15.1 Why “pathology”?
In the 1984 paper “The Viable System Model”, reprinted in Think Before You Think, Beer defends his medical vocabulary with a single alternative: “either we have a theory of viability, meaning ‘capable of independent existence’, or we have not” (TBYT 2009, “The Viable System Model”: 147). He describes the failures of organizations as pathology and speaks of diagnosis, etiology and prognosis, knowing that many readers object to such terms when they are applied to social institutions or to technical networks. If there is a theory of viability, it applies to any system that meets its conditions, and so do the terms that describe the failure of those conditions. Beer adds that the method is more than analogy: “an heuristic such as the yo-yo technique is in search of a mathematical invariance that transcends analogy” (ibid.). Earlier in the paper he describes that technique: once a VSM has been constructed by mapping, say, a brain onto a firm, each further viable system is tested against the model by running “down the chain of similes, analogies and homomorphs” with it “until the isomorph is reached”, then back up the chain with another fresh system, and so on, “hence the yo-yo metaphor (rather than model, note)” (TBYT 2009: 138). On this basis he describes breakdown in the terms of his own model. Viable systems of all kinds are subject to breakdown, and such breakdowns “may be diagnosed, simply in the fact that some inadequacy in the system can be traced to malfunction in one of the five subsystems, where in turn one of the cybernetic features that compose the rules […] will be found not to be functioning” (TBYT 2009: 148). The “rules” are those of the glossary in HOE (chapter 3). A diagnosis therefore has two stages: locate the subsystem that fails, then identify which rule it fails to meet. Beer continues “unabashed with medical-sounding talk”: “the etiology of the disorder may be traced, a prognosis may be prepared, and antidotes (even surgery) may be prescribed” (ibid.). The vocabulary also answers a common objection, that any organization which exists is by definition viable. Existence, Beer replies, proves little. The fact that a system is there does not guarantee that it will stay there, and “The fact that it is there does not prove that it is effectively there, witness universities, nor efficiently there, witness hospitals” (TBYT 2009: 148). Monopolistic systems in particular “often operate at the margins of viability” (ibid.). He names one frequent cause, especially in the social services: people looking for cheaper ways of doing things “attempt to repeal the Law of Requisite Variety itself”, and policing of every kind “often fails to recognize that only variety can absorb variety” (ibid.). The last phrase restates Ashby’s Law in the form Beer uses throughout (chapter 2.3). The hospital that is there but not efficiently there returns in 15.4, where Beer gives it a test.
15.2 Four diagnostic points
In the same paper Beer quotes four diagnostic points “made in a learned journal” (TBYT 2009: 148). The wording is someone else’s. Beer says that all four “have been expounded in my own writings, but not I think with such pith”, and reproduces them as direct quotations (ibid.). In summary, they ask whether management presides over a viable system at all, whether System Five represents the whole system to larger systems, whether managers understand the need for Systems Two and Four, and whether Three, Four and Five form a “Three-Four-Five” subsystem (TBYT 2009: 148–149). Beer judges them “extremely cogent and penetrating” and notes that they show how the language of the model allows subtle comments “in very few words” (TBYT 2009: 149). His own comments on the four points follow (TBYT 2009: 149–150):
- Takeovers. Acquired companies are always assured that their individuality and autonomy will be respected. Beer calls this “poppycock”, because a study of the embedment of the new System One in terms of the Law of Cohesion (chapter 13) shows that the connections between the two recursions inevitably take up variety from the new subsidiary (TBYT 2009: 149). “In the VSM, ‘autonomy’ is a precisely defined term, and it does not mean zero interference” (ibid.).
- Identity. Adaptation to a changing environment can involve so many adjustments that identity itself becomes the issue: “Will the system still be able to recognize itself? More particularly, will others be able to recognize it?” (ibid.). Beer refers to the Test of Identity in HOE (TBYT 2009: 150; see 15.5); identity as a condition of viability is discussed in chapter 5.2.
- Collapse of Five into Three. This is likely “insofar as Five people have usually been promoted from Three”. They are uncomfortable in their new role, and when something goes wrong “they are likely to dive down into the problem that they understand so well—never to emerge again” (TBYT 2009: 150). “But the collapsed metasystem is a special pathology. It is a decerebrate cat […]. It responds reactively, from the autonomic command centres at Three, and is incapable of planning and foresight (Four) and will and judgment (Five)” (ibid.). The Heart of Enterprise describes the same pattern: when a crisis occurs, a member of System Five “dives down from System Five straight into System Three, or (worse) straight into a muddy bog in System One. He is never seen again” (HOE 1979: 266). HOE draws the conclusion in the same image: “Thereby do our institutions become decerebrate cats. Management is noticeably failing the world, in the absence of metasystem. It can just manage to say miaow” (ibid.).
