory and the concept of emergent order

The study of non-linear dynamics has led to the development of theories such as chaos and complexity. These are now being applied to the study of organizations. The terms chaos, com- plexity, complex adaptive systems and complex- ity science are increasingly found in the strategy and organizational development and change literatures. In this article we focus on one of these, complexity theory, developed from the work of Prigogine and Stengers (1984). It has been applied to social systems and offered as a dynamic systems approach to the study of strategy (Parker and Stacey, 1994; Stacey, 1991; Tsoukas, 1998). Complexity theory deals with the nature of emergence, innovation, learning and adaption (Battram, 1998). It has developed along an interdisciplinary path, taking insights and inputs from mathematics, biology, computing and eco- nomics (McKergow, 1998). It can offer valuable insights into management and strategic issues (see e.g. Brown and Eisenhardt, 1997; Cilliers, 1998; MacIntosh and MacLean, 1999; Stacey, 2000). Different interpretations of the theory are

offered (see e.g. Anderson, 1999; Brown and Eisenhardt, 1997; Smith and Gemmell, 1991; Stacey, 1995) suggesting that it is not a coherent body of work underpinned by a robust theore- tical framework. Complexity theory concepts have been used to create a metaphorical language for describing organizational change and devel- opment (Lissack, 1997). Two interpretations have emerged that share the common theme of emergent order, but differ in their basic assump- tions about how order emerges. They can be described as the rules-based and connectionist approaches (Cilliers, 1998). The rules-based approach has its roots in

artificial intelligence, abstract mathematical mod- els and in linguistic concepts such as deep structure. Self-organization leading to the emer- gence of order is effected through the repeated

150 K. Houchin and D. MacLean

application of simple rules. Connectionist ap- proaches to complexity have neural networks of interconnecting nodes as their basis, and are supported by more interdisciplinary groups comprising neuroscientists, psychologists and engineers. The networks have no central control in the classical sense. Processing is distributed over the network and the roles of the various components change dynamically (Cilliers, 1998). Change and transformation are dependent on the capacity of the network to learn by changing the nature and number of the nodes in the network, the pattern of connections between them and the strength of these connections (Wood, 1999).

Applying complexity theory to strategic organizational development

Although there are different complexity theory interpretations, a number of common concepts are observable. These include the concepts we are using in this study: sensitivity to initial condi- tions, the presence of disequilibrium and feed- back processes, all of which interact to produce novel forms of order. These concepts form the theoretical basis for our study. Complexity theorists see organizations as

complex adaptive systems. They argue that one of the major insights the theory brings to strategy theory is that an organization can be viewed as a non-equilibrium system (Parker and Stacey, 1994). According to complexity theorists such as Stacey (1996) the concept of the organization moving from one stable state to another as a result of change is flawed. Rather, complex adaptive systems are non-linear and highly sensitive to initial conditions. The activities, events, routines, behaviours and human interac- tions existing in an organization at a specific point in time form the initial conditions for the emergence of future order. Some of these will be amplified through feedback and others dissipated through the system. Non-linearity can result in an initial condition having a disproportionate effect on an emergent outcome of the system as a whole. Apparently insignificant transactions within organizations can lead to large organiza- tional changes (MacIntosh and MacLean, 1999). Complexity theory is a departure from the

more traditional approaches to strategy, which downplay the unpredictability of the long-term

evolution of organizations. Unplanned action is inevitable and not necessarily a result of ignor- ance or incompetence. Complexity theory makes it difficult to avoid confronting the fundamentally paradoxical nature of management, in which managers must employ a rational loop within a shared mental process when they are close to certainty, and use other processes that generate instability if they are to bring about organiza- tional transformation. Strategy theorists commonly use negative feed-

back models, in that action is informed by the perceived need to ‘close the gap’ between desired and actual behaviour. Action tends to be ‘corrective’ and/or in accordance with a prior ‘designed’ order. Such models, for example Porter’s (1980) industry structure and value-chain analysis, promote the continuation of regularity and stability that are equated with success. Complex adaptive systems, in contrast, self organize. Self-organization is the natural result of non-linear interaction, not of any tendency of individual agents to prefer or seek out order. Such systems are driven by positive and negative feedback processes that are found in the different organizational systems. The formal rules and planning systems making up the legitimate organizational system govern day-to-day organi- zational behaviour. They use negative feedback processes, allowing the organization to carry out these activities in a rational and stable manner and push it towards equilibrium. The informal system (Schein, 1992) or ‘organizational shadow system’ (Egan, 1994) is where positive feedback is seen to operate, which takes the organization away from its desired state. There are interactions in the shadow system where no one is ‘in control’, but patterns of controlled behaviour appear, lead- ing to emergent order that profoundly influences the actual evolution of the organization.

Complexity theory limitations

Applying complexity theory to strategy and organizational development is not without its problems. The variety of definitions, the doubts expressed as to whether it is a theory, theories or a framework and the different meanings given to the terminology associated with complexity are all problematic (Morel and Ramanujam, 1999). Complexity theory has developed out of the

Complexity Theory and Strategic Change 151

study of physical systems. It is being applied to social systems in keeping with the tendency of the social sciences to follow the physical sciences. Gersick (1991) has shown us how understanding similar theories from different research domains can suggest thoughtful insight for others. Others ask if it is reasonable to apply theories that have their origins in the natural and biological sciences to social organizations, without necessarily tak- ing into account fundamental differences between the physical and social sciences (Chia, 1998; Goldstein, 2000; Reason and Goodwin, 1999). Complexity theory concepts have been developed by studying systems such as insect colonies, chemical phase transitions and physiological systems, in which the distinctive nature of the human being is not an issue. The temptation is to take findings from these systems and apply them directly to human systems. Gemmill and Smith (1985) for example, use the human immune system as an example of a dissipative structure, and then offer dissipative structures as a model for organizational transformation. Beinhocker (1999) states that, as both biological evolution and business evolution are complex adaptive systems, we can employ tools that help us understand biological evolution to help us under- stand the evolution of business strategy. Many of the results cited in the complexity

literature are the outputs of computer simula- tions and not firmly grounded in empirical observations (Goldstein, 2000; Rosenhead, 1998). McKelvey (1997, 1999) argues that we need a systematic agenda linking theory devel- opment with computational-model development and the testing of model structures with real- world structures. It is difficult to see how such models can account for the intricacies of human behaviour that include the role played by emotion, the options humans have to interpret and adjust or break rules, and the fact that humans belong to many systems which may or may not have easily defined boundaries. Our best use of complexity theory for understanding organization development may be as a metaphor giving us new insights, rather than trying to search for common principles across a variety of very different systems (Morgan, 1997; Tsoukas and Hatch, 2001). However, even that has to be treated with caution; given differences between human beings and, say, phase transitions in physio-chemical systems, metaphorically derived

assertions about the natural fluidity of organiza- tions have to be exposed to empirical scrutiny if they are to be made the basis of an alternative conception of organization and management. The fact relatively few academically robust

