Monday 17 September 2012


The Viable System Model and Structural Recursion

Stafford Beer’s Viable System Model and in particular his Recursive System Theorem are often the subject of discussion. In this blog rather than explaining Beer’s work I want to explain how I use it. Indeed I’ve used the VSM for a long time, since the old days in Chile, and in particular I’ve applied the idea of structural recursion to enterprises of all kinds. 
I see structural recursion as a possible consequence of the often tacit strategies used by members of an organisation to manage the complexity that they create, regulate and produce as they deal with a self-defined problematic situation in their medium/environment. I use the term create to point at the meanings (policies, purposes, values and products) that they define to guide their collective actions. These actions produce the organisation’s meanings as experienced overtime by stakeholders. Aligning meaning creation and production requires regulation supporting the organisation’s cohesion and adaptation to its medium/environment.
It is apparent that members of any organisation, whether they are senior managers or shopfloor workers, share the same biology and therefore have similar complexity management capabilities, but often their potentialities are very different. Since potentially problematic situations can imply exceedingly large complexities, a strategy to deal with them is chunking the global problematic situation into chunks (or primary activities as I call them in my work) that we may expect people strive to align with global policies, purposes and values. If these aligned chunks manage to be autonomous themselves and succeed creating, regulating and producing their particular policies and products, the structural and processual implications are amplifying, perhaps in orders of magnitude, the mutual constitution of organisational and environmental complexities. In the end achieving a constructive synergistic ‘unfolding of complexity’ hugely increases the overall situational and organisational complexities. This implies that in the same way that the global organisation articulates its problematic situation, the aligned chunks create, regulate and implement their own meanings; in other words are autonomous themselves creating and dealing with much more complexity than if they were following hierarchical instructions. Indeed this unfolding (i.e. chunking) may now happen within each of the chunks and therefore we are now witnessing a proliferation of situational and organisational complexities relying on the creativity, flexibility and inventiveness of their members. The number of structurally recursive levels in an organisational system is the number of autonomous structural levels producing the organisation’s self-defined meanings. 
Thus, each of the chunks needs resources and competencies to create, regulate and implement their self-defined products; these are systems 5, 4, 3, 2 and 1 in Beer’s VSM or the systemic functions of Policy, Intelligence, Cohesion, Coordination and Implementation in my own language.  These functions must recur in all viable chunks at all structural levels. More significantly, communications among them in the organisational system and with agents in their medium/environment produce recursive relationships of performance, cohesion and adaptation. In viable organisations these relationships emerge from effective complexity management strategies, which rely as much as possible on self-regulation and self-organisation for balancing what in general are asymmetrical varieties (I’m using variety in the Ashby’s sense). This recurrence of systemic functions and relationships for each autonomous chunk is for me the meaning of structural recursion.  The platforms for this recursion are the meanings or policies created by the members of the organisational system, which give them the closure of an autonomous organisational system. These policies are their purposes, values, norms emerging from their communications with agents constituting problematic situations and also most significantly, from their cultural contexts. These are multiple loops of circular causality, boot-strapping and so forth.
Beer’s first insights about this model came from his appreciation of the complexity management strategies of our nervous system as humans co-evolved with their medium. Methodologically, the insights coming from a scientific situation (neurophysiology) helped him to offer a model relevant to organisations and management. This methodology, which connects nature’s evolution and learning, is used in artificial intelligence and engineering today (e.g. mimicking the behaviour of ants, bees and birds).  With adequate methodological support, particularly accounting for people’s purposes, it can also be a most useful methodology for social situations.
I’ve applied structural recursion in a wide variety of situations; there are easy cases where the managerial situation clearly recognises autonomous chunks; this is the case of large corporations that over the years have evolved from functional, centralised structures, to autonomous divisions and strategic business units.  But, for smaller units in SBUs as well as for small and medium sized enterprises it is not uncommon to find that there is no structural recursion; simply, often, chunks are the outcome of hierarchical relationships. Indeed, empirical evidence may suggest that no recursion is happening.  Indeed in all cases recursion can only be established empirically. Of course we can always recognise shortcomings in these ‘unfoldings of complexity’ but in one form or another we start hypothesising chunks and levels creating, regulating and producing their own meanings and then test the hypothesis. There are structural and identity shortcomings that tend to be archetypical, which in practice limit the scope of recursive chunks (see Organisational Systems; Managing Complexity with the Viable System Model, by R. Espejo and A. Reyes, Springer 2011).
More interesting examples of complexity unfolding happen with reference to government policies,  such as health, education, nuclear waste management and so forth, where we find that processes of self-organisation evolve towards some form of structural recursion of the multiple institutional parts creating, regulating and producing that policy, which overtime develop connectivity that eventually may give viability to the policy.

