Author: Stacey, Ralph D.
Title: Complexity and creativity in organizations
Publisher: Berrett-Koehler
Year: 1966
Links: https://worldcat.org/en/title/940537303
open access: https://archive.org/details/complexitycreati0000stac
Introduction
Part One:
The Complex Nature of Human Networks 19
This book claims that the science of complexity offers a comprehensive new framework for drawing together a number of already existing insights about human systems into a coherent theory of organizational evolution that is dramatically different from the one currently dominating our thinking. The first step in sustaining this claim must, therefore, be to demonstrate that human systems are indeed the kind of system that the science of complexity deals with. That is the purpose of this part of the book. In Map C the central message of Part One is displayed as an expanded step 1a.
The introduction has already briefly defined the science of complexity. It is the study of systems consisting of large numbers of agents who interact with each other to produce adaptive survival strategies for themselves and hence for the system, or parts there of, that they belong to. Their system in turn interacts with others, making up a larger suprasystem in which they are agents that coevolve together. The total system, therefore, has a holographic or fractal aspect in which the parts interact continually to recreate the whole and the whole affects how the parts interact. (A fractal is a pattern that is repeated in a self-similar way at many different levels. The pattern is self-similar in that it is always recognizable but never exactly the same. It is a pattern that is repeated in an irregular way.)
Stacey_1966_OrgsComplexAdaptiveSystems.png
Throughout, the interaction between agents and systems is of a nonlinear form in which feedback on the consequences of behavior is used to construct models of the world from which rules of conduct, or schemas, are extracted. These schemas are then changed in the light of further behavior to produce more adaptive behavior. We are now talking about learning systems. In the following chapters, we will consider the sense in which human systems are all nonlinear feedback networks and the sense in which they are fractal, or self-similar, in nature.
It is immediately evident that the theory of organizational dynamics is derived in previous chapters from the science of complexity and the discipline of psychoanalysis has major implications for the practice of management. These implications are often experienced as anxiety-provoking because they challenge so many deeply held beliefs about what management ought to be, even if it is clear that it has not yet reached that ideal state. The following are typical concerns:
These responses and the questions they give rise to display how tightly the managers concerned are thinking within the same dominant schema as the one that is employed by the research community. Success is held to be a stable, predictable state that can be secured by comprehensive, prior, shared intention. Those who propose that instability and mess are essential are simply saying that coherent behavior is impossible and, therefore, what they are saying must be wrong. Consider, however, how one might respond to each of the concerns raised above from the complex system perspective, which indicates that not only coherent but also creative behavior is possible, and that it is possible precisely because the system is messy.
Criteria for Designing Actions in Conditions Far from Certainty
Members of an organization, whether they are in the role of manager or managed, all face much the same fundamental issue. They must take action and respond in the here and now to the consequences that the actions of others have just created for them. If individuals are to act coherently as a group, each of them has to answer a key question: What criteria are to be used in choosing the next action?
The dominant schema immediately points to one major criterion for selecting the next action, namely, the extent to which long-term outcomes of that action achieve some objective determined in advance of acting. The design of an action requires that people first identify the potential consequences of each action option open to them individually and as a group, then select the one that is reasonably likely to achieve the objectives of their organization. However, as soon as they try to do this, they are confronted with a problem: as they look into the future, trying to identify potential consequences, they find it harder and harder to do so the further they try to look into the future. At some point, and nowadays that point is not all that far in the future, they have to admit that they do know what the consequences of their actions might be, and the future becomes open-ended.
The dominant schema then prompts them to make assumptions about what they do not know so that the rational design of actions may proceed. This schema leads people to believe that the open-endedness they face is the result of their ignorance: they have not done enough research into the laws of causation and have not gathered enough data to work out the future consequences of their actions. Or the data and the analysis are available but others are either too incompetent or too badly behaved to use them properly. The immediate conclusion drawn is that ignorance can be overcome by greater investment in gathering information, funneling it to some central point where it can be analyzed, and then feeding it back to the actors. The dominant schema therefore leads people to believe that ignorance can be overcome by research into organizational excellence, incompetence can be overcome by training and developing managers, and systems can be used to prevent bad behavior. All of this is expected to enable members of an organization to be clear enough about the consequences of their actions, at least in a probabilistic sense or in terms of assumptions, so that these consequences may provide the necessary criteria for choosing their actions. The future is not unknowable; it merely is currently unknown.
