Journal of Knowledge Management Practice, Vol. 8, No. 1, March 2007

Societal Learning And Knowledge Management: Diagnosing And Enhancing Their Potential

Gabriel A. Ramírez, Central University of Chile


This paper argues that societies and economies learn and that existing theories on organizational learning and knowledge management provide valuable insight on how that learning takes place and can be improved. Making explicit what learning and innovation have in common, it states that National Innovation Systems are learning networks that facilitate innovation throughout the whole society, and that public policy making and promotion of science, technology and innovation, by government and private agencies, are knowledge management tasks.

On the idea that the main causes for societies failing to learn lie in fragmentations in their learning loops, it proposes a methodological outline to enhance learning capacity in groups and societies. The proposal is based on two complementary theoretical bodies: a form of the organizational learning process that relates sensorial activities to reflection and action through an evolving model of causality; and a general model of complex organizations that is recursive and not hierarchical.

Key words: Organizational learning, knowledge management, National Innovation System, dynamic modeling, learning networks, complex adaptive systems.

1.         Knowledge Creation And Innovation Systems

Since Cohen and Levinthal (1990) described the process of creating and using new knowledge by companies as learning, a number of specialists in industrial economics have used that idea to deal with innovation in productive systems (Nooteboom, 2001; Dal Zotto, 2003; Waalkens et al, 2004; Leahy and Peter, 2004). Cohen and Levinthal also introduced the expression “absorption capacity” for the learning capacity of companies; and concluded that the amount invested in R&D was an affective index to measure the innovation capacity in companies and economies.

Bart Nooteboom (2001) considers innovation and National Innovation Systems as phenomena analogous to organizational and societal learning; he states that learning occurs according to a heuristics of discovery that is equally applicable to individuals, organizations and innovation systems. Nooteboom contends that such a heuristics explains the relationships and differences between incremental and radical innovation; single and double loop learning; and the origin of exploitation and exploration of knowledge. Accordingly, innovation is understood as a process that is analogous to learning. On the grounds of that analogy, it is possible to use the principles and theories of individual and social learning to examine and diagnose the innovation phenomena in organizations, institutions and societies.

Nooteboom acknowledges two main stages in every learning process: the first consisting of a localized innovation cycle, and the second as a cycle of diffusion where knowledge is shared through the productive system. Nooteboom views are complementary to those of Senge and Kim in that both share the view of constructivism and the theory of first and second loop learning as used by Bateson (1978) and later on by Argyris (1977).

Characterizing the advanced postindustrial society as the Knowledge Society, Bueno (2004) states that an essential trait of the new advanced society is possession of a system that permits an ever faster rate of knowledge creation and an ever shorter transference time from basic knowledge to applied science and technology to business. He calls that system the Knowledge System. Within the knowledge system Bueno distinguishes the following four interacting sub-systems:

Ø      A Science System, where scientific research takes place, consisting of universities and other research  instutions.

Ø      A Technology System where the results of scientific research are transformed into technologies through institutions such as science parks and innovation centres.

Ø      A Productive System that comprises the firms that integrate those technologies into their production processes.

Ø      A public institutional system that, including non governmental organizations, promotes science, technology and innovation.

In essence, this concept of Knowledge System is equivalent to the concept of National Innovation System as defined by Nelson and Rosenberg (1993) and that according to Agosin and Saavedra-Rivano (1998; pp. 26) corresponds to “. . . Research and development laboratories that belong to innovative enterprises, the government owned basic and applied research organizations, the private and state universities, the government (local and national) agencies that promote science and technology, and private fundations that support scientific activities”.

Dealing with problems of innovation in the economy, Saviotti (2000) states that the innovation systems in companies are capable of learning and that a National Innovation System can be conceived as a large network of interacting institutions that create and apply knowledge. Saviotti takes a systemic - cybernetics stand, stating that companies display adaptive behavior, and that they change and evolve as a result of their interaction with an environment that is heavily populated by competitors. Pointing out the evolving character of the economy he states that: “In economic systems R&D, or more in general, search activities contribute to variation, while regulation and competition are the main forces responsible for selection” (Saviotti, 2000; pp. 8).

To explain the function that economic systems play and the dynamics of their complexity, Saviotti resorts to the Requisite Variety Law (Ashby, 1976) and uses variety as the key variable. He defines variety as the number of actors, activities and objects necessary to describe the system, and concludes that, in an international context that increases its variety, for a country to be able to keep or increase its share of world income, that country must also continuously increase its variety.