empirical organizational studies are reported in the literature is a concern for us. Case studies undertaken by Brown and Eisenhardt (1997) and MacIntosh and MacLean (1999) are examples of trying to move complexity theory onto a practical footing, but both are relatively short-term stu- dies. Neither explored in sufficient detail such issues as the effects of different social settings on the firms they were studying. Stacey (1996), Shaw (1997), Griffin, Shaw and Stacey (1998) and Seel (2000), give examples of consultancy assignments where they have applied complexity-theory con- cepts to organizational-change initiatives. They describe work that they did to create situations to increase connectivity and positive feedback so that novel forms of order could emerge. They do not say how effective their interventions were over the longer term. Relatively new and insufficiently tested theory is advocated as a sound approach to strategic change. According to McKelvey (1999) ‘without a programme of experimental testing complexity applications . . . will remain metaphorical and if made the basis of consulting agendas . . . are difficult to distin- guish from witchcraft’ (p. 21). A major motivator for this paper is to increase

the empirical examples of the application of complexity theory to strategic organization de- velopment. Our study is of a complete organiza- tion over a four-year period, allowing much more opportunity to explore emergent order. Of critical importance to this paper is human reaction to change, such as the desire to avoid anxiety, and the impact this has on human behaviour. During the research process, Author One was a contemporaneous participant in the development process rather than a researcher or consultant dipping in and out of the organiza- tion. This allowed us to access psychological attributes such as defence routines, which require close observation over a long period in order to be identified and surfaced. As a result, our conclusions differ considerably from authors of previous empirical studies in the assumptions they make about the application of complexity theory to social systems. Our research has allowed us to highlight the shortcomings in

152 K. Houchin and D. MacLean

using some of the complexity theory concepts, as well as to question the ease with which a theory developed from physical systems can translate readily to the social world of human systems.

The organization studied and the purpose of the research

AYTAG is a quango and its business is environ- mental regulation. Formed in 1996, it combined River Purification Boards (RPBs), industrial- pollution inspectors employed by central govern- ment and some environmental-health functions from local authorities. Author One joined AYTAG a month after its formation as training and development manager. In the reports pro- duced about how the new organization should be structured, in the first corporate plan and in statements the corporate management team made to staff, there were clear indications of the future order that the organization wanted to achieve. AYTAG set out to be an organization with:

(1) wide spans of control; (2) an emphasis on employee flexibility; (3) empowered managers, delegation to its low-

est point; (4) an emphasis on value for money; (5) a strong centre to knit the business together

out of the organization’s predecessors; (6) a reputation as an influencer in environmen-

tal matters as well as a regulator.

The authors’ interest in complexity theory and the organizational situation offered the opportu- nity for case-study research. The purpose of the research was to study the organization through the concepts of complexity theory identified earlier, to gain insight into the order that emerged in AYTAG and why it was similar to, or different from, what was initially desired. The research would allow us to take a view on the usefulness of complexity theory to our understanding of strategic organizational development.

The research methodology

We have highlighted the fact that complexity theory is relatively untested in social systems. Parkhe (1993) argues that the development of theory should follow a research route which

begins with exploratory research. A suitable starting point is a single case study (Brown and Eisenhardt, 1997; Eisenhardt, 1989; Parkhe, 1993; Yin, 1993). Tsoukas and Hatch (2001) suggest that a narrative approach indicates and supports the logic of complexity theory. The dominant frame of reference for research in management and strategy is quantitative. This informs the goals and methodologies of research- ers. From a complexity-theory perspective reductionist approaches to researching organiza- tions are likely to produce misleading conclu- sions. Stacey (2000) deals with this issue by arguing for a reflexive methodological position to seek out the patterns of behaviour that lead to emergent self-organizing processes. Such a posi- tion is consistent with complexity theory, with insights regarding the research coming from the broad principles of the theory itself. As an employee of AYTAG, Author One was

not able to stand outside the organization as an objective observer. This made it almost impossible to use a positivist or post-positivist methodology. Observations would be unavoidably observer- dependent (Weick, 1979). According to East- erby–Smith, Thorpe and Lowe (1991) the role of the researcher as an employee is appropriate when the researcher needs to become totally immersed, and experience the work or situation first-hand, as sometimes it is the only way to gain the kinds of insights sought. The authors took an ethnogra- phical approach. Such an approach favours local, historically contextualized social inquiry and is sensitive to the qualities of emergent phenomena. There were elements of participant inquiry, as Author One was a participant in the development of AYTAG. Deep insights were gained as Author One had an official role in a legitimate control system while simultaneously participating in the shadow system. Theory developed from deep insights gained from working in both systems is more accurate and more appropriately tentative because the researcher must take into account the intricacies and qualities of the particular context (Pettigrew, 1997).

The research process

Data gathering covered a four-year period from May 1995 to August 1999. For information about the setting up of AYTAG in the year prior

Complexity Theory and Strategic Change 153

to its becoming operational in April 1996, historical data were collected. The remaining data were collected as they were produced. Data collection was a real-time activity from May 1996 to August 1999. The majority of the staff in AYTAG were professionals who used research information on a regular basis. This meant that there was a great deal of easily available explicit information, which made secondary-data collec- tion straightforward. The authors’ interest in complexity theory ensured that both tacit and explicit data were collected (Stacey, 2000; Seel, 2000). The main method of data collection was participant observation. This allowed the authors to examine situations first-hand from partici- pants’ points of view and observe what people actually did, rather than reporting what people said they had done. Author One kept a personal diary. She had access to operational, policy and administrative staff at all levels of the organiza- tion through her work activities. During the four years, informal conversational interviews were conducted with all of AYTAG’s managers (approximately 150) and semi-structured inter- views were conducted with three members of the corporate management team. During the course of the research a great deal

of data were gathered. Hand-collected data such as interview notes and the personal diary were transcribed. All the data including corporate plans, emails, management-team reports, work- shop outputs and transcribed, hand-written data, were put in date order. Once the data were in date order, mind

mapping (Buzan, 1991) and process mapping (Langley, 1999) were used to group the data around individual themes. As a result of this exercise the dominant themes were:

� the Business of AYTAG; � the initial structure and recruitment of man-

agers and staff; � the Budget crisis and subsequent restructuring; � business planning in AYTAG; � the Science Reviews; � human-resource initiatives; � development of environmental strategies. Each of these was written up as a narrative

account telling a chronological story of what happened in AYTAG. The narrative accounts were then used to map events and actions against the key concepts of complexity theory – sensitiv-

ity to initial conditions, disequilibrium, feedback and emergent order. There was continuous checking back and forth between the original data and the narrative accounts. In the following sections is a discussion of the findings, ordered in reference to the key concepts of complexity theory. Complexity theory is relatively new; the

research path is less well-defined, so it may be perceived as high risk and overly subjective, with knowledge produced that is not necessarily generalizable. Throughout the data collection and analysis, validity procedures were used to minimize bias. These included a variety of data sources, exposing findings to others for comment (Yin, 1993) and maintaining a continuing aware- ness of the collusive processes that can occur – particularly when a researcher is also an employ- ee (Gilmore and Krantz, 1985). These procedures diminish some of the distortions which can occur in qualitative research. The perspective from which the research conclusions are derived can then be communicated more clearly.

The development of AYTAG

When AYTAG was formed there was a strong vision for the organization contained in its strategic objectives and mission statement. To- gether they expressed the new state which the organization wanted to reach. Its complex initial conditions influenced the way in which it moved forward. So many of its staff came into the organization over a very short period of time – not having worked together before and with so few policies and procedures to guide them. Once AYTAG became operational, disequilibrium-creat- ing actions occurred, positive and negative feed- back processes operated which altogether produced the actual future state of the organization in August 1999 when this research ended. This process is described in diagrammatic form in Figure 1.

