Introduction

Chile went into a highly significant political process in the early 1970s. It was an attempt to have a socialist revolution in freedom. The government of Presidente Salvador Allende, supported by political forces from the centre-left to the extreme left, was intent in transforming the country’s socio-economic relationships. It was a government for the workers and with the workers. The transformation had deep structural implications. From an economy that traditionally had supported the interests and consumption of the country’s privileged groups, the government wanted a people’s oriented economy. However, it was difficult to appreciate the complexity of this transformation.

Socialists’ experiences, such as the Soviet Union’s and Cuba’s, were driven by a centralized model of the economy and constructed on the shoulders of millions of people following the dictates of a planning system reflecting the views of a relatively small group of bureaucrats and experts. Enforcing centralized planning in Chile was not feasible; its long term democratic tradition had made that option difficult if not impossible. Tensions between a centralized planning and representative democracy were present throughout the three years of the Allende’s government. It is in this highly charged environment that the Cybersyn project emerged. The then General Technical Manager of the National Development Corporation (CORFO), Fernando Flores, saw both the conflicting nature of the government’s intended economic transformations and the opportunities offered by Stafford Beer’s organizational cybernetics. Beer was invited to Chile to discuss his insights about complexity and how they could be used in Chile’s political situation. This invitation was the origin of the Cybersyn project. With Beer in Chile the idea of a cybernetic alternative to centralized planning started to take shape. Presidente Allende gave the green light to proceed with this project, but, not so much as an alternative to the activities of the main departments of government responsible for the country’s economy but as a project to support CORFO’s management of the nationalized industry.

 

CORFO had been since the late 1930s responsible for a strategic State owned industry, controlling among others the oil, electricity, steel, and forestry industries. In 1970, the new government wanted to add to these industries medium sized enterprises; those producing electro-domestics, electronic products, furniture and so forth. These were strategic industries in the sense of offering a significant leverage to transform consumption patterns in the country. People at CORFO were grappling with this issue before Beer’s arrival to Chile. At his arrival, in November 1971, Beer proposed focusing on the cybernetics of the industrial economy, that is, on the communications and regulation of this exceedingly complex system. He had articulated this cybernetics in Decision and Control (Beer, 1966), the book that introduced several of us into his work in the late 1960s. This time, additionally, he had with him the manuscript of his yet unpublished book Brain of the Firm, the first of a trilogy about the Viable System Model (Beer, 1972, 1979, 1985); soon the project’s core group, about a dozen of us, was immersed in it.

During the two weeks of Beer’s first visit he managed to set up the Cybersyn project with a clear sense of urgency; supported by our briefings he produced several documents during those days. Perhaps a salient demonstration of his sense of purpose was encapsulated in a Plan of Action (Figure 1), in which he proposed precise tasks to braid organizational, informational and communication activities into a program for the management of the country’s industrial economy, which he named Cyberstride. In this Plan he also identifies specialized teams in Santiago (Team A) and in London (Team B). That Plan of Action was the beginning of the Project Cybersyn (Beer 1981, Schwenberg 1977, Espejo 2009, Medina, 2006 and 2011). The project lasted until September 11th 1973, when Salvador Allende’s government was overthrown by a military coup d’état.  Since then Cybersyn has received varied attention in the press (The Guardian 2003, The New York Times 2008) and other media, not always positive, among other reasons, because of its submission to the restricted knowledge generation capabilities of the computers of those days (Ulrich 1981), its unwarranted technical claims (Axelrod and Borenstein, 2010), and even for its Orwellian overtones (Baradit, 2010).

 

In this paper I offer personal reflections 40 years after the project’s demise in 1973. The emphasis of these reflections is contrasting its rather limited achievements at the time with its vision and relevance for our societies today. Its claims were large; it was presented as a project that made significant contributions to the management of the country’s economy in a short period of time. For its own sake I think necessary to compare its actuality with this claim. Particularly it is necessary to appreciate the project’s methodological and epistemological shortcomings to gain an understanding of its potentialities. My argument in this paper is as follows; first, I discuss Cybersyn’s actuality, what did we do in those two years? This is followed by a methodological revision of this work and reflections about the epistemological evolution of the Viable System Model in the last 40 years; the third section explores the project’s potentialities and vision; finally I discuss Cybersyn’s evolving meaning over the past 40 years.

Cybersyn’s Actuality

The Viable System Model (VSM) is one of Beer’s most important contributions to organizational cybernetics (Beer, 1972, 1981, 1979, 1985). The VSM emerged from Beer’s understanding that nature’s long-term evolution of viable systems had much to say about the viability of exceedingly complex systems such as firms (Beer, 1972, 1989). The focus of his attention was on the evolution of the human nervous system. From his understanding of this system he argued that any viable system had five systems, 1 to 5; System 1 was an operational system, the one producing the system’s products; System 2 was an anti-oscillatory function to coordinate the operational units of System 1; System 3 was distributing and optimizing the use of resources within the viable system; System 4 was responsible for the system’s adaptation to its environment and System 5 was responsible for policy-making (Beer 1972).  Additionally, a key aspect of this model was its structural recursion; each operational unit constituting System 1 had adaptation and production challenges just as the global system had, that is, each operational unit had S1, S2, S3, S4 and S5, and within each of these units the same structure for adaptation and production was responsible for the viability of their own units.