Saviotti adds that innovation in productive systems is never the result of isolated companies or research institutions, but the result of a complex network of organizations whose interaction produces an innovation system. An innovation system behaves as a complex learning network. Therefore theories and principles that explain the behavior of learning networks can be used to explain, diagnose and forecast the behavior of those complex social constructs that produce innovation.

2.         Learning And Societal Problems

Insufficient personal income and slow economic growth are at the root of the social problems in less developed countries; traditionally, every effort to increase their development has been dependent on economic growth. Until the work of Paul Romer (Carrillo, 2004) most of economists believed that the sole factors that determine the long range growth potential of nations were the availability of capital and labor. Romer’s work showed that economic growth was also associated with the availability of knowledge, which is considered an emerging property of labor, and with the use of scientific knowledge in production processes. By the same token, scarcity of knowledge is regarded as the main obstacle to social development, welfare and quality of life. In the last ten years, governments and international development agencies, such as the World Bank, have consistently aimed to increase human capital and promote applied scientific research to production system, within a framework of free market and higher competitiveness.

Discussing the origin of societal problems in a postindustrial society, Senge and Kim (2001) state that the big problems that the USA faces today, such as the crisis of the health system, the loss of credibility of Universities, the increasing dysfunction of government, taken together configure what they call a generalized institutional failure and are due to the fact that the knowledge-creating system, the method by which society’s institution improve and revitalize themselves, is deeply fragmented.

Beer (1984) in a classic article on the origin of the Viable System Model, identifies the source of the organizational pathologies as being a chronic incapacity to articulate a shared vision of the future in modern organizations and in the inadequate working of the assemblage of what he calls Intelligence, Control and Policy systems. The pathologies express themselves as fragmentation of the intelligence functions that, in companies, are exemplified by the typically isolated working of managerial units such as Strategic Planning, Marketing and Research and Development, and their lack of contact with daily operational activities.

The effects of that fragmentation are managers that, having to face an unpredicted present, tend to over emphasise the contingency and short term operations, putting excessive control on people, reducing organizational autonomy and the possibility to display the needed variety to deal with the increasing variety of the environment. Beer (1975) places the origin of the traumas that organizational changes causes to society in what can also be called “fragmentation of the learning cycles”. The Viable System Model fully complies with the characteristics of complex adaptive systems (McElroy, 2003; Seel, 1999) and provides a comprehensive framework to deal with the effects of connectivity, diversity and information flows in human organizations.

3.         Group Learning And Modeling

The view of learning organizations developed by Senge (1992) and Kim (1993) is essentially constructivist in the sense that it focuses on the manner by which the subject of learning builds ideas and concepts and acquires new behaviors. It starts by confronting experience and current knowledge, and pays special attention to the processes of reflection and to the understanding of how ideas, concepts, principles, rules and theories are internalized into useful models for effective action. This approach suggests that model building is a process of successive approximation to better knowledge. Models are constructed, de-constructed and reconstructed by trial and error, experimentation, simulation and theory building. New knowledge stems from comparing models with reality through experience, discarding those that fail to explain and sticking to those that provide a satisfactory explanation for action. In essence, the constructivist approach understands learning as a process of building and improving conceptual models of reality that are useful to face problems that emerge in the environment.

Senge (1992) also took Forrester’s developments on managerial systems dynamics and combined them with the contributions of Argyris, Schön and Kurt Lewin to group learning. To learn about the real world it is necessary to build causality models that can only emerge from the process of continuously verifying their validity through experience, where the effectiveness of the learning subject depends on its capacity to build models that are effective to deal with the emerging complexity of the environment.

Senge and Kim take from Argyris and Schön the ideas of learning type I and II and apply it to organizations and show how these two types of learning explain the drive for new knowledge, and also the natural reluctance to address new knowledge and change in organizations. They show that learning is a continuous cycle of activities involving observation, abstraction and conceptual understanding, and testing that understanding through action. The distinction between learning type I and II is important to understand ideas such as learning to learn, and the idea of exploitation and exploration of knowledge.