From the complexity perspective, however, we reach the opposite conclusion, namely, that the future is truly unknowable.
Creative futures emerge unpredictably from self-organizing interactions between members; therefore, they clearly cannot use some forecast of long-term outcomes to decide between one action and another. Furthermore, the complexity perspective would lead us to believe that because the identification of long-term outcomes is impossible, successful practitioners will use other criteria while, perhaps, pretending that they are not doing so.
For example, during a consultation to one organization, I attended a meeting of the senior management group, who were discussing whether or not to invest many millions in a new plant. They had before them a paper setting out twelve different scenarios for the next twenty-five years, together with the cash flows associated with investment in each of the twelve scenarios and the rates of return on their investment that these implied. In doing this they were, of course, following standard best-practice investment appraisal procedures. Soon into the discussion of the paper, however, confusion arose as to just what the differences were between one scenario and another; even the experts had forgotten how the huge number of assumptions made in each scenario differed from those made in the others. The confusion began to turn to anger, so the chief executive relieved the pressure by laughingly saying, "The one thing we know for sure is that none of these scenarios contain what will happen. We all know from past experience that reality will turn out to be completely different."
These intelligent and competent managers were using a procedure that is widely held to be rational and entirely appropriate in the circumstances. The very act of doing this, however, was completely irrational. Rationality requires that each individual expose, for joint scrutiny, his or her reasons for undertaking a joint action. In this way the reasons for acting jointly in a particular manner can be discussed and tested by the knowledge and experience of other players. When I asked the senior executives at this particular meeting whether they had already decided to back the proposed investment, most said that they had, and that they had reached their conclusion before they saw the paper on scenarios and rates of return. When they were asked if they had disclosed and discussed their reasons to each other, they replied that they had done so in casual conversations in twos and threes up and down the corridors. But now, in their formal, public meeting, they were acting as if they were seriously considering a decision based on the figures before them, and they had no intention of publicly discussing the real reasons.
Why not? Because those reasons had nothing to do with the long-term outcome and the profitability of the investment, and yet this is the basis upon which rational managers are supposed to make their decisions. The real reasons had to do with staying in the marketplace even though no one knew how it would develop, and with preventing competitors from building too strong a position. In fact, then, the real reasons for deciding to undertake the investment related to the here-and-now characteristics of the action itself -- the means, rather than the ends.
When asked why they were pretending to make the decision in one way when they had already effectively made it in another way, they indicated that they were not aware of doing this and then said that the real reasons were so general, and experience-based gut feeling is the pejorative for this, that they could not hope to persuade the nonexecutive members of their board and the financiers that the investment should go ahead. However, close questioning of nonexecutive board members and financiers revealed that they did not believe the forecasts any more than the managers making them did. The apparently rational method, however, protected them if the investment subsequently failed. The real criterion for making the finance available was the track record of the management team and the financiers' gut feeling. In this case and, I suggest, in all other cases in which discounted cash flow analysis is used, managers are acting out an apparently rational process, claiming that they are using a rational technique, when in fact, if they were really doing this, they would be behaving entirely unreasonably, because they would be making major decisions on the basis of information that they clearly knew to be worthless. However, they are being reasonable, because they are using the apparently rational as a cover behind which they are doing something quite real. The purpose of the cover is to reduce the anxiety levels that arise from dealing with the unknowable.