 

Within days of being briefed about CORFO and the nationalized industry, Beer mapped it into a recursive structure with CORFO as the system-in-focus embedding four ramas (groups of related industrial sectors), each of these ramas constituted by a set of so-called industrial committees or group of related industrial sectors, each of these committees embedding enterprises and enterprises embedding plants ¹. He then hypothesized that all these operations, starting from CORFO, had to have a viable organizational structure. In this sense the VSM was used as a heuristic rather than as a diagnostic or design tool.

‘Variety engineering’ is a key concept underpinning the VSM. For Cybersyn the design was reducing the large complexity of production activities at all structural levels, from plants to CORFO, and the disturbances buffeting them, to relevant information for management. The point was ignoring what deserved to be ignored and reporting significant changes. An aim for variety engineering in the project was offering a model driven approach to reducing situational complexity to a manageable level at the same time of improving performance.

The hypothesized recursive structure was used as a platform to design performance indices, based on the actualities (ACT), capabilities (CAP) and potentialities (POT) of essential variables for all the operational units, from the local to the global (Beer, 1981). The intention was measuring in real-time significant changes in the behavior of essential variables for workers and managers. Significant methodological and practical developments were made designing indices. Local people measured their daily actualities to compare them to their capabilities, or the best they could achieve with existing resources, and their potentialities, or the best they ought to achieve with investment to remove restrictions and bottlenecks. These indices were used to collect data in as near to real-time as practically possible and processed using a statistical formalism  2 ²; the Cyberstride suite was the software for this processing. The data collection was underpinned by a significant modeling capacity. Operational researchers produced quantified flowcharts for plants, enterprises and sectors to work out their capabilities and bottlenecks, and discuss with managers potentialities to design performance indices. Designing indices consumed the largest chuck of resources in Cybersyn. In practice the emphasis was designing production and human resources indices. Operational research modeling was used to design aggregated indices for enterprises and sectors. By the end of the project about 60% of the nationalized industrial economy was included in one way or another in this system.

Geographically industrial plants were distributed throughout a very long and thin country; capturing data required more than traditional mailing procedures. Real-time communications was the challenge; serendipitously we found a large number of spare telex machines in one of the state-owned enterprises and their installation followed in plants and enterprises throughout the country, as well as in industrial committees, CORFO and other government offices. A telex room with tens of machines was installed at CORFO; in practice it was an operations room for the state-owned industry that offered an incipient nervous system for the industrial economy; it was called Cybernet. Enterprises’ data were transmitted to the government’s computer center, where Cyberstride did the data processing. If significant changes were detected reports were sent back to the affected units. The expectation was that problems would be solved locally, however if problems persisted, after an agreed period with the affected managers, indices reports automatically jumped to the next level up under the assumption that these managers would have more chances to solve the related problems. This jumping up of ‘algedonic’ signals was intended for all structural levels (Beer, 1981).

Beer also wanted an economic modeling capacity in CORFO. Its purpose was having capacity to model the dynamic behavior of the industrial economy, particularly to balance the ear on the ground provided by Cyberstride’s indices with the eye on the future provided by these dynamic models. This part of the project received the name CHECO (CHilean ECOnomy); a small group of economists undertook this task with the support of a small team of system dynamic experts in London. Simple models of the Chilean economy were produced in collaboration by these two teams. This dynamic modeling of the Chilean economy used MIT’s Dynamo software (Forrester, 1971).

Finally, the display of indices’ reports, related information and dynamic models was focused on the design and construction of an Operations Room (Fig 2).  This room was envisaged by Beer as a ‘liberty machine’ (Beer, 1975, Medina 2011) a physical space to support policy-makers’ conversations. The emphasis was its ergonomics; a man-machine inter phase to improve decision-making. The room had several screens; one helped focusing the participants’ attention on the system-in-focus; the next two were used to project significant changes in performance indices relevant to the users of the room (e.g. energy performance indices if the managers were senior managers of the energy sector) and the algedonic signals of unsolved problems at lower levels of recursion.  The next set of four screens, called datafeed, gave information relevant to the actuality, capability and potentiality of relevant performance reports; information was projected by slide projectors controlled by managers from their chairs. The room also had two screens to project the outcomes of CHECO models and support discussions about the dynamics of the system-in-focus and its long term behavior. Decisions emerging from conversation in this room could be transmitted via telex (i.e. Cybernet) to the affected units and people. This room was a prototype that seldom was used by senior managers and politicians. However, the idea captured the imagination of people in the shopfloor and at least in a couple of plants the walls of workers’ meeting rooms were used to display performance indices and relevant information.

 

Of the four tools Cybernet was the one that changed our understanding of information and communications. At the beginning Cybernet was a tool for data transmission from plants, enterprises and committees to the telex room and from that room to the computer center where Cyberstride produced exception reports to send them back to the appropriate structural levels. These data flows were in themselves valuable contribution to management requirements, but CORFO managers soon learnt that Cybernet could be used for other purposes; rapidly documents, reports, and request of all kinds started to flow. These flows increased the use of Cybernet beyond the industrial economy; soon requests for machines came from ministries and other government institutions. What was unexpected was that when lorry drivers and small retailers went on a politically motivated strike in October 1972³ , Cybernet played a key role in defeating it. The network worked 24/7 and in practice became a powerful tool for horizontal coordination. Requirements and supplies were managed by enterprises among themselves reducing the need for hierarchical intervention. This was a most clear example of cybernetics in action. Not surprisingly this experience led many to relate Cybersyn to supplies and transportation activities beyond industrial production; an incipient value chain was in operation. The capabilities of Cybernet emerged from the turbulences of a country almost paralyzed by political fights. In politicians’ and managers’ minds Cybernet and the Operations Room became one; now the Presidential Palace wanted to have ‘this room’ in their own premises. The potentials of Cybernet, as an incipient ‘internet’, became apparent, but unfortunately it was too late to change the situation. At that time Beer became aware that Cybersyn had to be extended beyond industry and proposed a re-structuring of the project that for the first time had the chance to include the whole economy and not industrial production alone. He made a detailed proposal for that effect but it was already too late for Cybersyn to improve the cybernetics of the country’s economy. People in government were aware that a military intervention was in the offing and that there was no time for significant changes in relationships and management. Ironically, Cybersyn’s major success was the beginning of its demise.