On those bases, Kim (1993) explains how individual learning is transformed into organizational learning and how shared visions emerge and provide collective sense to organizational knowledge. Focused on the knowledge that individuals acquire within the organization, and on the knowledge that the organizations acquire through them, Kim remarks that learning has to do with firstly the acquisition of Know Why, implying the ability to articulate a conceptual understanding of an experience, and secondly the acquisition of the skill to know how, implying the actual capacity to display the right behavior.

Daniel Kim states that personal knowledge is embodied in mental models of reality and that those models determine the way people understand reality and behave. The idea of mental models is analogous to that of paradigm, in the sense of Kuhn (1970), in as much as mental models contain explicit and implicit assumptions that determines the way as people look at reality, interpret facts, conceptualize ideas and act. By the same token the idea that models evolve describing loops, where stable and unstable representation recurrently alternate, is analogous to the emergence of paradigms and the periods of crisis that precede changes of paradigms as discussed by Kuhn.

With the introduction of mental models, Kim is ready to clarify how learning type I is achieved without changing the prevailing mental models, whereas learning type II involves changes in the mental model.

To go from individual learning to organizational knowledge acquisition, Kim establishes the following premises:

1.      Organizations learn and know through their individual members.

2.      Organizational knowledge and learning is not the mere aggregation of individuals’ knowledge and learning.

3.      Group learning requires collectively shared models.

The link between individual and group learning is provided by the social interaction that generates the collective vision that the group or organization needs to achieve a collective purpose. Learning type I and II are also present in the group learning process. When talking of group learning, Kim uses the word “Welstanschauung” to refer to the collective assumptions that prevail in the organization and determine its culture, and the expression “Organizational Routines” to refer to the procedures, methods and practices that are consistent with the prevailing assumptions.

4.         Learning: Causality Models And Indices Of Performance

Figure 1 describes the organizational learning process, linking the sensorial aspect of information systems to reflection and action and their relations to self-regulation, self-organization and evolution. It shows how the learning process is triggered by unsuccessful attempts to respond to a changing environment and also by the need to anticipate to emerging change (Ramírez, 1981). In a very concrete sense, that description anticipated the work of Senge (1992) in using dynamics models to understand organizational learning. In economics, the advantages of using systems dynamics model instead of timeless models of optimal equilibrium resides in their ability to project into the future and to take into account the effect of time lags in the behavior of the modeled system. This has been discussed by economists such as Ormerod (1997) and Radziki (2003).

The idea of learning, implied in Figure 1, is constructivist in the sense that it states that the dynamic model that explains the interactions between the organization and its environment evolves into more complete versions according to its permanent confrontation with reality. It also suggests that monitoring the behavior of a set of relevant variables and parameters of a prevailing model correspond to actions for learning type I, whereas monitoring the relevance of the model itself corresponds to actions for learning type II.

Starting from the definition of the essential variables that in a company usually are the rates of profitability and levels of liquidity, Figure1 shows that to keep control on the behavior of a complex system, it is necessary to build a model of the causal relations that determine the behavior of the essential variables that defines their viability. In an economy the essential variables usually are economic growth, employment, inflation, public debt and trade balance.

























Figure 1 Organizational Learning, Models And Adaptation

5.         Learning, Information Systems And Knowledge Development

Figure 1 shows the place where model building fits within the process of learning and staying viable in the long term. Model construction, de-construction and reconstruction, along with change detection and re-organization, explains organizational adaptation, learning and evolution.  The figure also implies that the relationships between the organization and its environment, as time unfolds, become more and more complex. According to Ashby’s well known Requisite Variety Law it follows that to cope with the increasing complexity of the environment, the organization must generate models that are more complete and accurate and structures that are more flexible and rapid to adapt.

The figure shows that the model built around the set of essential variables, describes the causal relationships that exist between the variables and parameters that represent the organization and this variables and parameters that represent its environment. In companies, the essential variables are always related to liquidity and profitability, which, in generic terms, reflects the system capacity to maintain the required energy flows for the systems self-production. In the ambit of NIS the key variable are usually reduced those that in some way reflects the complex idea of innovation, such as the annual number of inventions in terms of registered patents, he annual number of scientific papers published in scientific journals (Lederman and Maloney, 2004) and the amount invested in R&D. To this purpose, the European Commission has defined a set of indices grouped into four categories containing 5 human resources indices, 4 about new knowledge, 3 about diffusion and application of knowledge and 7 indicators of investment in innovation, products and markets (The European Commission, 2003).