Initial conditions in AYTAG

AYTAG was a public body established by a government that favoured the centralization of services. It took over services previously provided by public-sector organizations that were more directly accountable to the local community. Its formation was disliked and opposed by locally

154 K. Houchin and D. MacLean

elected representatives. AYTAG was required to operate within a framework set by central govern- ment for managing its business. Initiatives such as ‘efficiency scrutinies’ and ‘key performance indicators’ had to be implemented. All the members of AYTAG’s Main Board

were appointed by the Secretary of State. A chief executive, a former civil servant was appointed in August 1995 and ‘a lean structure with wide spans of control’ was decided in September 1995. Five directorates were formed. Two, Corporate Services and Environmental Strategy, were based at the head office. The other three were regional operational directorates. The five directors, the senior and the middle managers, had all worked in predecessor organizations. With the exception of head-office staff, most employees had worked in former River Purifica- tion Board offices. The day-to-day operational business of AY-

TAG was carried out in the three Regions, in science laboratories or in multidisciplinary envir- onment-protection teams. Corporate Services contained the support functions; finance, person- nel, management services, information technol- ogy and public affairs. The Environmental Strategy Directorate developed policy in relation

to environment protection, sponsored research and promoted the non-legislative aspects of environment protection. Over 500 employees came from 63 different

predecessor bodies and around 150 were directly recruited. AYTAG’s transferred employees re- tained their terms and conditions and rates of pay. The RPB staff had worked in small organizations with relatively undifferentiated structures. Local-authority staff came from orga- nizations larger than AYTAG with clearly defined central-support departments. These for- mer organizations were hierarchically structured and paternalistic in their cultures. The nature of their regulatory work was reflected in an aversion to risk and a dependence on rules and proce- dures. On transferring into AYTAG, many managers found themselves managing interdisci- plinary teams and much more was demanded of them. Multi-skilled teams of professional envir- onmental-protection officers were seen as critical to the development of a ‘one-door approach’ to which AYTAG expected to move, as well as efficient and effective use of staff resources. To facilitate the integration of employees,

‘Introduction Workshops’ were delivered in 1996 and attended by every employee. The

Figure 1. The emergence of order

Complexity Theory and Strategic Change 155

workshop facilitator commented in his report in October 1996 that although staff were enthusias- tic and committed to making AYTAG successful, their public-service background did not prepare them well for the changes they were going through, nor for the relatively unstructured work environment that was currently in place. A comment by an environment-protection team leader continues this theme: ‘we feel like the rump of a benign dictatorship after the dictator had been got rid of. Floundering, unable to take decisions’. The nature of AYTAG’s work meant that it

employed a very high proportion of graduates who, in their previous organizations, had jea- lously guarded their professional expertise. The front-line professionals coming into AYTAG, and the managers that led it, were inexperienced in organizational mergers and their aftermath. In 1996 there was enthusiasm for environment- protection work throughout the organization. It was this enthusiasm that held the organization together in its early life.

Disequilibrium conditions in AYTAG

When looking at disequilibrium in AYTAG we identified six themes – which might be seen as facets (or drivers) of fluidity or instability. Their range meant that every employee was affected in some way during the period of the research. The first of these was the implementation of large- scale staff-development programmes such as the Introduction Programme, multi-skilling training for environment-protection staff and a manage- ment development programme for all managers. The second theme was the high level of

recruitment activity. There were several reasons for this: an initial shortage of professional staff, internal moves as individuals changed jobs to increase their salaries, new regulatory duties coming to AYTAG that increased its overall numbers and high levels of temporary appoint- ments in administrative positions. Implementa- tion of human-resource management initiatives was the third theme. Following on from organiza- tion-wide job evaluation and pay and grading exercises, new job descriptions, pay structures and terms and conditions were all introduced in October 1998. An appraisal and development scheme was implemented from March 1997. The first attempt to develop this into a performance-

related pay scheme in 1998 was rejected by over 80% of employees. A revised scheme was imple- mented without union agreement in March 1999. Senior management change was the fourth

theme. One director, two senior environment- protection managers and a regional scientist left. These departures were used as opportunities for restructuring, primarily affecting environment- protection teams and corporate support staff. During the research period for example, the personnel department had four directors and the finance department three. The 1997 Budget crisis was the fifth theme. In

March 1997, AYTAG recognized that it was unable to meet its tax bill and would be considerably in debt at the end of its first financial year. Discussions on funding continued until August when additional money was secured from central government. The budget crisis led to the demise of the Corporate Services Directorate. At their meeting on 23 July 1997, the Corporate Management Team agreed that the three Regio- nal Directors and the Environmental Strategy Director would have the responsibilities of Corporate Services distributed between them, except for finance. Part of the central-government agreement for giving additional funding was the recruitment of a finance director. While gathering data for this research we

became aware of a mismatch between what was formally stated as wanted and what actually happened. This formed our sixth theme. A diary extract from 27.6.97 gives a flavour of this. ‘J said he had spent six months with hydrology heads developing a strategy for flood risk. He said it was a requirement for AYTAG to do it. ‘‘I went to the CMT and they threw it out saying we don’t have to do it. I don’t know why they employ experts if all they do is ignore them’’. The mismatch between saying and doing caused conflicting priorities, arising most frequently be- tween policy advisers and the regions as opera- tional managers did not want to release resources to enable government-led strategy implementation. The variety of the disequilibrium incidents was

considerable. They affected every employee to some extent. Employees could find themselves either prevented from receiving training or spending too much time on training courses as a result of large and sudden fluctuations in budgets. They could find themselves working long hours because of recruitment shortages,

156 K. Houchin and D. MacLean

reporting to a rapid succession of line managers with different ways of working or carrying out tasks that they were told were important only to find that their work came to nothing.

Feedback processes in AYTAG

Feedback processes acting against achieving the desired new state

We identified feedback processes acting against achieving the desired new state which we grouped into seven themes.

(1) The moderation of expressed plans. Many initiatives introduced in AYTAG failed to be implemented in full, such as multi-skilling environment-protection officers, and the im- petus to achieve them tended to die away.

(2) Continuation of traditional ways of working. There were decisions taken and actions left unchallenged that indicated to employees both overtly and covertly that they could maintain their old ways of doing things. For example, inherited hierarchical structures remained and budgets were not devolved.

(3) Perceived lack of a cohesive Corporate Man- agement Team. One of the themes that continued throughout this research was employees’ views of the corporate manage- ment team, who were seen as not working as a team, having disparate views, each member working to their own agenda.

(4) Actions weakening a corporate approach. These actions included abolishing the post of corporate services director and splitting the sections he managed between other directorates, creating a corporate planning team chaired by a regional director and failure to recruit new information technology staff to the Corporate IT department.

(5) Non-acceptance of ownership and responsibility. The Introduction Programme run in 1996 highlighted concerns relating to lack of trust, ownership of issues and an inability to take action through a perceived lack of authority. These themes continued to be raised through- out the period of this research.

(6) Relationships are more important than tasks. The consultants running the 1998–1999 man- agement-development programme reported that in AYTAG the maintenance of relation- ships between people was more important

than carrying out tasks required if the organization was to achieve its stated strate- gic aims. An example is the slow implementa- tion of management information systems. When new work came to AYTAG, organiza- tion-wide systems were installed. Inherited local systems for capturing environmental data had been developed by managers who moved into AYTAG. They were incompati- ble and could not be used on an organization- wide basis. However they were not replaced.