One of Beer’s preoccupations as the project evolved was the inclusion of the people in policy-making processes. This was a concern for an inclusive democracy, giving policy-makers the chance to align their purposes with those of the people. This concern was articulated in an off-shoot of Cybersyn; the project Cyberfolk. Presidente Allende’s utterance “at last the people” as Beer explained to him the VSM was its origin. Cyberfolk was a technology aimed at including the people in policy processes; it was an attempt for a real-time response of the people to politicians as they discussed a policy in a public space (see figure 3). Underpinning Cyberfolk was Beer’s attempt of designing a homeostat to balance, in this case, the high complexity of the people  —their individual concerns — with the low complexity of a relatively small number of politicians, managers and experts dealing with these policy concerns. In terms of the VSM Cyberfolk wanted to give closure to policy-making (the VSM’s System 5). His paradigmatic contribution for this purpose was the “algedonic meter” (Figure 3), a device to measure people’s satisfaction or dissatisfaction with progress in public conversations (Beer, 1981); this was a tool to help them steer these conversations.

Critical Review

In what follows I offer a critical review of Cybersyn’s methodological implementation and discuss the epistemological lens used in it. The aim is preparing a platform to make visible its potentialities for a better society in the next section.

About Methodology

It can be argued that methodologically Cybersyn emphasized technology at the expense of a significant involvement of those transforming the Chilean economy on the ground.

A good cybernetics of the Chilean economy was, and indeed is, equivalent to its effective organization (Beer, 1975) 4.⁴  For a good economy it was, and is, necessary an effective organization and to achieve that the Viable System Model is a powerful tool; it offers an effective recursive structure for the implementation and adaptation of the government’s economic policies. Specifically, from plants to CORFO, it was necessary to produce relationships of autonomy and cohesion.

A first challenge for Cybersyn was setting a system for the effective management of the complexity of the industrial economy through well-articulated recursion levels. That was a tall order for Cybersyn and there was not much time to work out an effective unfolding of the industrial economy’s complexity. The recursive structure of ramas, sectorial committees, enterprises and plants was agreed on pragmatic political grounds and not on sound cybernetic principles for managing complexity. Cybersyn did not diagnose and design the industrial economy’s structure; instead it used hierarchies emerging from political processes. These were power hierarchies and relationships between levels driven by politics. Although the prevailing ideology was respecting the autonomy of enterprises, committees and so forth the organization structures of these operational units were not mature; the cybernetics of the nationalized industry was weak.

Structural recursion is by and large the outcome of self-organizing processes, which can be enabled by organizational design. The assumption was that ramas, industrial committees, enterprises and plants were all autonomous units embedded within autonomous units. This was a strong assumption that hid hierarchical relationships which denied structural recursion (Espejo, 2011). The pressure to make things happen in the economy and Cybersyn did not time to reflect too much upon relationships and it is clear that these were hierarchical rather than recursive.

Cyberstride’s focus was on designing performance indices for the essential variables of enterprises and plants. In practice these indices were designed to measure the performance of their internal operations, at the expense of measuring relations with economic agents in the environment. This focus reduced the chances of using some form of market relations to build up an effective economy. The focus was on existing production processes and not on the dynamic capabilities that were necessary for organizational adaptation to environmental changes (Teece, 2008). This was an important methodological issue not considered in Chile at the time but it implementation was latent in Cybersyn. Over the years this methodology has evolved in the context of multiple organizations, for instance at Hoechst AG in Germany (Schuhmann, 2004). At a more general level since several methodologies for indices design, such as Critical Success Factors (Rockard, 1979), Balanced Scorecard (Kaplan and Norton, 1996) and others, have been implemented in enterprises of all kinds.

Overall Cybersyn emphasized the filtration of operational complexity rather than the amplification of organizational complexity. Not much attention was given to enabling lateral coordination within and among plants and enterprises. Coordination was necessary to increase the chances of distributed ‘local’ problem solving. Coordination systems are huge amplifiers at the local level. Unless people in an organizational system, in this case units of the Chilean economy, share operational standards as well as values, mores and purposes they will find it very difficult to coordinate their actions by mutual adjustment. Lacking these coordination systems the natural orientation of communications was vertical, that is, hierarchical. Today, with our current understanding of the VSM, it is clear that System 2, or its coordination function (Espejo, 1989), is a powerful function to enable autonomy. With current technologies, such as internet, social networks, unit to unit software and many more, this systemic function plays a role that was not thinkable let alone possible in the early 1970s. There were no information and communications technology available in those days to implement real-time coordination and not surprisingly Cybersyn emphasized filtration, that is, indices of performance, rather than amplification, that is, coordination systems; however, the VSM helps us seeing that both aspects are necessary for effective performance. Enabling autonomy in organizations is an aspect that I have since emphasized in applications of the Viable System Model (Espejo, 2001, Reyes 2001, Espejo, Bula and Zarama, 2001, Espejo and Reyes, 2001).