- Three–Four–Five. The fourth point makes the case for a management center (chapter 11), where “the Three-Four-Five metasystem has a chance to find its own cohesion” (TBYT 2009: 150). Beer ends the paper with a proposal that follows from the medical vocabulary. Since much has been learned about viable systems from those that are “viable no more”, a small team of “organizational pathologists” should be formed, organized “more like a lifeboat crew”, whose first task would be to resuscitate a dying organization and, failing that, to perform a post-mortem “before rigor mortis had set in” (TBYT 2009: 150). He observes that “history is rewritten in these circumstances with breathtaking speed” (ibid.).
15.3 System Four missing
The collapse of Five into Three has a precondition that HOE states directly. In the preceding chapter of that book Beer had argued that System Four is “often, indeed usually, virtually empty”, because its components are scattered, not in communication with one another, and labeled “staff” or “merely advisory” (HOE 1979: 265). The consequence follows from the structure of the metasystem: “Without a System Four clearly in place, and with a System Five whose very nature is ambiguous, there is no Three-Four interaction, and no Five monitoring of that interaction. Then the whole metasystem collapses into System Three” (HOE 1979: 265). Beer calls this a practical point, “Everywhere one turns, the reality is the same”, and adds the theoretical one: in terms of the neurophysiological model, “the institution is a decerebrate cat” (HOE 1979: 265). He works the image out with some care. A decerebrate cat lives on, but “its viability is ensured by a bogus environment, and is sustained by artificial sustenance. If you prod its leg, it kicks back. And this is called ‘living’” (ibid.). Thought through in institutional terms, “we diagnose a major pathology of our times” (ibid.). The pathology links two parts of the model. The absent System Four (chapter 10) removes the Three–Four homeostat (chapter 11), and without it System Five has nothing to monitor and no function of its own, so it is drawn into the day-to-day management of System Three.
15.4 Pathological autopoiesis
In the last chapter of HOE Beer goes a step further. The pathologies discussed so far concern one level of recursion, and he summarizes them: “it should be very clear by now why Systems Two and Four are normally vulnerable subsystems, why the monitoring devices of the Three-Four balance often fail, why System Five often collapses into System Three, and why the metasystem often lacks a capabilty [sic] to speak the metalanguage” (HOE 1979: 408). Then he names a deeper failure: “The ultimate pathology of the viable system concerns the failure of its cohesiveness, and of its inter-recursive algedonics. This turns out to be an aberration of its autopoietic function” (HOE 1979: 408). To explain it he takes up Maturana and Varela’s concept of autopoiesis. The concept enters HOE as a statement about identity. In the overview Beer writes: “The primary characteristic of a viable system is not the power of self-reproduction, but of self-production. It is continuously in business to produce itself, to be what it is, to preserve identity” (HOE 1979: 277). Institutions such as a hospital or a university replace all their members over time and yet remain recognizably themselves: “There is, as in our bodies, every kind of change. But there is no alteration. In the concept of autopoiesis we have the final testimonial to viability” (HOE 1979: 405). Already in the overview, however, Beer points to the danger: “Too much organizational effort may go into maintaining the status quo as an end in itself, so that the job that the enterprise is supposed to do becomes almost a by-product. This is the last of our diagnoses” (HOE 1979: 277). He states the pathology with a biological comparison: “We need a thyroid gland; its pathology is to become a goitre. Enterprises need to be autopoietic; their pathology is to turn self-production into an end-in-itself” (HOE 1979: 410). For the hospital he proposes a simple test: “what proportion of everybody’s time is spent on healing, and what proportion on the autopoiesis of the medical profession?” (ibid.). Where, then, does the healthy proportion lie? Beer first considers a numerical constant, a fixed percentage of effort for necessary self-maintenance, and rejects it: purposes vary between systems, and “to seek a numerical constant of the x-percent kind for the maintenance of necessary autopoiesis, is to pursue a chimera” (HOE 1979: 411). The invariant is structural: “The proportion of total time that needs to be devoted to maintaining necessary autopoiesis is exactly the time required to maintain System One — which is the viable system at the next level of recursion — whatever that may be” (HOE 1979: 411). The Law of Cohesion (chapter 13) determines this proportion in each case. The criterion follows. Only a viable system is autopoietic, and “Systems Two, Three, Four, and Five are not in themselves viable systems at any level of recursion” (HOE 1979: 411). Healthy autopoiesis is therefore only that of System One. Systems Two to Five are “services to the autopoiesis of System One” and should “exhibit no internal autopoietic behaviour whatever” (HOE 1979: 412). “It is precisely when these subsystems of the viable system exhibit internally autopietic [sic] behaviour (whereupon the total viable system becomes pathologically autopoietic), that people start to talk […] about ‘The Establishment’” (ibid.). Beer gives two biological illustrations: a tree brought down by ivy, which he recognized as “a System Two failure”, and “a cancerous tumour” as an example of pathological autopoiesis beyond failed symbiosis (HOE 1979: 412). His definition reads: “‘The Establishment’ in any social system comes into being at the point when the vital principal [sic] of autopoiesis consumes energy greater than that needed to maintain cohesiveness through the appropriate number of viable recursions that mark its claim to organizational identity as a set of embedments of System One” (ibid.). He adds: “‘The Establishment’ presents autopoietic activity on the part of Systems Two, Three, Four, or Five; and this constitutes a pathological symptom of the viable system” (HOE 1979: 412). Here several threads of the book meet. The Law of Cohesion defines how much metasystemic effort is needed to hold the recursions together. The algedonic channel (chapter 12) is the mechanism that keeps the recursions in contact. On my reading (an interpretation, not Beer’s statement), when the metasystem begins to produce itself instead of serving System One, it consumes more than cohesion requires and at the same time loses its responsiveness to the signals from below. That is why Beer can call the failure of cohesion and of inter-recursive algedonics a single pathology, and why he locates it in the autopoietic function (HOE 1979: 408).