(7) Running away from ‘Bad News’. On a number of occasions, ‘bad’ news had to be given. When this happened the ‘giver of the news’ distanced themselves from it either by taking leave or arranging for more junior staff to inform others. A high-profile example of this was the Chief Executive going on holiday in April 1997 after sending all staff an email announcing the size of the budget deficit.

Feedback processes working towards achieving the desired new state

There were many feedback processes taking place in AYTAG that encouraged the organization to move forward to new ways of working. We identified six themes in these feedback processes.

(1) Support for a strong centre and a corporate approach. The initial structure of AYTAG with its two central directorates, one controlling administration and the second directing strat- egy and policy formulation and implementa- tion provided the basis for a strong centre.

(2) Encouraging a sense of belonging to the organization. One of the aims of the 1996 Introduction Programme was to act as a unifying influence to help staff feel part of the new organization. AYTAG also encouraged a sense of belonging through its communica- tion systems: all staff had access to email and an intranet. There were regular social and environmental newsletters, and well-attended sporting and recreation clubs.

(3) Encouraging staff empowerment. Throughout the period of this research, staff were encouraged to communicate their ideas and opinions, and take on more responsibility through participating in working groups, project teams and attending corporate-train- ing and development events.

Complexity Theory and Strategic Change 157

(4) Encouraging staff with broader skills and knowledge bases. The need to change skill bases was recognized very early for environ- ment-protection staff and scientists, all of whom had previously worked only in one environmental medium. Training and devel- opment programmes were provided to all staff to achieve the changes required to skill bases.

(5) Developing the management process. The majority of AYTAG’s managers had come from small organizations where they had managed others with the same technical skills. In AYTAG they were increasingly required to act differently. Team leaders managed multifunctional teams. Managers divided their time between their operational work and contributing to the corporate development of the organization.

(6) Increased focus on delivering the business. During the research period, AYTAG became much more business-focused through its cor- porate and business planning systems, through target-setting, thorough its staff appraisal schemes and through its external communica- tions. The communication of its work to industry and to the public at large continued to expand. High-quality external communica- tions were considered essential to its business.

Emergent order in AYTAG

Statements made and actions taken at the start of AYTAG tell us about an organization that had decided it needed a strong centre to unite its members and drive it forward – one that wished to be flexible and empowering, seeking to influence as well as regulate. Early actions taken to promote this way of working included the design of the organizational structure, the Introduction Pro- gramme and the work done by the Corporate

management team thinking through the culture they wanted in the organization. The subsequent development of management competences, the management-development programme and multi- skilling training also helped take the organization towards a new equilibrium. However, over four years, the changes that these initiatives were meant to bring about did not take root and the organiza- tion did not reach the published desired state. Table 1 briefly describes the order that did emerge.

From wide spans of control to hierarchical organization

In 1996 AYTAG had a flat structure, with no more than three layers of management between the chief executive and front-line staff. The only exception to this was in its science function. From the beginning of 1998, the chain of command across the organization began to increase. Incre- mental structure changes in the operational directorates and in central administration units increased the layers of managers. Informal hierarchical arrangements began to appear in the policy teams. The reviews of the science function brought little change and the inherited organizational structures with their long chains of command were retained. Over the period of the research, AYTAG left in place hierarchical structures that it had inherited, and acted to increase hierarchy in areas where it initially had a flat structure. What emerged was a traditional hierarchical organization.

From employee flexibility to emphasis on traditional professional specialisms

Two important interlinked initial conditions for AYTAG were the skills and traditional career- development paths of its incoming staff. These employees were skilled in well-defined, single

Table 1. Emergent order

April 1996 – desired future state Actual future state – August 1999

The desired equilibrium The new equilibrium

(1) Wide spans of control (1) Hierarchical organization

(2) Emphasis on employee flexibility (2) Emphasis on traditional professional specialisms

(3) Empowered managers, delegation to its lowest point (3) Increasingly restricted managers, increase in bureaucratic procedures

(4) Emphasis on value for money (4) Emphasis on cost reduction

(5) Strong centre (5) Regional independence

(6) Influencer and Regulator (6) Regulator – target-driven

158 K. Houchin and D. MacLean

media disciplines; their career development was based around membership of single-discipline professional institutes. Flexibility was very de- pendent on multi-skilling front-line staff and building the confidence and management exper- tise of the team leaders so that they could manage multi-skilled professionals. Early actions that AYTAG took to broaden

the skills base of its workforce did little to move the organization away from its initial conditions. The influences of traditional working practices remained strong, and they were reinforced through delays to training programmes caused by the 1997 budget crisis and the introduction of new pay and grading structures in 1998, which favoured specialists. The effects of incrementally changing the structure, creating specialist teams, reducing the number of managers and increasing the number of single-discipline specialists all acted as organizational-level positive feedback, taking the organization away from its desired equili- brium. Interestingly, at the individual level, this was manifested as negative feedback, as employees sought to dampen-down organizationally sanc- tioned change in order to retain the individual practices that they brought with them to AYTAG. Three-and-a-half years after AYTAG became operational, the skills base and working practices of its scientific and environment-protection em- ployees remained much the same. The organiza- tion had limited success in shifting from its initial conditions; its workforce, with few exceptions, did not become more flexible. What emerged was an organization with a workforce that retained its traditional skill mixes.

From empowered managers to increasingly restricted managers

Looking at AYTAG’s initial conditions, we see that most of its employees came from hierarchi- cally structured organizations with managers having limited freedom to take decisions. All of AYTAG’s managers came from such organiza- tions, and were unfamiliar with other ways of operating. The organizational structure implemen- ted in April 1996 gave managers considerable freedom, but they opted out of taking control. They did not take decisions they could have taken. Instead they passed them on to higher levels. These actions were largely tacit and they created individual feedback processes. One result of this

behaviour was that the corporate management team spent much of its time on non-strategic issues and, as a result, was perceived as not giving strong direction to the organization. In the absence of direction, managers continued with individual tacit actions to dampen-down change and pursue a reversion to the conditions they had known before AYTAG. As the organization developed, managers’ freedom to operate was gradually eroded. Budgets, for example, became less devolved. The desire for empowered managers lessened, and what emerged were managers increasingly occupied in complying with proce- dures. Some felt very frustrated, while others felt more comfortable, as they had disliked the less- structured environment they had found them- selves in when they joined AYTAG.

From value for money to efficiency and cost control

Although value for money was regarded as an essential aim for AYTAG, in reality it was never pursued. The emphasis from the beginning was on the negative feedback mechanisms of effi- ciency and cost control. The yearly corporate plans made minimal reference to value for money, referring only to cost control and efficiency. It is likely that the organization perceived value for money and efficiency as the same thing. The cost control and procurement measures put in place were feedback processes that prevented the organization looking for value for money, and what emerged was an organiza- tion preoccupied with efficiency.