The Chilean Economy Model (CHECO) was a relatively under resourced component of Cybersyn; the idea of producing a model of the economy without the participation of key actors, such as the Ministry of Economics, the Ministry of Finance, the Central Bank, the National Planning Office and so forth, condemned it to being no more than a learning exercise within the Cybersyn team. The system-in-focus for the CHECO modeling was the full economy. This was a pragmatic but inadequate choice; Cybersyn’s system-in-focus was not the total economy. Good cybernetics for this modeling would have implied distributing modeling and planning capacity throughout the recursive structure of the industrial economy as a contribution to assess the dynamic capabilities and potentialities of all autonomous units, from CORFO to the local units. In practice CHECO had neither influence on the management of the Chilean economy nor on the management of CORFO and its operational units. From a VSM’s perspective CHECO had to be a modeling tool to support ‘inside and now’ and the ‘outside and then’ debates within all autonomous units (i.e. S3, S4 and S5 in the VSM). It had to be a tool to relate the productivity of operational units such as CORFO, ramas, committees and enterprises to their longer term needs to contribute to their adaptation to turbulent environments. However, as said above, these units were not autonomous and the dynamic modeling of their environments was not one of CHECO’s concerns. In the end the planning of the industrial economy was a centralized function of CORFO driven by macroeconomic concerns, not distributed to the ramas, committees and enterprises. The relationships of CORFO’s Planning Department were with the National Planning Office rather than with the Cybersyn project.

As for the Operations Room (Figure 2) from a methodological perspective Cybersyn stressed constructing conversational technology rather than ‘designing conversations’; it offered a technology to include people in policy processes but did not offer a methodology for their meaningful inclusion. These have been controversial aspects of Cybersyn that over the years have been criticized as science fiction and fanciful technology (Axelrod and Borenstein, 2010) 5 .⁵ However, the vision of a conversational space for self-reference was, and is, a strong one. It has been replicated in multiple enterprises and other institutions (Holtham et. Al. 2003). Those who see it mainly as flashing technology quite naturally will find it wanting; those who see it as a conversational space for distributed policy-making throughout the organizational system will see operation rooms as a shorthand for policy-making requiring steering conversations between those running the operations ‘inside and now’ and those dealing with the ‘outside and then’. This is a place for balancing the stretching produced by the internal operations and the environmental demands. The vision was blurred by the excitement of constructing a Liberty Machine (Athanasiou 1980, Beer, 1975 and Medina, 2011). In practice this meant that Cybersyn did not pay adequate attention to enabling conversation for workers, managers and politicians to work out distributed but aligned purposes for the autonomous plants, enterprises, committees, ramas and CORFO. The overall global political strategy of developing a people’s oriented industrial economy, in the context of a highly uncertain political environment, restricted necessary conversations for distributed local self-reference and autonomy. CHECO and Cyberstride together could have supported these conversations; however, the Operations Room failed to engage the right stakeholders. It was an instance of a technology dominated agenda at the expense of enhancing the autonomy of enterprises and therefore contributing to the performance of the national economy. Beer’s later work in Team Syntegrity (Beer, 1994) was a powerful methodological contribution to designing conversations. Indeed, varied technologies are now available to enable these conversations; they were unthinkable in those days and give credibility to Beer’s vision 40 year ago.

Finally, Cyberfolk was yet another example of vision and technology anticipating future developments. Extending this idea of inclusion to on-going policy issues received limited methodological attention in those days and was restricted to a small group of scientists and experts. In the late 1990s and early 2000s, together with Clas-Otto Wene, I used the VSM and Habermas’s communicative competence to propose an approach for inclusion and transparency in nuclear waste management policy in Europe (Wene and Espejo 1999, Espejo 2003) and later on with German Bula for a discussion of inclusion in Colombia (Bula and Espejo, 2011); again Beer’s vision was ahead of its time.

Epistemological lenses

Over the past 40 years significant social, organizational, economic and technological developments have increasingly helped transforming Beer’s vision into reality; real-time management and coordination, autonomy within organizations, communication networks, conversational spaces, regulation of the economy and so forth. New information and communication technologies have gone hand in hand with an evolving epistemology to account for the complexity of interactions and communications in organizational systems. The VSM’s information management epistemology, which dominated our work in Chile, is now being replaced by an operational epistemology (Espejo and Reyes, 2011).

This new epistemological lens for the VSM started, for me, in the days of Cybersyn. Its stronger information management epistemology, what I will refer below as the black box description of organizations, was increasingly challenged by a communications epistemology or operational description that highlighted the need to account for the moment to moment complexity of organizations striving for their viability. Today, our understanding of the VSM is much more sophisticated than in those days; we understand much more the accounting for the complexity of organizational systems. It was serendipitous that Heinz von Foerster, Humberto Maturana and Francisco Varela were working in Santiago precisely as Cybersyn unfolded. Conversations with them during the early 1970s helped us start seeing the economy with the lenses of second order cybernetics — the cybernetics of the observer — (von Foester 1984), operational closure, structural determination and structural coupling (Maturana and Varela, 1989, Maturana 2002, Varela 1979). Their work has influenced the operational epistemology of today’s Viable System Model (Espejo and Reyes, 2011).

A clean epistemological accounting of the complexity of social systems (Varela, 1979) should account for both the complexity of their input/output transformations, as observed by external observers, and the complexity of the relationships between the observer participants producing these systems. These are two complementary perspectives, one is accounting for external observations of transformations of inputs into outputs and the other is accounting for observer participants’ recurrent interactions or structural couplings. The latter is a far larger complexity but both are necessary.