So instead of concerning themselves with the ends, which they cannot determine in advance, managers are compelled to judge what it is appropriate to do entirely on the basis of the means. The criteria for quality actions become, not ends, but ethical considerations and criteria having to do with maintaining positions, keeping options open, retaining flexibility, and revealing errors as fast as possible. The quality action is not one with a predetermined outcome, because that effectively excludes all creative actions, but the action that is morally good in itself, the action that keeps options open by allowing an organization to stay in the game and not yield to competitors, the action that allows managers to detect their errors as soon as possible.
Consider the revolution that this last criterion implies. At present it is common to hear top executives saying to their subordinates, "Make sure that I do not have to deal with surprises." This instruction quite naturally leads subordinates to cover up their errors until it is too late, a procedure that is disastrous in conditions that are far from certainty (Schwartz, 1990). If, instead, senior executives say, "Make sure that you surprise me as early and as frequently as possible," they change the criteria for a quality action to include one that allows the rapid uncovering of errors (Collingridge, 1980). For such a change to occur, managers will have to develop a tolerance for errors.
Note how the complexity perspective facilitates awareness of the distinction between espoused and in-use management theories of action and assists in understanding why managers do what they actually do instead of what the management literature suggests that they should. The complexity perspective indicates that many of the messy processes that managers employ behind the cover of technical rationality are entirely appropriate. In that sense, the complexity theory of organizations does not present anything new, anything that is not happening or has not happened. What it does present is a more comprehensive and more useful way of making sense of what managers actually do. In other words, it is a useful framework for organizational self-reflection. It is an approach that assists managers in reflecting upon automatic behaviors that are driven by rules lying beneath the level of awareness. Such reflection is the essence of double-loop learning, without which no new knowledge is created.
Freedom and Control in the Space for Creativity
The dominant schema encourages managers to assume the following: Quality actions are based on some knowledge of their outcomes.
Such knowledge is best obtained from some centralized point to which data has been funneled and where it is then analyzed before being fed back to those who must act. At least the objectives, if not the actions themselves, need to represent either the intention of a small number of the most powerful or the democratic intention of the whole group.
The result of thinking in this way is that most members of an organization are left with very little individual freedom. It is the role of the majority of the members to implement the actions that are likely to achieve the outcomes intended by the most powerful or by the majority.
From this perspective, legitimate authority flows from the roles occupied by the powerful, who may or may not be democratically elected. That authority will, of course, be diminished if the members of an organization take their own authority; it is believed that disorder would ensue if large numbers of individuals acted according to their own local rules of behavior in the absence of some blueprint. Thus, the notion that all members must act only to achieve some predefined outcome not only obstructs creativity and encourages us to avoid actions with long-term outcomes; it is also an enormous constraint on individual freedom in an organization.
However, the science of complexity shows that complex adaptive systems produce order for nothing, that is, without any blueprint, plan, or set of instructions being fed to them. In organizational terms, this means that we do not need to fear that when individual members take their own authority for their actions it will automatically lead to anarchy. Taking our own authority does not mean doing anything we feel like. Authority is the legitimate use of power, so we need to look for the source of legitimacy when we take our own authority. That source lies in two locations: in the task itself and in our own humanity. Taking our own authority means taking the steps necessary to accomplish the task, as dictated by that task rather than by some figure in an authority role, within the internal constraints we set ourselves for ethical behavior. The constraints lie in what we believe to be right and in the need we all have to sustain the support of those we interact with.
Mature members of an organization who take their own authority for performing the tasks of the organization, that is, who self organize, will not produce anarchy but may well produce creative new strategic directions.
There are, then, two diametrically opposed ways of answering some of the most basic questions to be posed about organizational life. One way is based on the belief that order is put into organizational behavior by prior shared intention, by either the most powerful or the majority. This leads to ways of managing and organizing that greatly constrain individual freedom -- the few do the thinking and the creating while the many do as they are told. The opposite view is based on the notion that a creative new order emerges unpredictably from spontaneous self-organization. This leads to ways of managing and organizing in which individuals are free to take their own authority, constrained only by the nature of the task, the need to sustain support, and the imperative to behave ethically. The new science of complexity offers hopeful justification for freedom in organizational life.