From a strong centre to regional independence

AYTAG’s ‘Centre’ was made up of two directo- rates – Corporate Services and Environmental Strategy. Directors and senior managers in AYTAG’s predecessor organizations who trans- ferred into AYTAG were in powerful positions in regional directorates. Prior to AYTAG, these managers had had control of their administrative and planning functions. From the beginning, the regions were quick to complain and criticize the centre. This criticism came to a head after the 1997 budget crisis, with the Corporate Services Director taking the blame for it. The collective power of the central administrative departments was weakened by giving control of both the department and of corporate planning to regional

Complexity Theory and Strategic Change 159

directors. These changes took AYTAG away from the desired equilibrium it was allegedly attempting to reach. Gradually it became acceptable for the regions

‘to do their own thing’. Regional directors were able to take independent actions such as restruc- turing and offering salaries outside personnel guidelines. These actions continued to strengthen the directorates at the expense of the organization’s centre. What emerged was an organization with a weak centre, which was unable to offer direction, and AYTAG failed to reach its desired new state.

From wanting to regulate and influence to regulator

The predecessor bodies making up AYTAG were experienced in working locally in regulatory activities. An initial condition for AYTAG was the lack of experience of the members of its corporate management team operating nationally as influencers. Non-regulatory matters were drawn together into one policy section in the Environmental Strategy Directorate. During 1996 this role began to be downplayed. Requests to appoint an education officer were turned down. Two years later, the jobs of the employees in this section were downgraded in relation to policy sections covering regulatory functions. Reference to AYTAG’s influencing role was

made in corporate plans. However by the time the 1998/1999 plan was produced the organiza- tion was still stating ‘AYTAG is also increasingly seeking to develop policies to influence the approach of others’ implying that very little development had occurred. In August 1997, when the production of the corporate plan was put under the control of a regional director, the emphasis shifted to target-setting for the regula- tory functions, but there were no targets for its influencing role. Each year in the annual reports there was less

emphasis on reporting AYTAG’s educating and influencing role. The way in which AYTAG was funded made it difficult to increase spending on non-regulatory issues. Increased grant-in-aid was not available to fund non-regulatory activities so they had to be met by reorganizing and reprioritizing other activities, which AYTAG did not do. AYTAG did not shift significantly away from its initial regulatory condition, and what emerged was an organization that was

primarily a regulator with its influencing role underdeveloped.

The effects of anxiety on emergent order

For some employees the move to AYTAG was exciting; for others it was an anxiety-inducing experience and they took action to attempt to reduce the anxiety they felt. The anxiety-reducing measures had two aspects: first, they were actions taken in order to feel in control; second, they were attempts to avoid conflict. Anxiety-reducing activity created individual negative-feedback pro- cesses, which attempted to bring the organization back into a state of equilibrium as experienced by individuals in previous organizations. Negative feedback reduces connectivity be-

tween the elements in a system. AYTAG’s anxiety-reducing measures all gave rise to a reduction in connectivity, as they progressively reduced the quantity and diversity of interaction between individuals. Initiatives in AYTAG gra- dually involved fewer and fewer people. Working groups became smaller, training and develop- ment more localized. Such actions strengthened anxiety-reduction. As a result, change initiatives to bring about new ways of working and suitably supporting styles of management were minimized or died out altogether.

Order emerging at the boundary

When we looked closely at where order emerged in AYTAG, we found that it was at the boundary between the organization’s legitimate and sha- dow systems. In this respect our work echoes that of Stacey (1993, 1995, 1998), who has highlighted the importance of this boundary. The legitimate system tries to pull the organization towards a new order; it is the source of stability. The tacit processes of the shadow system, in contrast, are the source of instability. In other words, the boundary between stability and instability is the boundary between the shadow and legitimate systems. AYTAG was sensitive to its initial conditions.

In the legitimate system, these were to do with its formation, structure, remit and the appointment of staff. In the shadow system there was a mixture of anxiety and excitement. There were memories

160 K. Houchin and D. MacLean

of a prior organizational life and pre-existing networks – some strengthened, others weakened as a result of the positions key staff were given. Disequilibrium was found in the legitimate system, brought about by the expressed desire to create an organization considerably different from any of its predecessors. The shadow system of the organiza- tion was closer to a state of equilibrium, as there was scope for individuals to retain old working practices and networks. In the legitimate system, negative feedback operated to drive the organiza- tion towards a new order. Feedback in the shadow system was negative when viewed at the level of the individual, as employees attempted to get back to the way in which they worked previously. At the level of the organization, however, that same shadow-system feedback could be viewed as positive, a source of instability, as it took the organization away from its desired new state. As AYTAG developed, what emerged was a

different organization from that which was originally envisaged. Its published aims were not realized. The former director of Corporate Services made the comment after structural changes following the 1997 budget crisis ‘the window of opportunity is shutting. Our niche will be regulation. AYTAG’s an immature organization. When it settles it may be very stodgy’. What emerged was a hierarchical, regu- latory organization, retaining traditional profes- sional specialisms, rather than a flexible, influen- cing organization with broad-based professionals and empowered managers.

Discussion

Earlier in this paper we briefly commented on existing strategy theory, suggesting that a more holistic approach is required for the study of strategic organizational development. We can now examine how complexity theory enables us to describe the development of AYTAG. It has helped us create a rich description of its devel- opment and allowed us to make use of and interpret a wide range of data within the one framework. It has enabled us to take a holistic approach to studying the organization, and to include issues often ignored by the more tradi- tional approaches. Using key concepts of complexity theory to

describe the development of AYTAG has identi-

fied its emergent properties and provided an explanation of why these properties developed in the way they did. The emergent properties apply to the AYTAG system as a whole. It is these properties that shaped AYTAG’s dynamics, which in turn shaped its approach to strategic develop- ment. However, we also found in this study that there were difficulties in applying complexity- theory concepts to a social system. These difficul- ties justify Chia’s (1998) and Goldstein’s (2000) comments on the need for circumspection in attempting to transfer complexity-theory formula- tions from the natural to the social world. A general view put forward by complexity

theorists is that instability in systems results in positive-feedback processes dominating to bring about change, and new or novel forms of order emerge (Mitleton-Kelly, 1998; Prigogine and Sten- gers, 1984). Empirical organizational studies by Brown and Eisenhardt (1997), Shaw (1997) and MacIntosh and MacLean (1999) reflect this view. However, the present research, in contrast, shows that destabilizing a social system such as an organization does not inevitably lead to novel forms of order. Although there were continuous destabilizing activities, ranging from training pro- grammes to a budget crisis, the order emerging in AYTAG was very little different from that which had existed in its predecessor organizations. Calls for the creation of disequilibrium and

positive feedback to facilitate the emergence of novel forms of order are based largely on the study of complexity theory in physical systems (Prigogine and Stengers, 1984). Social systems are different from physical systems in their complex- ity. The rules that determine the interactions in social systems are socially constructed, and are not fixed by laws of nature (Berger and Luck- mann, 1966; Weick, 1979). Human agency can, for example, alter the parameters and structures of social systems (Levy, 1994). In a social system, everyone has a psychological state, and this has to be factored in when considering the behaviour of that system. We found that negative indivi- dual-level shadow-side feedback leads to the social system attempting to return to a previous order. That same feedback is perceived as positive at a higher level. Thus, when infused with human issues such as memory and desire, the role and nature of feedback, and the distinction between positive and negative var- iants, become somewhat blurred.