In the former description observers observe organizational systems as black-boxes; they are in a privileged position where they can observe both these systems and their environments simultaneously and establish correlations between the two through time. Their observations are associated with a mode of inference in which output information affecting the inputs of these systems determines, assuming a good model of their transformations and no unexpected disturbances, their future behaviour. It is a mode of inference that has associated with it a discourse about controlling a system’s behaviour by choosing the appropriate controllable inputs. In this type of description control is understood as restricting the system’s behaviour to reach desirable outcomes or goals (Rosenblueth et al., 1943). This type of description helps accounting for the complexity of organizational transformations (Espejo and Reyes 2011, Chapter 3). It recognises that often ‘it is not necessary to enter the black box to understand the nature of the function it performs’. This is Beer’s First Regulatory Aphorism (Beer, 1979, p. 59). This aphorism implies that the transformation of inputs into outputs is governed by regularities and that these regularities can be established through observation. Though this type of description is referred to as functionalist and often is dismissed as mechanistic, it is valuable to account for the complexity of the boundary interactions of an organization’s transformation (Espejo 1989). However for autonomous systems choosing variables to observe and control depend on the purposes ascribed to the organizational system and therefore, depend additionally on a wide range of inner conversations. Clarifying purposes permit us to work out the inputs and outputs relevant to the observers’ purposes in the situation. The boundaries of the system are thus defined by the variables the ‘inner’ observers choose to regulate. The design of indices of performance in Cybersyn was mainly influenced by the government’s policies and less by conversations about purposes and boundaries at the levels of CORFO, ramas, committees, enterprises and plants. In that sense the chosen essential variables to monitor in real time were less the outcome of autonomous conversations of operational units at different levels of recursion and more the outcome of global policies. No doubt conversations about purposes were happening in all those units, but the Cybersyn teams responsible for indices design, in general, were not involved in them. This was not a shortcoming of the VSM, but of Cybersyn’s implementation. Critics of Cybersyn (Ulrich 1994) failed to understand this distinction.

For operational descriptions of organizational systems there are no inputs or outputs (Varela, 1979, p.85); observers are accounting for their systemic experiences standing on its ‘inside’. The focus is on the relations that increase the chances for the system’s viability; its cohesion and adaptation to the environment’s perturbations, naturally, including other economic and social agents. In other words, because observers are not in privileged positions anymore (i.e. ‘outside the system’) there is neither an environment nor a set of inputs, outputs or a transformation process (i.e. a function relating the outputs with the inputs) to account for the system’s behaviour. All that observers have at hand are their interactions and communications constituting the system as a whole. These are operational descriptions. External perturbations may trigger changes in the internal structure of the system but they do not determine its future behaviour. This is why this mode of description is more appropriate with a discourse about autonomy and, therefore, for describing the behaviour of autonomous systems. This is an operationally closed system that makes its own decisions. In this type of description control is understood in terms of self-organization and self-regulation. Again, as explained before, Cybersyn designers did not focus enquires on these relations/conversations; they were not accounting for their complexity. This accounting would have been possible if these designers had had the chance to study and influence the cybernetics of the industrial economy at all its recursion levels; indeed a major endeavour.

Under the black box type of description it is commonly said that the organizational system operates with a representation of its environment. On the other hand, an operational description makes of the organizational system a cognitive non-trivial machine (von Foerster, 1984); this cognition arises as a result of the structural interconnectivity of operationally closed components. Under these circumstances, it is the structure of the organizational system that selects which patterns of disturbances in its environment are going to be ‘seen’, ‘heard’, or in general ‘perceived’. It is its internal structure that makes sense of the world ‘out there’. It is in this sense that we say that organizational systems are structure-determined (Maturana, 1988; 2002). In Cybersyn, CHECO models were (limited) representations of the Chilean economy and not an outcome of politicians and managers’ structural couplings producing, through these recurrent interactions, shared models of the economy for its regulation, adjusting their relationships, that is, their models as unexpected disturbances hit the economy.

The concept of information changes dramatically under these two types of system’s descriptions. Information as referential, instructional and representational is a concept that pertains to the black-box type of descriptions (Simon, 1981). On the other hand, in the operational type of descriptions, we use the word information to endow its environment with meaning (Varela, 1986, p. 119). With this distinction, we are moving from questions about semantic correspondence to questions about structural patterns.

For Cybersyn black boxes were the dominant mode of description; the Viable System Model was used as a set of black box descriptions of the industrial economy; they took the form of quantified flowcharts. Today, the VSM is a far more complex and sophisticated tool to study and design organizations; the operational type of descriptions and related forms of intervention were not used in Cybersyn, and we now are clear that this type of description is necessary to make sense of black-box descriptions; it is necessary to account for conversations to clarify purposes, to build up responsible trust between autonomous units, to coordinate their operations and so forth.

Yet, in spite of recent developments in complexity theory, the study of economic and social phenomena still depends to a large degree on black box descriptions; there is little about relationships and mutual regulation to enable effective ecologies of organizational systems. In the book the 23 things they don’t tell about capitalism, the Cambridge economist Ha-Joon Chang (2010) highlights that there is no such thing as a free market, that free-market policies rarely make poor countries rich, that contrary to the centre-right view that governments cannot pick up winners (in resources allocations) they often do pick up winners, that contrary to the view that self-regulating markets make the right choices, leaving these choices to the markets alone is not smart, that even if we believe that free-market economies are not regulated, in general they are over regulated, that contrary to the view that financial markets need to become more efficient they need to became less efficient, and so forth. The traditional views about economic policies are supported by black box descriptions of the economy; limited attention is paid enabling self-organization and supporting effective relationships. The VSM makes possible systemic views that counter fragmentation and support understanding economies as operational descriptions of relationships to make them more viable and just. Cybersyn’s vision, albeit not its actuality, offered this regulatory and structural framework. This is our concern next.