Leading in the Space for Creativity
The dominant schema that currently drives the thinking of both the research community and practicing managers leads to the belief that organizational joint actions must be selected according to how likely they are to achieve desirable outcomes. The desirability of an outcome is to be determined by those who occupy authority roles. Such views lead to particular notions of leadership: leaders determine and articulate the direction in which a group or organization is to develop and then employ a number of motivational methods, ranging from force to inspiration, encouragement, and facilitation, to persuade others to move in that direction.
A complex adaptive system theory of organizations does not reject such notions of leadership; rather, it puts them into a context that enables us to see that they are limited, special-case notions. They are the concepts of leadership that apply to ordinary management through the medium of the legitimate system, which is confined to single-loop learning and to making more efficient what an organization already does well. When it comes to articulating where an organization has come from and then sharpening our understanding of how it may continue in much the same direction, the notions of leadership set out above are entirely appropriate.
However, far from certainty and equilibrium, leadership has rather different meaning. In these circumstances the leaders of the legitimate system cannot know, any more than anyone else can, where their organization is going. It is extraordinary management, the self-organizing process of double-loop learning pursued in the shadow system, that determines creative new directions. The leaders of the legitimate system are simply participants in the functioning of the shadow system, albeit rather important and influential participants. They are particularly important and influential because other members of the shadow system project leadership and authority into them. The manner in which they cope with these projections has a profound effect upon the degree to which the anxieties of creative learning are contained rather than avoided.
Leaders, then, do not determine direction when they take up their roles in the shadow system. Instead, they become important participants whose primary function has to do with the containment of anxiety. They need to be involved in the group processes of the shadow system, but from a position on the boundary, where they can understand the processes but not get sucked into them (Miller and Rice, 1967). As people operate in the shadow system, leadership roles emerge spontaneously from the interaction. When the creative space is participatively occupied, then, the role of leadership shifts around the system according to which people have contribution to make and how effectively they are able to attract the attention of others to that contribution.
The other possibility is that a leader will emerge who occupies the creative space on behalf of others. Such a leader contains the anxiety of the others and articulates and initiates potentially creative thought, discoveries, and behaviors. Such a leader still cannot be in control of the outcome of this creative behavior; the dynamics of the system make this impossible. However, as part of the anxiety-containing function, the leader may subsequently present what has happened as intentional and this myth will be gratefully accepted by the followers. When they are caught up in this kind of dependent specialization in occupying the space for creativity, followers will be furious with, and dismissive of, leaders who admit that they do not know what the future direction is, and the leaders will be quite fearful of making any such admission. Together they will collude to make meaningful empowerment impossible.
Leadership from an extraordinary management perspective is thus very different from leadership from an ordinary management perspective. The leadership required in extraordinary management includes the capacity to contain anxiety for others, on the one hand, and the ability to provoke and contribute to the double-loop learning process on the other. Anxiety-containing capacity is a function of the behavior of the leader that has to do with the manner in which power is used and with compassion for the feelings and fears of others in a group. Leaders contain anxiety when they are able to empathize with others and articulate or interpret what they are experiencing (Carr and Shapiro, 1995). Provoking double-loop learning requires the capacity to play with metaphor and images and pose stretching challenges for others and the ability to listen and hold oneself open to changing one's mind.
A complexity theory of organizational development therefore ascribes very important and very difficult roles to management in addition to the currently dominant notions that also continue to be important from an ordinary management perspective. Complexity theories of management lead to a very rich, paradoxical theory of leadership in which leaders have to be both the conventional directors of others and the far more subtle containers of their anxiety and provokers of their double-loop learning capacity. These different attributes of leadership do not blend harmoniously with each other. Instead, they conflict with each other; directing and intentionally not directing are diametrically opposed ways of behaving and both are required of an effective leader in a complex adaptive system.