Complexity Theory and Strategic Change 161

Critical to our understanding of emergent order in organizations is the role of anxiety and the actions taken to reduce it by individuals and groups (Stacey, 1993). Anxiety is present as a general sense of unease. It is generated when change is indicated or implemented and it is found in the shadow system. Strategy theorists in general focus their theories on the legiti- mate system. With the exception of Stacey (1993, 1995, 1998) and Shaw (1997), the shadow system is largely neglected by them. In order to study it, it is necessary to draw on theories that do not have their roots in physical systems. Such theories usually come from the behavioural sciences. Destabilizing activities and actions taken to

reduce anxiety were continuing features of AYTAG’s development. Both can be traced back to the organization’s initial conditions. This finding caused us to consider that the concept of disequilibrium as described by complexity theorists may be less helpful in facilitating the emergence of novel forms of order in a social system. Novel forms of order will not necessarily emerge as a result of destabilizing events. We need another way of visualizing the concept of disequilibrium if it is to be helpful to our understanding of the behaviour of complex adaptive social systems such as organizations. Complexity-theory researchers, for example

MacIntosh and MacLean (1999), applying the theory of complexity to the analysis of social systems describe disequilibrium as a dynamic between stable and unstable states. We offer a possible alternative view of disequilibrium in complex social systems such as organizations. As a result of this research we would like to suggest that this dynamic can be thought of as anxiety. Indeed, in such systems, disequilibrium and anxiety can be considered synonymous. It is the presence of anxiety that keeps a complex social system in disequilibrium. The causes of anxiety or disequilibrium are

found in the organization’s legitimate system which in AYTAG included changes to working practices, reporting lines and organization-wide development programmes. When organizations make changes such as restructuring, merging, shifting skill bases this causes anxiety levels to rise or expressed in complexity-theory terms causes disequilibrium in the system. Anxiety is something that human beings try to

get rid of and they seek to replace it with the

equilibrium of comfort and security. Menzies (1960) and Argyris (1992) are among the authors who have written on how humans working in organizations build defences against anxiety. The natural tendency for humans is to seek equili- brium and this can be done by reducing anxiety through conflict avoidance, taking actions to retain control and minimising change (Janis and Mann, 1977). These activities give rise to individually-driven negative feedback and reduce connectivity between the individual elements in the social system. Disequilibrium is then dis- placed as the system tries to regain equilibrium. Actions to dampen-down anxiety that move the organization towards an equilibrium position are found initially in the shadow system; they are usually tacit and at an individual level as they were at first in AYTAG (Egan, 1994). All systems including social systems are subject

to constraints. Stacey, Griffin and Shaw (2000) draw on the work of Elias and Scotson (1994) to introduce the constraints placed on group mem- bers in social relationships. Elias and Scotson state that a basic social impulse is the creation and maintenance of power differentials. This theme is picked up by Streatfield (2001), who argues that power is a constraint that excludes some communicative actions and includes others. Reducing connectivity between elements in a social system results in an exclusion of some communicative actions and so leads to power differentials. Power differentials in organizations are expressed as hierarchy.

Are organizations naturally complex adaptive systems?

Authors such as Anderson (1999) and Mitleton- Kelly (1998) argue that one of the major insights that complexity theory brings to strategy theory is that the organization can be viewed as a non- equilibrium system. As noted earlier, viewed from this stance, the concept of the organization moving from one stable state to another as a result of change is flawed. However, the present research leads us to question this view. Our suggested proposition is that disequilibrium is anxiety; organization is a defence against anxiety, and organizations thus tend towards stability. Para- doxically, though organizations can self-organize, from the perspective of this case study when they

162 K. Houchin and D. MacLean

do so, they tend to retreat from the very fluidity which enables self-organization. In the state towards which they are attracted, they are not naturally occurring complex adaptive systems such as are studied in the natural sciences – if anything they are our defensive responses to such systems. Our research findings indicate that the ‘natural’

tendency of a complex social system is the creation of equilibrium rather than novelty. When disequilibrium is present in the form of anxiety, the shadow system initiates the creation and maintenance of power differentials and self- organizes into hierarchy to reduce anxiety and suppress novelty (Stacey, 2000). The desire to reduce anxiety produces patterns of stability, first within the organization’s shadow system. Such patterns include the repetition of behaviours which are present in an organization’s initial conditions. As these patterns stabilize they become part of the legitimate as well as the shadow system, and the drive back towards a previously recognizable equilibrium is reinforced. In AYTAG, what initially began as individual actions, such as managers opting out of taking decisions, resulted in organizational action that took AYTAG away from having a devolved structure with a multi-skilled, professional work- force, back towards a traditional, hierarchical organization with single-discipline professionals. In social systems, hierarchy is the result of self-

organizing, and anxiety-reduction actions are entirely natural emergent outcomes. The sugges- tion here that, in the absence of anxiety and power differentials, organizations would exist as complex adaptive systems is tantamount to sug- gesting that organizations would exist in the absence of the fundamental processes of organizing! The concept of an organization as a natural

complex adaptive system may well be a myth. We are suggesting that equilibrium expressed as a desire for harmony and security is the norm in social systems as they seek to avoid anxiety. Complex systems made up of human elements tend not to exist away from equilibrium. Such systems are not pushed towards chaos when the system is in disequilibrium. They tend to grav- itate towards a previously experienced equili- brium position. This is the exact opposite of what complexity theorists claim about living systems (Brown and Eisenhardt, 1997; Pascale, 1999). Nevertheless, this is what happened in AYTAG as it tried to return to previous operating styles

that had been in place in its predecessor organi- zations. Stacey, Griffin and Shaw (2000) and Streatfield (2001), in acknowledging that hierarchy is an emergent property of organizations, appear to be moving to a position which corresponds with our observations that organizations stabilize over time. However, rather than viewing organizations as complex adaptive systems, following the lead of Brans and Rossbach (1997), we can draw on the work of Nicholas Luhmann (1982) and regard them as complex recursive systems, continually trying to reproduce themselves in the same way. Our understanding from the study of living organisms tells us that the majority only experi- ence short-term survival. They do not adapt and this is may also be true for organizations.

The circularity of key complexity- theory concepts

Tsoukas and Hatch (2001) argue that interpreting the features of complex systems from a complex- ity theory perspective is dependent upon the position of the interpreter. From our position, we encountered problems in defining and trying to separate different key complexity-theory con- cepts. There was a degree of circularity as initial conditions, disequilibrium and feedback all ap- peared as facets of one dynamic. It was difficult to define what exactly is meant by initial conditions. The actual circumstances created by the formation of the organization such as its designed systems mattered less than the precon- ditions that existed as memories. The latter formed a legacy which dominated the initial conditions. We found that connectivity drove feed- back, which in turn influenced connectivity. Work- ing at the boundaries of different organizational systems was essentially working on connectivity. Anxiety produced connectivity changes which altered feedback mechanisms. Anxiety-reducing actions caused people to withdraw, which re- duced connectivity and created negative feed- back. This in turn further reduced connectivity, so that connectivity and feedback appeared to be different sides of the same coin. The unifying phenomenon of anxiety was different from what we expected from the complexity-theory litera- ture. This may not have come to light previously, as a result of there being relatively few examples of empirically rich studies.

Complexity Theory and Strategic Change 163

The need for complexity theory to be informed by psychology and social theory

This study of AYTAG shows that complexity theory in its current form is only partially successful as a device for describing organiza- tional development and change. Our study has highlighted limitations in using the concepts of complexity theory to describe human systems. We recognize there are many questions still to be answered for the theory to be as useful as some of its advocates would suggest. Unlike physical systems human systems have histories that cannot be ignored. They contain psychological drivers and do not necessarily tend towards chaos. The study of psychological drivers in management writings on complexity-theory ap- plications is underrepresented, with the exception of the work of Stacey (1996, 2000) and those who work closely with him (Streatfield, 2001; Stacey, Griffin and Shaw 2000; Shaw, 1997). For com- plexity theory to be really useful to management practice it has to move away from its reliance on exemplars from natural-science systems and embrace theories and principles from psychology and social theory.