6.         Learning And Reflexive Observation

The model, to be coherent needs to be complete and to have closure.  Completeness in the sense that contains all the variables and parameters that are necessary to have a satisfactory causal explanations, and must have closure in the sense that exist, at least, one complete feedback loop. In Figure 1, the two feedback loops that act upon the modeling of the system –inside the box labeled ambit of reflection- are equivalent to the idea of obtaining a satisfactory explanation through successive approximations. Those loops, taken together with those that connect from the box “Ambit of Action” make of Figure 1 a constructivist description of the learning process.

Once the model is complete and coherent, it is necessary to determine which, among the whole set of variables and parameters of organization-environment’s dynamic model need permanent monitoring. That can be ascertain by sensitivity and robustness analysis, which is equivalent to find out what policies have the highest capacity to arrest the harmful effects of changes in the external conditions. This is equivalent to formulate hypotheses on the effects that decisions will have and therefore is equivalent to the process of defining strategies, courses of action and actions.

On the other hand, designing information systems is essentially about establishing routines that determine well-defined patterns of interaction between the components of an organization. It is specifically to do with establishing data capture procedures, creation of data bases, computer processing, data integration, filtering and access to information. In terms of learning, providing ITC infrastructure and developing information systems can be understood as processes that are analogous to wiring “perceptual neural circuits” that allow pattern recognition, evoke memories and trigger responses. These “circuits” determine regular patterns of behavior that can be considered organizational routines within a prevailing “Welstanschauung”, all of which are factors that provides stability and coherence to the organization, but also limits its perceptual and learning capacities.

In Figure 1 the operation of the information system –or the working of a well established neural circuit- is summarized in the activities of data gathering about the value that the essential variables take together with some other highly correlated non-essential variables, plus data on the behavior of the parameters to which those variables are sensitive (Beer, 1975).

7.         Learning: Sensing, Sense And Action

If the dynamic model, on whose assumptions the information system has been built, is a valid representation of reality, every detected change should be explainable by the causal relations that are present in the model; and every departure from the desired values of the variables under monitoring, could be arrested by applying a predetermined policy. If that policy is effective and can be implemented, the organization is applying learning type I, in the sense that the organization displays behaviors that are known to be effective.

Unfortunately, dynamics models as every other type of model are never complete, in the sense of being a definitive explanation of reality. Therefore, there is always the possibility that the variables under monitoring display changes that cannot be explained on the assumptions –or Welstanschauung- that prevailed when the circuits were established. In this situation, the model must be re-built to include new causality factors capable of explaining the observed odd behaviors. Building and re-building models on a different set of assumptions is, essentially, learning type II.

Finally, when there is an adaptive policy that is effective, in terms of the new prevailing model, but that cannot be implemented because some structural condition in the organization, the only option left is to redefine the organization in relation to its environment. In terms of managerial practice, that means carrying out actions such as diversifying, increasing the scale of operations,  de-investing, re-designing processes and/or products and developing strategic alliances. These options are, in the end, strategic changes; and this is how Figure 1 shows that strategic responses to environmental changes are trials, successful or not, in the generation of learning type II. Figure 1 reveals how self-regulation, self-organization, and organizational evolution relate to learning and innovation.

The building and re-building of the dynamic model, that is triggered by lack of coherence, incompleteness and structural incapacity to implement adaptive strategies,  correspond to the loop that constructivist descriptions of learning, establish between reflexive observation, concept formation and model building.

8.         Diagnosing And Enhancing Learning Capacity

Given that fragmentation of learning loops is at the root of the difficulties to learn that organizations and societies have (Kim and Senge, 2001), and that the viable system theory and the organizational learning approach explain the origin of such fragmentation; it is possible to resort to these theories to diagnose organizational learning capacity and find a way to enhance it.

For a complex entity such as a National Innovation Systems, improving its organization and functionality and designing and implementing effective policies can be considered knowledge management actions at a societal level. What follows is a first approximation to a methodological framework to facilitate organizational learning at any organizational level and a synthesis to look at knowledge management as the processes of enhancing the creation and application of new knowledge .

8.1.      Identify The Learning Unit Of Concern And Its Systemic Components

The starting point is to identify the entity, organization or human activity system whose learning and knowledge management capacity is to be diagnosed and enhanced. Typically, this will be a company, a business unit, an academic department, a research institution or any organizational instance that potentially has the fundamental systemic characteristics of a viable system; namely: the capability of independent existence, a recursive structure and the explicit or emerging presence of the five functions that are necessary and sufficient to ensure long term viability (Beer, 1981a). Having a recursive structure means containing viable systems and being contained by viable systems.