Cybersyn’s Potentiality: towards Beer’s vision

What did Cybersyn mean in the context of the Chilean government and in particular of its industrial economy? To what extent did it influence people’s decisions in the industrial economy?  Did it increase the reach of managers or enhanced the autonomy of the workers? Were its tools deskilling workers or increasing their problem solving capabilities? By highlighting Cybersyn’s actuality, in particular its limitations, this paper has offered a dispassionate view to reflect upon these questions.

Cybersyn was a platform for change that still is unraveling. Tough the project had technocratic overtones and limited political influence, its true value was its vision, which still has much to offer to improve society. In contrast to the traditional emphasis that economists put in mathematical modeling to support policy processes, Beer’s Viable System Model points towards enabling by design the self-organization of economic agents; Beer acknowledges both the huge complexity of any economy and the need to support by design the regulation and self-regulation of its agents.

 

The VSM was at the core of Cybersyn however the methodology for its application was blurred. We did not question whether CORFO, and its embedded autonomous units, had viable structures able to cope with the demands of the people and of other internal and external economic agents and whether by structural design and improved communications we could have stretched the structures and improved their performance. Though its conceptual framework was powerful and made valuable technical contributions, Cybersyn failed making the organization of the economy more effective; it neither produced inclusive policies for workers and the people in general nor did it improve the productivity and overall performance of the economy.

The challenge was making productive a state-owned industry; the transformation should have gone far beyond making available information and communication systems. Indeed, the methodological capabilities of the Chilean Team were incipient. Our emphasis was the implementation of information and communication technologies, and except for the October strike and Fernando Flore’s political influence, we failed achieving any significant economic and political influence. Cybersyn did not have much influence developing people’s potentials towards a more effective industrial economy, let alone towards an effective organization of society. The project had the Viable System Model as its foundation and indeed its key ideas, such as managing complexity, ultrastability, adaptation, recursion and others, were most important for Cybersyn designers but their embodiment in the practices of managers and workers were indeed limited. Success for these ideas required increasing the productivity, autonomy and entrepreneurship of economic agents within a regulatory framework that steered them towards aligning their actions with the government’s policies. The acid test should have been producing a cohesive economy with productive autonomous enterprises. These were potentialities of Cybersyn.

Chile’s economy was weak. It was dependent on minerals for its foreign currency and foreign technology and expertise for its industrial development. For consumables the policy had been protecting the local economy through high import barriers. Nationalizing a large number of enterprises in a short period of time had implied replacing relatively experienced managers with inexperienced ones and exploited workers with workers more focused on social justice than on innovation and entrepreneurship. In that context, let alone the context of an antagonistic international western world, achieving a high performance industry was a tall order and was going to require many years to become reality. The realistic challenge for Cybersyn would have been accepting that its meaning was opening the horizons for a better future rather than for an effective economy in the short term. However, at that time, it was difficult to see this distinction; it was ‘realistic’ privileging the technological implementation of a visionary project rather than building in the short-term more effective relationships.

We accepted the VSM as a liberty machine rather than as means to support processes to transform society; rather than focusing on social transformation our focus was on producing information systems and artifacts. The VSM had to be far more than a heuristic to map a hypothesized recursive structure for an industrial economy and its use required far more development to achieve social transformation. Beer’s emphasis in designing and implementing a liberty machine made less necessary both developing methodologies and clarifying epistemological considerations for the application of the VSM. However, it can be argued that because Cybersyn succeeded in designing and implementing it a platform for further exploration and future developments was created. It was the extraordinary energy that Beer imbued in an impossible task that avoided sending his vision into oblivion. Paradoxically, though Cybersyn was insufficient for its purpose of improving the cybernetics of the Chilean industrial economy, it was an anticipation for necessary technologies and means for achieving a better cybernetics in future societies; it made more likely the designing of social economies beyond the extremes of centralized bureaucracies and poorly regulated free-markets. These two extremes are ineffective as illustrated by the collapse of socialist economies in Eastern Europe and the short-comings of capitalist, free-market economies (see above the “23 things they don’t tell about capitalism”). Accepting that the VSM is still a paradigm waiting for its time, today it offers an option for designing self-organized social economies. Beer had clear the scope of Cybersyn for ‘designing freedom’ (Beer 1975), for ‘designing the system for organizations’ (Beer 1985), for providing a ‘platform for change’ (Beer 1975) and so forth. In Chile, through Cybersyn, Beer’s vision was offering organizational cybernetics as an alternative to achieve fairer social economies. Unfortunately, a wider appreciation of this vision was constrained by an impossible socio-political context, a weak implementation methodology and an information oriented epistemology; all these aspects contributed to blurring an appreciation of its longer term meaning. The project’s implementation failed matching the complexity of the social processes involved.