Complexity theory and strategic change

Earlier in this paper we suggested that complexity theory could offer a different conceptual frame- work for the study of strategic change. As a result of this study we argue that organizations may not be naturally-occurring complex adaptive systems. We also argue that a behavioural-sciences dimen- sion has to be incorporated into complexity theory for it to be of greater use. The theory has provided us with a different perspective on how order emerges at the interface between the shadow and legitimate systems. This has not been done previously with such empirical detail. We have suggested that organizations stabilize over time, as stabilization is one of the primary defences against anxiety. Defensive behaviours, stabilization and inertia

are recurrent themes in the strategy and organi- zational-behaviour literature, but less attention is given to them than strategic change (Greenwood and Hinings, 1988; Weick, 1979). Concepts developed by Argyris (1992) such as single-loop

learning and defence routines and Menzies’ (1960) study of social defences offer explanations of how we attempt to limit change and remain stable. This body of work and that of Hirschorn (1990) and Kets de Vries and Miller (1994) on the psychodynamics of organizations helps us gain insight into why organizations stabilize. In the strategy literature there are models and ap- proaches which focus on a continuation or a return to stability. Miller and Friesen (1984) suggest that most organizations gravitate to- wards a ‘design archetype’ and remain there for a lengthy period of time. Greenwood and Hinings (1998) propose that organizations use ‘interpre- tive schemes’ to move along an organizational ‘track’, which can help explain the absence of change. Johnson (1988) writes about the persis- tence of cultural paradigms that bring about resistance to change. Our single case study is insufficient, and what we would suggest is further empirical study on stabilization and novelty in organizations to allow direct comparisons to be made between what can be revealed through the lens of complexity theory and through other approaches to strategy. Whittington’s (1993) framework with its four conceptions of strategy, which we briefly drew attention to at the beginn- ing of this article, may be a good place to start.

Conclusions

Complexity theory as it is being used is inade- quately sensitive to the dynamics of individual and group behaviour, which are the key to understanding the detailed phenomena of strate- gic change. When it is used to describe the pro- cesses of strategic change, its specific usefulness is still unclear. As a metaphor, it offers the possibility of a rich set of concepts to explore the phenomenon of emergence in organizational systems, particularly at a macro level. However, more empirical research is needed in organiza- tions, as without this complexity theory is in danger of becoming a short-lived linguistic fashion statement. This might then deprive us of the promise it shows as an alternative and integrative set of ideas and insights on the dynamics of pattern development. Complexity theory applied to organizations

has to be more seriously informed by our understanding of the dynamics of human beha-

164 K. Houchin and D. MacLean

viour in order to take account of a range of familiar issues such as socially constructed rules, anxiety and other forms of embodied expression. The role of such phenomena has to be incorpo- rated into complexity-theory concepts to help explain the creation and evolution of patterns of interaction. This will enhance the attractions of complexity theory as an integrative framework for understanding strategic change. In this regard, an increased level of empirical inquiry is much needed; until such time as it materializes and yields fruit, our experience would suggest that normative statements on management based on complexity theory would be better framed as conjecture.

References

Anderson, P. (1999). ‘Complexity Theory and Organization

Science’, Organization Science, 10(3), pp. 216–232.

Andrews, K. R. (1987). The Concept of Corporate Strategy. (3rd

Edition). Irwin, Homewood, IL.

Ansoff, H. I. (1984). Implementing Strategic Management.

Prentice Hall, Hemel Hempstead.

Argyris, C. (1992). On Organizational Learning. Blackwell,

Oxford.

Battram, A. (1998). Navigating Complexity. The Industrial

Society, London.

Beinhocker, E. D. (1999). ‘Robust Adaptive Strategies’, Sloan

Management Review, 40(3), pp. 95–106.

Berger, P. L. and T. Luckmann (1966). The Social Construction

of Reality. Penguin, Harmondsworth.

Brans, M. and S. Rossbach (1997). ‘The Autopoiesis of

Administrative Systems: Niklas Luhmann on public admin-

istration and public policy’, Public Administration, 75,

Autumn, pp. 417–439.

Brown, S. L. and K. M. Eisenhardt (1997). ‘The Art of

Continuous Change: Linking complexity Theory and Time-

paced Evolution in Relentlessly Shifting Organizations’,

Administrative Science Quarterly, 42, pp. 1–34.

Buzan, T. (1991). The Mind Map Book. Penguin, New York.

Chia, R. (1998). ‘From Complexity Science to Complex

Thinking: Organization as Simple Locations’, Organization,

5(3), pp. 341–369.

Cilliers, P. (1998). Complexity and Postmodernism. Routledge,

London.

Colado, E. I. (1995). ‘Strategic Analysis of Organizations: A

model from the Complexity Paradigm’, Human Systems

Management, 14(1), pp. 51–70.

De Wit, B. and R. Meyer (1998). Strategy Process, Content,

Context. (2nd Edition). West Publishing Company, St Paul.

Easterby-Smith, M., R. Thorpe and A. Lowe (1991). Manage-

ment Research – An Introduction. Sage, London.

Egan, G. (1994). Working in the Shadow Side: a guide to positive

behind the scenes management. Jossy Bass, San Francisco.

Eisenhardt, K. M. (1989). ‘Building Theories from Case Study

Research’, Academy of Management Review, 14, pp. 488–511.

Elias, N. and J. Scotson (1994). The Established and the

Outsiders. Sage, London.

Gemmill, G. and C. Smith (1985). ‘A Dissipative Structure

Model of Organization Transformation’. Human Relations,

38, pp. 751–766.

Gersick, C. J. G. (1991). ‘Revolutionary change theories: A

multilevel exploration of the punctuated equilibrium para-

digm’, Academy of Management Review, 16, pp. 10–36.

Gilmore, T. N. and J. Krantz (1985). ‘Projective Identification

in the Consulting Relationship: Exploring the Unconscious

Dimensions of a Client System’, Human Relations, 38, pp.

1159–1177.

Goldstein, J. (2000). ‘Emergence: A construct amid a thicket of

conceptual snares’, Emergence, 2(1), pp. 5–22.

Greenwood, R. and C. R. Hinings (1988). ‘Organizational

Design Types, tracks and dynamics of strategic change’,

Organization Studies, 9(3), pp. 293–316.

Griffin, D., P. Shaw and R. Stacey (1998). ‘Speaking of

complexity theory in management theory and practice’,

Organization, 5, pp. 315–339.

Hirschorn, L. (1990). The Workplace Within: Psychodynamics

of Organizational Life. MIT Press, Cambridge, MA.

Janis, I. L. and L. Mann (1977). Decision Making: a

psychological analysis of conflict, choice and commitment.

Free Press, New York.

Johnson, G. (1988). ‘Rethinking Incrementalism’, Strategic

Management Journal, Jan/Feb, pp. 75–91.

Kets de Vries, M. F. R. and D. Miller (1994). The Neurotic

Organization, Josey-Bass, California.

Langley, A. (1999). ‘Strategies for Theorising from Process

Data’, Academy of Management Review, 24(4), pp. 691–710.