The five necessary and sufficient functions are: An operational one, performed by a set of highly autonomous operational units that carry out what the system is essentially suppose to do; a coordination mechanism to harmonize the working of those operational units; a control system to monitor and make sure that the units do what they have to do in the short term; an intelligence function to survey the long term behavior of the environment and to formulate strategies; and a policy function to balance short and long term priorities through resource allocation for action (Beer, 1981b).

This theory makes an important logical distinction between the components of a viable system. It establishes that the coordination, control, intelligence and policy functions are meta-systemic to the operational function in the sense that they take care of the successful survival of the whole system rather than to the success of the individual operational units. It also establishes that the operational function is performed by units that are also viable systems, hence also highly autonomous components.

For instance, if a National Innovation System is the unit whose learning capacity is to be enhanced, and given that its essential purpose is to facilitate innovation in the economy, one can visualize the knowledge producer institutions and productive organizations, such as research institutions, schools, universities, and companies as its operational components.

Figure 2 shows those components nested into three highly autonomous systems: A productive, an educational, and a science and technology system. The figure also displays the meta-system that supports the success of the whole and looks for synergic integration of the operational units. The icons of the triad policy–intelligence–control, in each component unit, represent the recursive nature of the NIS, implying that its three main components also meet the requirements of viability. One of the most important consequences of recursion is that the same principles and methodologies can be applied to any other level of recursion to facilitate learning and knowledge management.






















Figure 2 A National Innovation System: Components And Meta-System

The consistency of this nesting affects the effectiveness of the whole system: The more consistent is the nesting, the more resilient the system is. Figure 2 represents a very complex social system, where the integrity of its learning loops is in the essence of the way in which they are nested and they nest their components.

Although problems of effectiveness and management in complex social systems are often acknowledged, they are seldom considered learning problems. For instance, the National Council for Innovation and Competitiveness of Chile (2006; pp 56) stated in its final report: “The best international practices show that an efficient National Innovation System needs at the top, an institutional framework to drive, coordinate and provide it of direction. In that sense, foreign specialists had agreed that our National Innovation System is underdeveloped, ill coordinated, lacking guidelines and an integral and coherent approach”.

Figure 3 shows the NIS according to the customary description of the VSM, displaying two levels of recursion. The triad in its operational components means that the components are also viable systems that contain specific institutions such as schools, universities, companies and supporting organizations such as Regional Innovation Systems and some other local development agencies. 

Saviotti (2000) observed that the institutions that comprise a NIS, emerge during its evolutionary development. That has been the case in Chile (Ramírez, 2005) where that sort of emerging process expresses itself in the formal creation of some regional innovation systems and in the emergence, in some industries, of novel organizational arrangements such as clusters around mining, salmon, wine and forestry, and in the arising of a number of new supporting public and private agencies and high rank commissions such as the National Innovation Commission for Competitiveness.

Although the VSM is usually dismissed for being a too complex tool for analysis, the fact is that reality is still much more complex. In reality, viable systems are often part of more than just one recursive dimension. Business schools, for instance, are usually part of a viable system such as a university, but they are also part of regional or international associations of business schools that are also viable system. That means that any viable unit may be a component of more than one viable system which defines more than one recursive dimension (Leonard, 1999).






























Figure 3 The National Innovation System in terms of the VSM

In most countries, the Educational and the Productive Systems have a long and well known history of development and reforms, whereas Science and Technology Systems is a fairly recent development still far from maturity.

8.2.      Design A Meta-System

In the case of a well structured system such as universities and companies the elements of the meta-systems are usually embodied in organizational units that are clearly recognizable. In the case of universities, those units typically comprise the ones that do extension, promotion, planning and policy making. In companies those elements are usually embodied in departments such as R&D, market research, operational and strategic planning, and in a board of directors.

According to Beer (1981a), the existence of the meta-system components does not guarantee viability, since its poor integration is paramount in organizations of the postindustrial society which are an expression of fragmentation in the learning loops (Kim, 2001). The lack of integration of the meta-system’s component remains as one of the main challenges to innovation and societal learning in developing countries. In National Innovation Systems that integration turns out to be especially complex, since the components belong to more than one recursive dimension and, in many cases, the area of common interests is very thin, as it is the case of companies that are competitors within the same productive cluster.