15.5 The Test of Identity
In the 1984 paper Beer writes that “The Heart of Enterprise includes a highly sophisticated Test of Identity” (TBYT 2009: 150). HOE does not use that name. It contains one text that tests identity: “Identity: an exercise”, placed after the last chapter and its bar dialogue, at the end of Part Three (HOE 1979: 415–433; listed in the contents at p. 417). That this is the text Beer means is my identification, not his statement; the exercise matches the point the paper raises, whether a system and others can still recognize it. The exercise asks a single question: “Does the organization have the power to do what it says that it will do — given that it has done its best to define its terms, state its goals, and set up regulatory systems to these ends?” (HOE 1979: 417). Power here is neither legal nor financial capability. It is “the ability of the organization to gather itself together into an effective identity, to act as a unified whole” (HOE 1979: 417–418). Beer uses the image of a jigsaw: “The problem of identity is this: have we seen the picture on the lid?” (HOE 1979: 418). The puzzle came without a lid, and in a large organization it is too big to be seen whole from a distance, so the identity has to be understood from within. Consensus therefore has to begin with “a common view of the identity itself” (ibid.). The exercise is written without jargon, for managers (HOE 1979: 417). The organization is divided into its main blocks, the Elements, which cohere in the Whole; service activities are left out (HOE 1979: 419). Fifteen questions follow. Each is answered by marking a card on a scale from low to high, without numbers or weights, and laying it face down to avoid halo effects (ibid.). Questions 1 to 4 ask how far the senior team shares a picture of what it manages, how far the Elements agree with each other about the identity of the Whole, how far their view matches the Whole’s own picture, and whether the Whole correctly assesses the gaps. Of the fourth Beer says “we could call it a meta-question”: it concerns “the ability and honesty of the Whole to observe its own behaviour correctly and in context” (HOE 1979: 423; questions 1–4 on pp. 420–423). Question 5 asks how well the Whole’s identity is transmitted to the Elements (HOE 1979: 424). Questions 6 and 7 turn to autonomy. The “Degree of Autonomy is a crucial factor in Identity”, and Whole and Elements rarely agree on it (ibid.). The principle Beer proposes is “that the amount of intervention necessary in Elemental operations is only that required to support the coherence of the Whole”, and question 7 asks whether the information passing between them is judged correctly for volume and level (HOE 1979: 425). Questions 8 to 11 concern the network of committees, task forces and judgments that decides matters bearing on identity, which Beer calls “Identity Regulation”: is it adequate, do the Elements understand it and use it, and is it self-organizing, “LEARNING”, “ADAPTING”, “EVOLVING” (HOE 1979: 426–428)? Questions 12 to 14 ask how much of the effort of management, of the work force and of the respondent is “honestly addressed to the intended job (as distinct from ritual in all its forms)” (HOE 1979: 429). Question 15 asks whether the enterprise is aware of these issues at all (ibid.). Beer adds: “Perhaps we should think of institutional failure less in terms of an unsupportable balance sheet, or industrial unrest, and so on, and more in terms of the collapse of its identity” (HOE 1979: 430). For the evaluation the fifteen marks are transferred to one chart, Figure 70, “An impression of the power of the enterprise to act as an effective integral whole” (HOE 1979: 430). What counts is the uniformity of the marks more than their position; “an almost random scatter” suggests that “the integrative system is quite probably out of control” (HOE 1979: 431). For a group of managers Beer proposes digitizing the cards and clustering them by computer, both by manager and by question. Clusters of managers show differences in perception between parts of the enterprise. Clusters of questions have a structural meaning, since the questions are based on the viable system model: “So it ought to be possible to discover through this exercise and its evaluation just how successfully the law of cohesiveness is working in a given enterprise; and maybe (one day) thresholds will be established, to exceed which will predict collapse” (HOE 1979: 432). In the first trial, managers who “cheerfully asserted that they shared a corporate identity” fell into clusters that separated their two divisions (HOE 1979: 433). On my reading, questions 12 to 14 are the working form of the pathology in 15.4: the ritual that preserves “accepted structures in the organization” (HOE 1979: 428) is what Beer, in the last chapter, calls self-production turned into an end in itself.
References
- HOE — Beer, Stafford (1979). The Heart of Enterprise. Chichester: John Wiley & Sons.
- TBYT — Beer, Stafford (2009). Think Before You Think: Social Complexity and Knowledge of Knowing. Edited by David Whittaker. Charlbury: Wavestone Press.