Thursday 22 March 2012

Review of Cybernetic Revolutionaries by Eden Medina, MIT Press, Cambridge Mass, 2011

March 2012

In this book Eden Medina offers a historic review and reflexion of an unlikely project; Cybersyn. It happened in the Chile of Presidente Salvador Allende in the early 1970s. The book offers a most compelling account of an attempt to use science and technology in support of the management of a highly turbulent political process. I was its operational director and therefore had a privileged participation in its unfolding.
Cybersyn, the brainchild of Stafford Beer, was conceived and developed at the same time of publishing Brain of the Firm, the first of his trilogy about the Viable System Model (VSM). He had been pondering about this model for some years and Chile offered the opportunity to use it.
Eden’s historic account is well researched and balanced. As a participant, this account resonates fairly well with my memories. Indeed I disagree here and there with particular recollections and in occasions, as I illustrate below, I would have put the emphasis differently, but overall I feel comfortable with her account even if sometimes it is not favourable to individuals or the project as a whole. Yes, I would have emphasised differently the roles of the Operations Research and Design teams in the project. I would have liked to read more about the huge contribution of the Operations Research group at the State Technology Institute. This team was responsible for modelling some of the technological processes of the public and nationalised enterprises and for designing performance indices to support managers. On the other hand I felt that the role of the team that designed the Operations Room (also operating from the same Institute) appeared over emphasised. No doubt that this latter team made a most powerful contribution to the project, but its overall scale was smaller than that suggested by the book.
Cyberstride, the idea of managing in real time the economy, was the driving force of Cybersyn and in my view it was Stafford’s most powerful vision at a time when global and local management were dominated by historic reporting operating with huge time lags. His concern was reducing the complexity natural to the industrial activities of the country to meaningful levels for effective managerial action, respecting the autonomy of the people, enterprises and overall industrial economy. He referred to this concern as variety engineering. In this engineering he saw computers as nodes in action networks rather than as number crunching machines. At a time when computers were used to routinize operational activities, Beer saw the role of computers in society and the economy as machines to give information to the people for them to act and control their destinies.
Eden rightly gives to Beer’s Viable System Model an important role in the project; however she makes apparent that “Beer was more interested in studying how systems behaved in the real world than in creating exact representations of how they function.”… “Beer’s emphasis on action over mathematical precision set him apart from many of his peers in the academic operations research community who, Beer believed, privileged mathematical abstraction over problem solving”
Stafford arrived to Chile with the manuscript of the book “Brain of the Firm”, which instantly captured the imagination of all of us and also of people further afield. Eden clearly states that from a historic perspective her explanation of the VSM had to be rooted in that book and not in its further developments after Cybersyn and she offers a good introduction to the model. Perhaps what this introduction does not make apparent is that Stafford had not spent much time in methodological considerations for its application and that in fact an important contribution of the local team was unravelling its use. However, the VSM was not used to model the Chilean economy but as a reference for engineering its variety. It gave us a heuristic for designing indices of performance at a number of structural levels, which were hypothesized as recursion levels of the industrial economy in line with the insights of the VSM.
Eden gives compelling evidence about the unavoidable interdependence of technology and politics. Cybersyn, as a technological device, could not free itself from the on-going politics of the day. Its relevance to politics is clearly instantiated by the increasing influence of Fernando Flores - the political leader of Cybersyn- on President Allende’s decisions. At the same time her account of the project itself makes apparent the limited relevance of Cybersyn in the Chilean economic scene. This was the case in spite of Beer’s efforts to catch up with the political chaos. She illustrates this conflict between politics and technology with reference to the publicity received by the project at the time and the project’s schizophrenia. Stafford’ main public speech about Cybersyn at the time was the Richard Goodman Lecture in the UK and she states with reference to this lecture that “By emphasizing technology instead of Cybersyn’s relationship to the social and economic goals of Allende’s nationalization program, Beer failed to definitively separate himself from the technocrats he criticized.”
This book offers a wonderful story about unlikely events that happened 40 years ago that are still relevant today. Personally, with the benefit of hindsight, I could make many criticisms to the work of those difficult but adrenaline-charged days, but in a book with a historic emphasis it would be unfair to criticise Cybersyn with the eyes of the 21st Century and certainly Eden Medina does not do that as she offers a balanced a well contextualised account of Cybersyn.


Raul Espejo