Designing the meta-system is finding out, and if necessary creating the organizational units to perform the intelligence and control functions, to coordinate the working of the operational units and  to balance short and long term priorities. In the case of highly diffuse systems, such as National Innovation Systems (Devine, 2005), some of the component of the meta-system may be in the process of being created or even missing, their integration is usually poor and they seldom develop a coherent shared vision.

The creation of a properly working meta-system can not be done through an act of design by a omniscient individual or groups of individuals, this is a process than can only be facilitated by the combined effect of an open and transparent conversational process throughout the whole system tissue, and the presence of a mature leadership willing to enhance autonomy and capable of providing the necessary ethos for self-regulation.

The big challenge for people at the highest level of public policy making is to provide conditions for the emergence of those units and for the interactions that are necessary to arrive to widely shared visions on what the NIS is and should do.

8.3.      Coordination And Performance

The meta-system must also provide short range coordination to avoid clashing and oscillations between the components of the operational units. “As the operational units are semi-autonomous, the system needs ways of generating synergy between the units to reduce the effect of destructive self-interest” (Devine, 2005). In the case of a NIS the operational units are the educational, productive and science and technology systems made up by a large number of organizations that compete for resources, markets and recognition. One of the biggest challenges that open society faces is to achieve shared visions and coordinated behavior at every level of recursion.

Although the VSM does not prescribe the rules of coordination nor provides the foresight to get the right vision, it does establish the organizational conditions that will favor their rising and development. Its recursive nature provides a road map to deal with the complex task of orchestrating a shared vision at every level of recursion and to develop inter-recursive coordination means. The VSM theory states that autonomy is as an essential feature of viability, meaning that coordination in societal systems can only be attempted through recommendations and incentives such as subsidies and grants; by autonomously self generated rules and by voluntary collaboration.

It is also a task of the meta-system to establish the monitoring procedures to follow up the performance of the operational units and the behavior of the environment, in the short term. In the case of NIS that is to verify the contribution that firms, public services, NGOs and other organizations make to innovation. This is to define a set of essential variables whose behavior give account of the robustness and capacity to survive over the short and long term of the object of learning system (See Figure 1). According to the requisite of autonomy, this set of variables, plus the parameters that determine their behavior, should be permanently monitored at the operation level and, by exception, at the meta-systemic level.

The integration of recursively nested systems of indices was put forward by Beer (1975) and re-elaborated, around models of dynamic simulation, by Ramirez (1981). In the last few years, several simplified versions of those ideas have been developed under the generic name of Business Performance Measurements, where the best known general approach is the Balanced Scorecard of Kaplan and Norton (1996).

8.4.      Develop Models That Are Relevant To The Organization Viability

The task of visualizing the environment time ahead is a key activity that requires identifying the causal relationships that may exist between the essential variables and those variables and parameters that may affect their stability. Discovering causal relationships is in itself the essence of the reflection process of Figure 1.

Building a model is a key cognitive activity: It involves finding out the variables that should be monitored, it reveals the feed-back loops that generate self regulation as well as potentially explosive growth, showing the counterintuitive effects of information delays and emerging virtuous and vicious circles.

The ambit of reflection, in Figure 1, includes strategic thinking to act on the operational level and to adopt the organizational forms that the system should take to face unfolding environmental complexity. The results of that reflection are new initiatives that should be taken to ensure long term viability. According to the VSM theory a process like the one in Figure 1 should take place at every level of recursion.

8.5.      Policy Design And Strategic Action

As a result of the modeling process above, a strategic diagnostic will emerge providing the overall policies and corporate actions that should be taken.  This process is an essential activity of the intelligence function whose auscultation of the environment should reveal opportunities to catch and threads to avoid. Effective policy design must be anchored in solid internal reality; therefore it is part of that modeling process to realistically acknowledge internal weaknesses and strengths.

In Figure 1, those actions are also shown within the square “Ambit of Reflection” and labeled “robustness analysis and policy design”. In the case of a NIS, this is to define action programmes that provide a general framework of objectives to define more specific sub-programmes and projects to be implemented at lower levels of recursion. This belongs to the strategic action level that will make possible to display behaviors that are completely new. Policy making implies to carry out initiatives such as developing new products, renewing productive processes and developing and using new technologies.