Coda

In 2008 Jorge Baradit published in Santiago, Chile, the novel Synco. The novel is about the Cybersyn Project. Synco was the Chilean name for Cybersyn. It was a composition of the ideas of information and control the Spanish word CINCO, ‘five’, and the number of systemic functions constituting the Viable System Model. It starts showing the Operations Room (figure 2), and setting it as a ‘retro-futuristic’ novel that takes place in 1979. Its assumption is that the coup of 11th September 1973 did not succeed and that the Allende’s government had continued for all those years with the support of General Augusto Pinochet. For Baradit the project’s purpose was “…converting Chile in the first Cybernetic State in history, underpinned by a network which anticipated in decades the Internet as we know it”. Chile appears as a neo-fascist State, dominated by the SYNCO machine, which controls all aspects of private and public lives. One of the protagonists, as he tries counteracting the state’s drift towards a technocratic rightwing society, says: “SYNCO, a god made of wires and a shared mind, a beehive, will establish the first technological dynasty in history…But we are building up an army of code breaking children. We have educated them in the secrets of SYNCO … a battalion of mind focused soldiers which will face up with their keyboards a new type of war for which they (the government) are not prepared”.  Another person states in relation to the government’s socio-economic direction that “The third way is an illusion” produced by a network of black covered cooper wires. Baradit appears to accept that a successful 1973 military coup was the lesser evil for Chile; the alternative was too awful to contemplate.  It is sad that this popularization of the SYNCO project appears to give credence to the fears expressed by the right wing political press just before the coup in 1973. It re-enforces the view that the cybernetic project could only have led Chile to a neo-fascist, totalitarian outcome. This trivialization betrays a profound lack of grasp of organizational cybernetics as a science for democratic governance and not for autocratic control.

Cybersyn did not succeed in re-constructing the nature of the Chilean society. Its vision was offering an alternative to the extremes of a centralized planning system or an unrestricted free market; it offered a third way for social cohesion and economic fairness. Cybersyn was a project ahead of its time. Its creation was visionary but unfortunately its intended implementation did not have requisite variety. The necessary social and organizational contexts to re-construct the nature of social relationships did not exist; however desirable it might have been to provide information in real-time and by exception, the necessary relationships for the cohesion and adaptation of the social fabric were not there. A mooted point is whether a longer period of implementation, uninterrupted by the coup d’état of September 1973, would have supported this requisite learning. Some of us in the project had a vague appreciation of the need to embody these relationships in the social fabric of the economy but collectively most of us did not see Cybersyn beyond being a powerful theoretical framework and our practice was biased towards a technical implementation at the expense of the values of building up a truly autonomous decentralized industrial economy and furthermore an inclusive democracy.

In conclusion, Cybersyn did not succeed in reconstructing a more humane and just social nature in the Chile of the 1970s; however, the safeguards against technocratic tendencies were precisely in the very implementation of Cybersyn, which required a social structure based on autonomy and coordination to make its tools viable. Without a culture of autonomy, coordination and inclusion these tools were too weak to have any social impact. The control against any autocratic tendency was intrinsic to its conceptual framework, the VSM. Of course politically it was always possible to use information technologies for coercive purposes however that would have been a different project, not SYNCO; its political and conceptual underpinnings were those of a democratic society and its tools were orders of magnitude less complicated than those required for centralized control. With the benefit of hindsight I think that had the 1973 coup failed, and had the people and its socialist government supported the 3rd way offered by Beer’s vision, Chile would have emerged, after several years of painful learning and development, as a more cohesive and fair society with a large social capital of engaged citizens.

References

Athanasiou, T., 1980. The Liberty Machine, Undercurrents, The Magazine of Radical Alternatives and Community Technology, No 38, February-March.

Axelrod, J. and G. Borenstein, 2009 “Free As In Beer: Cybernetic Science Fictions”  A paper delivered at the 2009 Pacific Ancient and Modern Languages Association Conference http://www.youtube.com/watch?v=hCO3vXyR-c4& list=PLF56B008626140E0A&index=1

Baradit, J. 2008. SYNCO, Ediciones B, Santiago, Chile

Beer, S. 1966. Decision and control. The Meaning of Operational Research and Management Cybernetics. Chichester: Wiley.

Beer, S. 1972. Brain of the Firm. London: Allen Lane, The Penguin Press.

Beer, S. 1974. Designing Freedom. Toronto, CBC Publications.

Beer, S. 1975. Platform for Change. Chichester: Wiley.

Beer, S. 1979. The Heart of Enterprise. Chichester: Wiley.

Beer, S. 1981. Brain of the Firm (Second Edition). Chichester: Wiley.

Beer, S. 1985. Diagnosing the System for Organizations. Chichester: Wiley.

Beer, S. 1994. Beyond Dispute: The Invention of team Syntegrity. Chichester: Wiley.

Beer, S. 1989. “The Viable System Model: its provenance, development, methodology and pathology” In R. Espejo & R. Harnden (Eds.) The Viable System Model: Interpretations and Applications of Stafford Beer’s VSM: 11-38. Chichester, UK: Wiley.

Chang, H.-J. 2010. 23 Things They don’t Tell You about Capitalism London, Allen Lane, Penguin

Espejo, R. 1980. ‘Cybernetic Praxis in Government: the Management of Industry in Chile 1970-1973.’ Cybernetics and Systems: An International Journal, 11, 325-38.

Espejo, R. 1989. ‘A Method to Study Organizations.’ In R. Espejo & R. Harnden (Eds.) The Viable System Model: Interpretations and Applications of Stafford Beer’s VSM: 361-82. Chichester, UK: Wiley.

Espejo, R., G. Bula, R. Zarama, 2001. “Auditing as the Dissolution of Corruption.” Systemic Practice and Action Research 14(2): 139-156.

Espejo R., Reyes A., 2001. The state of the state: introduction. Syst Pract Act Res 14(2):135–137

Espejo R. 2003. The Riscom Model, Boundaries and Contextualisation, In: Andersson K (ed) VALDOR; values in decisions on risk, Symposium, Proceedings. Stockholm, Sweden, pp 444–451

Espejo, R. 2009. “Performance Management, the Nature of Regulation and the CyberSyn Project.” Kybernetes 38(1/2): 65-82.