Self-Reflecting, Changing Mindsets, Designing the Use of Power, and Managing Anxiety
We come now to the requests managers make for the prescriptions for success that a complexity theory of organizational life leads to and the proof that it does indeed lead to success. The way in which this request is normally phrased indicates that it is posed from the stable equilibrium perspective that so dominates current thinking. The complex adaptive system theory outlined in this book is seen as a "new way of managing," and people want to know how they should go about implementing this new way, just as previously they were advised how to implement, say, Total Quality Management or Business Process Reengineering.
The first point to make is that I am not talking about a new way of managing. All the building blocks used to construct a view of organizational life based on complex adaptive systems are already reasonably well known and are also used in practice. In that sense nothing is new here. I am not, therefore, suggesting a new set of prescriptions, a new system along the lines of Business Process Reengineering. What I am saying is that a new overall framework is now available with which to think about and try to make sense of what people in organizations are already doing.
Thus, extraordinary management is not a new process; it is what people in organizations are already doing in an automatic way without much awareness or reflection. What is new is a coherent overall framework for people to use in reflecting upon what they are doing so that they may make more useful sense of it. I have I argued that on a very widespread scale, people say that they are using one kind of management approach as a cover behind which they are actually doing something else. Rarely do any of us reflect in a consistent and public way on what that something else is.
The prescription, then, is to use the insights of complexity science as a framework for individual, group, and organization wide self-reflection, that is, an examination of, and dialogue about, the processes of extraordinary management. What is the point of this? It is that self-reflection is the way to improve the kinds of messy processes required for extraordinary management rather than continuing to believe in the fiction of stable equilibrium. The prescription is for people to think about complexity and what it means and to develop their own responses to it. People who begin to think differently will almost certainly begin to act differently, and they will then almost certainly affect someone else who will begin to behave differently.
I am firmly convinced that attempts to provide rather precise sets of general prescriptions are simply an invitation for people to stop thinking. Sometimes this might be a good thing to do in the interest of action. However, when it comes to creativity in organizations, it is more important to present ideas and then leave it to the actors to make what they will of them in their own situations. Let me quote from the recollection of one articulate manager after he had been exposed to the kind of thinking that the science of complexity suggests:
As I have read about complexity theory I feel like I have become aware of a new world around me. I discovered ethnography and more importantly action learning and reflection-in-action techniques. ... Almost straight away my behavior changed. I began openly asking people for their thoughts and assumptions behind statements when they came up in discussion and when I disagreed with them. Almost immediately I felt much more satisfied with my inputs and with the responses was getting from people, particularly with the factory manager l am working with. ... We have a time set aside to reflect now -- it kind of happens though not in a planned way. This line of thinking opened the box though. ... Time after time we are now exploring the motivation and potential assumptions behind behaviors we see and attempting to modify ours to continue to develop support and momentum for change. ... We have also begun to identify patterns of behavior among our work force... Something we have identified is that there seems to be an underlying fear of failure and an expectation of coming second.
What this manager is describing is a process of individual self reflection being amplified to his closest colleagues. This process is changing mindsets and getting people to examine how they are using their power and what impact it is having on others, particularly what it has to do with the anxieties and fears of others and how they might be addressed. This call for self-reflection and the consideration of how anxiety is to be managed is perhaps the most important prescription of a complex adaptive system view of organizational life. Strategic plans cannot be about creative outcomes, but they could be about installing appropriate psychological and emotional conditions to encourage the spontaneous self-organization that might produce creative outcomes.
Redundancy, Slack Resources, and Play
Extraordinary management is a process that is practiced in a spontaneously self-organizing shadow system that is strongly akin to play. It is true dialogue in which people engage with each other, not to be in control, but to provoke and be provoked, to learn and contribute to the learning of others, to change their own minds as well as the minds of others. This process is like play in that it invites operation in the transitional zone of the mind, where reality and fantasy come together in the form of metaphors, analogies, and images. Experience-based intuition, rather than sequential, logical analysis, leads to the creative insight.
Play, however, takes up time, and much of it has no visibly beneficial outcome. A great deal of it will turn out to be futile, but some probably will not. Without it, however, an organization can not innovate. Creativity and innovation, therefore, require slack in an organization's resource utilization. Creativity and efficiency are enemies: the first requires slack resources and the second requires that there be none. They are enemies in another way, too. Efficiency requires that there be no redundancy, no repetition of the same tasks in different parts of an organization or at different times. Creativity, however, requires redundancy. First, it allows for the repetition of different ideas and experiments in slightly different ways, and small differences can have large outcomes. It also means that the organization will be more resilient in the face of inevitable failures in the innovative experiments going on.
There Are No Guarantees
Many managers, on first being exposed to the kind of theory outlined in this book, ask for evidence of organizations that have used it to lead to success. Using it means engaging in self-reflection; which sometimes leads to successful changes in behavior and sometimes does not -- the space for creativity has no guarantees. Furthermore, a period of success in any organization's life will have so many and such complex causes that trying to identify thinking along any particular lines as the cause of success would be to fall into the trap that the dominant paradigm is currently caught in. The fact is that no one knows what causes success until he or she has tried it and seen if it does. A complex adaptive system approach to understanding organizations offers, not a guarantee of success there is no such thing -- but more useful framework for making sense of experience, reflecting, and thus potentially designing more effective actions.
Keep It Boiling
Finally, a complex adaptive system approach to understanding organizational life contains a warning against complacency. The forces that operate to lock an organization into a successful strategy, to suck it into the stable zone, seem to be extremely powerful. The antidote is continually to seek to keep the shadow system on the boil, to keep coming up with novel ways of doing this and then containing the anxiety that is raised.
Conclusion
This book has attempted to draw the attention of researchers, consultants, and managers, those concerned with life in organizations, to new efforts being undertaken to understand life in nature. I suggest that these new efforts, making up the science of complexity, provide an overall framework for pulling together many existing building blocks in the literature on management and organization into a new way of approaching organizational life. I believe that this new way, which is built firmly on a psychodynamic approach, provides a more useful way of making sense of life in organizations than the stable equilibrium paradigm that currently dominates attempts to understand the problems of managing in organizations. I am not suggesting that the science of complexity provides us with a new set of comprehensive prescriptions for managing and organizing. I am suggesting that it provides a framework for making more sense of what we have been doing all along. The novelty of what am proposing lies in the suggestion that we reflect in public on what we are doing, using the science of complexity to inform that self-reflection. Why should we be concerned about such self-reflection at this point in our history? The answer, I believe, is that the speed of change is faster than ever before and the level of complexity we all must deal with is greater than ever before. If we are to contain the anxiety of creative activity in the midst of such complexity, we must find a way of making sense of our experience of life in organizations that resonates more with that experience and the way we feel about it. I suggest that the science of complexity provides us with just such a framework.
Antichaos: This term is used interchangeably with self-organization.
Archetype: An archetype is a potential behavior that preexists experience and awaits specific experience to be actualized or real ized. Although the archetype exists in a recognizable general form, its specific actualization is always unique and depends upon the specific experience. An archetype is therefore a similar concept to an immanent, implicate, or enfolded order as used by Bohm (1980) and by Aristotle. It is also similar to Plato's concept of an ideal form.
Attractor: This is a state of behavior into which a system settles if left undisturbed. Equilibrium states are attractors of the single point and periodic oscillation, or cycle, types. Strange attractor is another term used for low-dimensional chaos or fractal behavior. Bounded instability: Bounded instability is used in this book as a general term to include all forms of behavior found in a phase transition between the stable and unstable zones of behavior for a system. It therefore includes low-dimensional chaos, but not high-dimensional chaos. It includes fractal behavior and behavior at the edge of chaos.
Chaos: Low-dimensional chaos has a precise mathematical mean ing: behavior that has global structure but is specifically unpredictable over the long term. Its presence is indicated by a positive Liapunov coefficient and it is generated by a few simple math ematical rules. High-dimensional chaos is generated by a set of many rules and it displays very little structure; it is close to randomness and thus closer in meaning to the normal usage of the term chaos.