Levy, D. (1994). ‘Chaos Theory and Strategy: Theory,

application and managerial implications’, Strategic Manage-

ment Journal, 15, pp. 167–178.

Lissack, M. (1997). ‘Mind your metaphors; Lessons from Com-

plexity Science’, Long Range Planning, 30(2), pp. 294–297.

Luhmann, N. (1982). ‘Autopoiesis, Handlung und kommuni-

cative Verständigung’, Zeitschrift für Soziologie, 9(1), pp.

5–17.

MacIntosh, R. and D. MacLean (1999). ‘Conditioned emer-

gence: a dissipative structures approach to transformation’,

Strategic Management Journal, 20, pp. 297–316.

McKelvey, B. (1999). ‘Complexity Theory in Organization

Science; Seizing the Promise of Becoming a Fad’, Emergence,

1(1), pp. 5–32.

McKelvey, B. (1997). ‘Quasi-natural Organization Science’,

Organization Science, 8(4), pp. 352–375.

McKergow, M. (1998). ‘The East Asian Economic and

Financial Crisis: Lessons from Complexity Theory’. APEC

Economic Outlook Symposium, Xiamen, China, 16–17 May.

Menzies, I. E. P. (1960). ‘A Case Study in the Functioning of

Social Systems as a Defence against Anxiety’, Human

Relations, 13, pp. 95–121.

Miller, D. and P. Friesen (1984). Organizations: a quantum view.

Prentice Hall, New Jersey.

Mintzberg, H. (1988). ‘Opening up the definition of strategy’. In

J. B. Quinn, H. Mintzberg and R. M. James (eds), The

Strategy Process. Prentice Hall, New Jersey.

Mintzberg, H., B. Ahlstrand and J. Lampel (1998). Strategy

Safari. Prentice Hall, Hemel Hempstead.

Mitleton-Kelly, E. (1998). Organisations as Complex Evolving

Systems. OACES Conference, Warwick, 4–5 December.

Morel, B. and R. Ramanujam (1999). ‘Through the Looking

Glass of Complexity: The Dynamics of Organizations as

Complexity Theory and Strategic Change 165

Adaptive and Evolving Systems’, Organization Science, 10(3),

pp. 278–293.

Morgan, G. (1997). Images of Organization. (2nd edition). Sage

Publications, California.

Parker, D. and R. Stacey (1994). Chaos, Management and

Economics: The implications of non-linear thinking, Hobart

Paper, Institute of Economic Affairs, London.

Parkhe, A. (1993). ‘‘Messy’ research, methodological predis-

positions, and theory development in international joint

ventures’, Academy of Management Review, 18, pp. 227–268.

Pascale, R. T. (1999). ‘Surfing at the edge of Chaos’, Sloan

Management Review, 40(3), pp. 83–94.

Pettigrew, A. (1997). ‘What is Processual Analysis?’, Scandina-

vian Journal of Management, 13, pp. 337–348.

Porter, M. E. (1980). Competitive Strategy: Techniques for

analysing industries and competitors. Free Press, New York.

Prahalad, C. K. and G. Hamel (1994). ‘Strategy as a field of

study; why search for a new paradigm’, Strategic Manage-

ment Journal, 15, pp. 5–16.

Prigogine, I. and I. Stengers (1984). Order out of Chaos; Man’s

new dialogue with Nature. Bantam Books, New York.

Reason, P. and B. C. Goodwin (1999). ‘Towards a science of

qualities in organizations: lessons from complexity theory

and post modern biology’, Concepts and Transformations,

4(3), pp. 281–317.

Rosenhead, J. (1998). ‘Complexity Theory and Management

Practice’, Human Nature Review, http://human-nature.com/

science-as-culture/rosenhead.html, accessed 22 November 2001.

Schein, E. H. (1992). Organizational Culture and Leadership

(2nd Edition). Jossey-Bass, San Francisco.

Seel, R. (2000). ‘Culture and Complexity: New Insights on

Organisational Change and Organization Development’,

Organisations and People, 7(2), pp. 2–9.

Shaw, P. (1997). ‘Intervening in the shadow systems of

organizations – Consulting from a complexity perspective’,

Journal of Organizational Change Management, 10, pp. 235–

250.

Smith, C. and G. Gemmill (1991). ‘Change in the small group: a

dissipative structure perspective’, Human Relations, 44(7), pp.

697–716.

Stacey, R. D. (1991). The Chaos Frontier, Creative strategic

control for business. Butterworth Heinemann, Oxford.

Stacey, R. D. (1993). ‘Strategy as order emerging from chaos’,

Long Range Planning, 26(1), pp. 10–17.

Stacey, R. D. (1995). ‘The Science of Complexity; an alternate

perspective for strategic change processes’, Strategic Manage-

ment Journal, 16, pp. 477–495.

Stacey, R. D. (1996). Complexity and Creativity in Organiza-

tions. Berrett-Koehler, San Francisco.

Stacey, R. D. (1998). Creativity in Organisations: The impor-

tance of mess. Complexity and Management Papers, No I,

University of Hertford.

Stacey, R. D. (2000). Strategic Management and Organisa-

tional Dynamics. (3rd edition). Pearson Education Ltd,

Harlow.

Stacey, R. D., D. Griffin and P. Shaw (2000). Complexity and

Management: Fad or radical challenge to systems thinking?

Routledge, London.

Streatfield, P. J. (2001). The Paradox of Control in Organiza-

tions. Routledge, London.

Tsoukas, H. (1998). ‘Introduction: Chaos, Complexity Theory

and Organization Theory’, Organization, 5, pp. 291–314.

Tsoukas, H. and M. J. Hatch (2001). ‘Complex thinking,

complex practice: The case for a narrative approach

to organizational complexity’, Human Relations, 54(8), pp.

979–1013.

Weick, K. (1979). The Social Psychology of Organizing.

Addison-Wesley, Reading, MA.

Whittington, R. (1993). What is Strategy and Does it Matter?

Routledge, London.

Whittington, R., A. Pettigrew and H. Thomas (2002). ‘Con-

clusion: Doing More in Strategy Research’. In A. Pettigrew,

H. Thomas and R. Whittington (eds), Handbook of Strategy

and Management. Sage, London.

Wood, R. (1999). The Future of Strategy: The Role of the New

Sciences Part One. Paper for the first Conference of the New

England Complex Systems Institute on Organizations and

Management.

Wood, R. (2000). The Future of Strategy: The Role of the New

Sciences Part One. Paper for the first Conference of the New

England Complex Systems Institute on Organizations and

Management.

Yin, R. K. (1993). Applications of Case Study Research. Sage,

Newbury Park, CA.

Kate Houchin is responsible for the curriculum in management, human resource management, trade union studies, training and development at Stevenson College Edinburgh. Before joining the College her career was in public sector organizational development and training. She has worked for local authorities and for the Scottish Environment Protection Agency. She holds a masters degree in Human Resource Development and a PhD from the University of Glasgow.

Donald MacLean received his PhD from the University of Cambridge and spent ten years working in the optoelectronics industry before joining the University of Glasgow in 1993. A senior lecturer in strategic management, his research interests lie in the development of novel theories and methods of/for organization, strategic management and transformation. He co-founded SCOT, the Scottish Centre for Organization and Transformation, which conducts collaborative research with a broad range of public and private sector organizations.

166 K. Houchin and D. MacLean