8.6.      Determine Relevant Recursive Dimensions

Once the object of learning system has been properly recognized and diagnosed, it is necessary to identify the complementary recursive dimensions of the system in focus. For instance, if the system in focus is a business school, several complementary recursive dimensions can be identified such as its nesting within a national and international organization of business schools, and its nesting within a regional productive cluster and around one or several productive sectors. Although the complementary dimensions of recursion are subordinated to the immediate institutional dimension, they provide an important source of coordination and possibilities. In a highly competitive market, establishing the right strategic alliance may make the difference between survival and disappearance. The topic of multidimensional recursion  has been discussed by Leornard (1999).

The point of identifying the complementary recursive dimensions is to find out opportunities for collaborative action. In practice coordination, in multi-recursive dimensions, express themselves in voluntary and compulsory regulations such as code of conducts in guilds and trades, and legal frameworks for business. Social organizations and people are always members of two or more recursive dimensions and those memberships, to some extend, condition their behavior providing opportunities for both: variety reduction and amplification.

Considering more than one recursive dimension at a time makes the analysis and interventions more complex, but it also provides a more complete understanding of the real complexity of social systems. From the point of view of organizational learning and knowledge management, it is important to emphasize that the considering additional recursive dimensions provide opportunities to increase learning capacity and improve the stability and reliability of the viable systems in focus, and also the viability of the wider learning network that nests it. This is what happens with communities of practices, learning communities and the emergence of all sort of Web based knowledge communities, hence the importance of the IT global infrastructure and ease of access to people in developing countries at every level of the societal tissue.

9.         Applying The Methodological Outline

Although this outline has been described in mere 6 steps, it involves high complexity. The strategic analyses and modeling that are necessary to determine indices of viability, create collaborative relationships, establish inter-recursive dialogues and keep interactions balanced between semi-autonomous self organizing components are, in themselves, the object of management and its more recent approach: Knowledge Management.

Applying this outline does not mean discarding using, in a systemic way, whatever knowledge the behavioral, organizational and managerial sciences have to offer to deal with the specificities of any organization in particular. Organizational learning and knowledge management should not ignore the knowledge that managerial practice, academic research and the entrepreneurial culture have to offer.

This methodological outline can be modularly and incrementally applied to specific problem areas. For that purpose it is necessary to keep a systemic attitude looking to and finding out the “holons” or systemic units that display or should display the properties of viable systems. Within that framework, any technique or method from conventional management and practices can be used. It will be usually necessary to complement these interventions with the use of organizational behavior methods and techniques to improve group and interpersonal communication, creative problem solving, and the use of IT means to do distance work and keep collaborative work in remote learning networks.

This outline can be considered a means to improve the learning of social networks. The complementarity of the VSM with the constructivist approach to learning and its close relation to the modeling methods of Figure 1 shows it as a tool that goes far beyond being a mere organizational structure blueprint. The introduction of the concepts of recursion and recursive dimensions forms a bridge between a learning unit and a multidimensional environment, where that unit can find collaboration to learn and increase the viability of the whole societal network.

10.       References

Agosin, M.R., Saavedra-Rivano, N. (Eds) (1998), Sistemas Nacionales de Innovación ¿Qué puede América Latina Aprender de Japón?, Dolmen Ediciones, Caracas – Santiago.

Argyris, C. (1977), Organizational Learning and Management Information Systems, Accounting, Organizations and Society, Vol. 2; pp 113-123.

Ashby, W. R. (1976), An Introduction to Cybernetics, Methuen and Co. London

Bateson, G. (1978), Steps to an Ecology of Mind,, Granada Publishing Ltd., Paladin.

Beer, S. (1975), Platform for Change, John Wiley & Sons, New York

Beer, S. (1981a), The Heart of Enterprise, John Wiley & Sons, New York

Beer, S. (1981b), Brain of the Firm, 2nd. Edition, John Wiley & Sons, London.

Beer, S. (1984), The Viable System Model: Its provence, development, methodology and pathology. Journal of the Operational Research Society, Vol. 35; pp. 7 - 26.

Bueno, E. (2004), ¿Qué hace que la Sociedad Actual pueda ser llamada Sociedad del Conocimiento?, Sociedad & Conocimento Issue 2; pp. 51.