Espejo, R. 2011. “Seeing a case study through a cybernetic epistemological lens” Kybernetes,,40:9/10, 1273-96.

Espejo, R. & Reyes, A. 2011. Organizational Systems: Managing Complexity with the Viable System Model. Heidelberg: Springer.

Forrester, J. 1971. World Dynamics. Cambridge, Mass, Wright and Allen,.

Harrison P.J. and Stevens C. R., 1971. A Bayesian Approach to Short Term Forecasting. Operational Research Quarterly, December 1971.

Holtham, C., Lampel, J., Brady, C, and Rich, M., 2003. “How far can business war-rooms provide an effective environment for management learning?” Educational Innovation in Economics and Business (EDINEB 2003), Salzburg, June 2003

Maturana, H. and Varela F., 1992. The Tree of Knowledge. Boston & London, Shambhala

Maturana, H., 1988. “Reality: The Search for Objectivity or the Quest for a Compelling  Argument.” The Irish Journal of Psychology 9(1): 25-82.

Maturana, H., 2002. Autopoiesis, Structural Coupling and Cognition: a History of these and other notions in the biology of cognition. Cybernetics & Human Knowing 9:5.

Medina, E., 2006. “Designing Freedom, Regulating a Nation: Socialist Cybernetics in Allende’s Chile ” J. Lat. Amer. Stud. 38: 571-606

Medina, E. 2011. Cybernetic Revolutionaries: Technology and Politics in Allende’s Chile. Mass: MIT Press.

Kaplan, R. and Norton D., 1996. The Balanced ScoreCard; translating strategy into action. Boston, Mass, Harvard Business School Press.

Reyes A., 2001 Second-order auditing practices. Syst Pract Act Res 14(2):157–180

Rockart, John F., “Chief executives define their own data needs”, Harvard Business Review 1979 (2), pages 81-93.

Rosenblueth A, Wiener N, Bigelow J., 1943. Behavior, purpose and teleology. Philos Sci 10:18–24

Schuhmann, W., 2004. Observing Experiences with the VSM, in Tribute to Stafford Beer, Espejo, R. Guest editor Kybernetes, Vol. 33 No. 3/4

Simon, H. 1981. The Sciences of the Artificial, 2nd edition. Cambridge, Mass.: The MIT Press.

Teece, J. D. (2008). Technological Know-How , Organizational Capabilities, and Strategic Management : Business Strategy and Enterprise Development in Competitive Environments. London, World Scientific Publishing Co.

The Guardian, Monday 8 September 2003 “Santiago Dreaming” by Andy Beckett

The New York Time, March 28, 2008, “Before ’73 Coup, Chile Tried to Find the Right Software for Socialism” by Alexei Barrionuevo

Ulrich, W. 1994. Critical Heuristics of Social Planning. Chichester: John Wiley.

Varela, F., 1979. Principles of Biological Autonomy. New-York, Elsevier-North Holland.

Varela F. 1986., Steps to a cybernetics of autonomy. In: Trappl R (ed) Power, autonomy, utopia. Plenum, New York

von Foerster, H., 1984. Observing Systems. California, Interpublishers.

Wene C-O, Espejo R., 1999. A meaning for transparency in decision processes. In: Andersson K (ed) VALDOR; values in decisions on risk, Symposium, Proceedings. Stockholm, Sweden, pp 404–421

………………………………………………………………………………………………………………………………………………………………………….

Source: This article was originally published as Espejo, R. (2014b). “Cybernetics of Governance: The Cybersyn Project 1971-1973”, in Metcalf, G. S., Ed. (2014). Social Systems and Design. Translational Systems Sciences. Japan, Springer.

………………………………………………………………………….……………………………………..…………….. 

[*] Raúl Espejo (r.espejo@syncho.org) was, from 1995 until his recent retirement, professor of Systems and Cybernetics at the University of Lincoln UK. Before that date, from 1977 to 1994, he was senior lecturer at the University of Aston UK. His work is focused on organisational cybernetics, complexity and digital technology, mostly in social systems and policy making. He is currently president of the World Organisation of Systems and Cybernetics (WOSC). He is the author of the VIPLAN Method to study organisations, which was published originally as an e-book in 1996 and more recently in his book, with Alfonso Reyes, Organizational Systems: Managing Complexity with the Viable System Model (Springer, 2011). He is also responsible for the VIPLAN Methodology and the VIPLAN Learning System. Almost half a century ago he was operational director of the Cybersyn project in Chile, under the scientific direction of Stafford Beer.

…………………………………………………………………………………………………………………………………………………………………….

Notes

1. In fact more recursion levels were hypothesised but for the sake of this argument no more are necessary.  
2.   It was used a short-term forecasting model by Harrison and Stevens (1971).  
3. CIA orchestrated according to USA Congress reports
4. Beer defines cybernetics as the science of effective organization (Beer, 1975 p.425)
5. Jeremiah Axelrod and Greg Borenstein delivered the paper “Free As In Beer: Cybernetic Science Fictions” at the 2009 Pacific Ancient and Modern Languages Association Conference. The link for the video recording of this presentation can be found in the references (Axelrod and Borenstein, 2010).The paper describes “how British cyberneticist Stafford Beer’s writing, infographics, and industrial design for his ambitious Cybersyn Project worked together to create a science fictional narrative of omniscience and omnipotence for Salvador Allende’s socialist government in Chile”.