Collapse of chaos: This term means the same thing as self-organization.
Complex adaptive system: A complex adaptive system consists of a number of agents interacting with each other according to schemas, that is, rules of behavior, that require them to inspect each other's behavior and adjust their own in the light of the behavior of others. In other words, complex adaptive systems learn and evolve, and they usually interact with other complex adaptive systems, They survive because they learn or evolve in an adaptive way: they compute information in order to extract regularities, building them into schemas that are continually changed in the light of experience.
Complexity: When information about behavior is so irregular that its description cannot be compressed, then the algorithmic information complexity is maximal. Information cannot be summarized, only reproduced in full; that is, only a short computer program (schema) is required to describe the extremely few regularities that do exist. Effective complexity is defined as the length of the schema, so it is low when the environment is random, although the algorithmic information complexity is very high. For example, the schema might be a mean and a standard deviation that would allow the boundaries around the environment's behavior to be established, but nothing more. This obviously has a very limited usefulness in terms of guiding behavior. A system trying to adapt or learn in these circumstances will not be able to extract much in the way of regularities, nor will it be able to predict much in specific terms. Randomness equals high information complexity but low computational or effective complexity. Furthermore, when information is highly orderly, it is easy to compress, or summarize it. A short computer program, or schema, will be able to reproduce such information. Here, the algorithmic information content, or algorithmic complexity, is very low and the effective complexity, the length of the schema that captures the regularity, is also very low. In this situation, nothing much happens, nothing much changes, and there is no need for much learning or adapting.
A complex adaptive system will therefore only function in the sense of adapting when effective complexity is sizable, that is, in conditions that are intermediate between order and disorder. Effective complexity is defined in terms of the length of the adapting system's schema and, as such, it is an internal property of the system just as much as it is a feature of the environment. How ever, it has to be supplemented by the notion of potential complexity. This is the potential that complex adaptive systems have for creating a great deal of new effective complexity from only a modest change in their schema. An example of this is provided by humans, whose genome (biological schema) varies only slightly from that of apes, but who have much greater effective complexity in terms of their behavioral schema. Potential and effective complexity together amount to a form of bounded instability. Systems in this state operate in an intermediate phase between stability and instability.
Control: This book distinguishes between behavior that is controlled and people who are "in control." For people to be in control they must be able to specify desired outcomes and identify actions that are likely to produce those outcomes, and then be able to employ negative feedback to keep actual outcomes close to desired ones. People can therefore only be in control in rather limited circumstances. However, even when no one can be in control, the behavior of groups of people can display the characteristics of control -- coherent pattern, connection, and constraint -- through the process of spontaneous self-organization.
Creativity: Creativity is defined in this book as some alteration in the recessive schema of an individual, a group, or an organization that leads to a change in the dominant schema that then turns out to improve fitness. A change can only be judged to be creative, therefore, after the event.
Dissipative structures: Dissipative structures have stable, recognizable forms that are continually being dissipated and renewed, as when the cells in a human body are replaced.
Dominant schema: The dominant schema of an agent or a system is the set of rules or symbol system that models an agent's or system's perception of the current primary task and thus drives the performance of the currently perceived primary task.
Double-loop learning: Double-loop learning occurs when a system adapts its behavior to the stimuli presented to it in a beneficial way as a result of changing its schema. This is sometimes called complex or deutero learning and is to be contrasted with single-loop learning. Double-loop learning is therefore the change of a dominant schema and this requires a change in recessive schemas. Double-loop learning results in innovation and creativity.
Edge of chaos: This is a form of bounded instability found in the phase transition between the order and disorder zones of operation for a complex adaptive system.
Emergence: Emergence is the production of global patterns of behavior by agents in a complex system interacting according to their own local rules of behavior, without intending the global patterns of behavior that come about. In emergence, global patterns cannot be predicted from the local rules of behavior that produce them. To put it in another way, global patterns cannot be reduced to individual behavior.
Feedback: This refers to the process in which information about the outcomes of an action is fed back into the decision-making, or regulation, process to affect the next action. Feedback is negative when the information about a gap between expectation and outcome is fed back to dampen deviations from the expectation. Positive feedback does the opposite, feeding back information to amplify the gap between expectation and outcome.
Fitness landscape: This is a conceptual frame for thinking about the evolutionary journey of a system. Strategies that make the system fitter for survival represent movement up a hill, whereas disadvantageous strategies represent movement down into a valley. Each system's landscape is determined by the strategies of the other systems it interacts with. Evolution is therefore a journey across a heaving landscape. Smooth landscapes represent the ordered zone of operation and very rugged landscapes represent the disordered zone of operation. Landscapes that are rugged but not too rugged are optimal for evolution and constitute the edge of chaos.
Fractal: This refers to behavior that "fractures" so as to produce self-similar copies of itself. It is found in the phase transition be tween the stable and unstable zones of operation of a system and is a form of bounded instability. It is very close to low-dimensional chaos.
Implicate, immanent order: This is a pattern of behavior that exists as a potential because of the properties of some set of rules. Such order is enfolded and is then unfolded by the running of the rules, the experience of the system. See also Archetype.
Innovation: Innovation can be potential or actual. Potential innovation occurs when an agent or system alters its dominant schema. This is the same as saying that it alters its perception or model of current primary tasks or their manner of performance. Actual innovation occurs if this alteration is beneficial to the agent or system in the sense of increasing its fitness, and that happens when the change in behavior delivers what those agents and systems being interacted with demand or accept as the price for further interaction.
Legitimate system: This refers to the hierarchy, bureaucracy, and shared ideology that members of an organization recognize as having the authority to sanction actions and allocate resources.
Nonlinearity: A system is nonlinear when actions can have more than one outcome and when actions generate nonproportional outcomes, in other words, when the system is more than the sum of its parts.
Power law: This refers to a typical pattern of distribution of many small events and few large events that is typically found at the edge of chaos.
Primary task: The primary tasks of an agent or a system are the tasks that must be performed if the agent or system is to survive. Primary tasks produce what the other agents and systems being interacted with demand as the price for further interaction. The test of whether or not an activity is actually a primary task is whether it produces what those being interacted with demand in return for continued interaction. Performance of primary tasks is driven by the dominant part of an agent's or system's schema and that dominant schema models what the primary tasks are as far as the performing agent or system itself is concerned. This may or may not coincide with the actual primary task; survival follows only when the dominant schema models the actual primary task as determined by other agents and systems.
Recessive schema: This refers to the part of a system's symbol system that is not being utilized to form the rules driving the system's performance of the current primary task. It is therefore the symbol system that can be employed in play.
Schema: A schema models regularities in the stimuli experienced by a system. A schema consists of a set of rules that reflects regularities in experience and enables a system to determine the nature of further experience and make sense of it. A schema also contains rules indicating how the system should respond to its experience, which may include extending, modifying, or changing the rules comprising the schema. The rules in a schema are coded in the form of symbols, such as changes in electrical currents, chemical interactions, mental images, and numbers, that stand for some aspect of real experience. A schema is thus a symbol system.
Self-organization: This is the process by which agents in a system interact with each other according to their own local rules of behavior without any overall blueprint telling them what they are to accomplish or how they are to do it. The concept includes but does not coincide with double-loop learning, because deterministic systems, which do not learn, also display spontaneous self organization.
Shadow system: This is the set of interactions among members of legitimate organizational system that fall outside that legitimate system. It comprises all social and political interactions that are outside the rules strictly prescribed by the legitimate system. It is the arena in which members of an organization pursue their own gain, but also the arena in which they play, create, and prepare innovations.
Single-loop learning: Single-loop (sometimes called simple) learning, or conditioning, occurs when a system employs its schema without change, adapting its behavior to the stimuli being presented to it so that its behavior becomes more beneficial to it. This is to be contrasted with double-loop learning.