Carrillo, J. (2004), Desarrollo Basado en Conocimiento, Transferencia, Vol. 17, No. 65. Instituto Tecnológico de Estudios Superiores de Monterrey. México.

Cohen, W.M., Levinthal, D.A. (1990), Absorptive Capacity: A new perspective on learning and innovation, Administrative Science Quarterly, Vol. 35; pp. 138.

Dal Zotto, C. (2003), Absorptive Capacity and Knowledge Transfer between Venture Capital Firms and their Portfolio Companies, Proceedings of the The DRUID Conference, Copenhagen.

Devine, S. D. (2005), The Viable Systems Model applied to a National System of Innovation to inform policy development, Systemic Practice and Action Research; pp. 491-517.

Kaplan, R S., Norton, D P. (1996), Using the balanced scorecard as a strategic management system, Harvard Business Review, Jan – Feb; pp75-85.

Kim, D.H. (1993), The Link Between Individual and Organizational Learning, Sloan Management Review, Fall; pp. 37 – 50.

Kim, D.H. (2001), Organizing for Learning. Strategies for Knowledge creation and enduring change, Pegasus Communications, Inc., Waltham, MA.

Kuhn, T.S. (1970), The Structure of Scientific Revolutions, 2nd Edition, Chicago University Press.

Leahy, D., Peter, J. (2004), Absortive Capacity, R&D Spillovers and Public Policy, Neary University College Dublin and CEPR, available at

Lederman, D., Maloney, W. (2004), Innovación en Chile: ¿Dónde estamos?, Expansiva, Serie en foco ISSN 0717-9987, available at

Leonard, A. (1999), A Viable System Model: Consideration of Knowledge Management, Journal of Knowledge Management Practice, August, available at

McElroy, M. (2003), The New Knowledge Management. Complexity, learning and sustainable innovation, Butterworth Heinemann, USA.

National Council for Innovation and Competitiveness (2006), Final Report, March 2006, Santiago, Chile.

Nelson, R., Rosenberg, N. (1993), Technical innovation and National Systems in National Innovation Systems: A comparative Analysis, Oxford University Press, R. Nelson (Ed), Oxford.

Nooteboom, B. (2001), Learning and Innovation in Organizations and Economies, Oxford University Press, New York.

Ormerod, P. (1997), The Death of Economics, John Wiley and Sons, New York.

Radzicki, M.J. (2003), Mr. Hamilton, Mr. Forrester and a Foundation for Evolutionary Economics, Journal of Economics Issues, Vol. 37, No.1.

Ramírez, G.A. (1981), Management Information: Design of a System for Change Identification, PhD Thesis, Aston University, England.

Ramírez, G.A.. (2005), Societal Learning in the Knowledge Society. The case of the transition in Chile,. Research Project No 42. Faculty of Economics and Managerial Sciences. Universidad Central, Chile.

Saviotti, P.P. (2000), Networks, National Innovation Systems and Self-organization, 4th International Conference on Technology Policy and Innovation, Curitiba, Brasil; available at

Seel, R. (1999), Complexity and OD. An introduction, available at

Senge, P. (1992), La Quinta Disciplina, Granica Vergara, Buenos Aires.

Senge, P., Kim, D. (2001), From fragmentation to integration: Building learning Communities, in  Organizing for Learning. Strategies for Knowledge creation and enduring change, Pegasus Communications, Inc., Waltham, MA.

The European Commission (2003), European trend Chart on Innovation European Innovation Scoreboard, Technical Paper No 5, National Innovation System Indicators, October 31, 2003.

Waalkens, J., Jorna, R., Postma, T. (2004), Learning of SMEs in networks: the role of Absorptive Capacity, University of Groningen, Research Institute SOM (Systems, Organizations and Magement). Resaerch Report  04B14.

 Meet the Author:

Dr. Ramiez took his first degree at University of Chile and his MSc and Phd at Aston University, UK. He is an international consultant, professor in management and systems, and the former Director of the Graduated Programme in Business Administration at the Central University of Chile. He is Editor of the “Sociedad & Conocimiento” review and coordinator, in Chile, of the “Comunidad Iberoamericana de Sistemas del Conocimiento”.

Dr. Gabriel A. Ramírez, Managing Director AIM Ltd., Los Leones 643/201, Providencia, Santiago, Chile; Telephone: 56-2-233.10.58; Cel: 56-9-251.12.14; e-mail: