content

15 min read

4.1 Why the nervous system?

In the summary of Part Two of Brain of the Firm (1972; 2nd ed. 1981) Beer states what the whole exercise is for: “The object is to construct a model of the organization of any viable system. The firm is something organic, which intends to survive — and that is why I call it a viable system” (BOTF 1981: 75). The book develops the model by comparing the firm with the human nervous system. Readers who know the model only from diagrams tend to take that comparison as a teaching aid, a vivid picture that could be removed without loss. Beer’s own account of the choice gives a different reason, and his steps are best followed in order. The sentence just quoted is the first step, and two points in it carry the later argument. The model is meant to apply to any viable system, not only to the firm. And the firm qualifies as a viable system because it intends to survive; viability is defined through survival, not through a particular form of organization (see chapter 5). The second step is a criticism of the usual way of describing organizations. Instead of taking a system known to be viable as the model, management uses organization charts, “that are really devices for apportioning blame when something goes wrong” (ibid.). The criticism recurs in the other books (MS 1967: 98–99; DTS 1985: i; BD 1994: 3). In Diagnosing the System for Organizations Beer gives the reason in one sentence: “the organization chart makes no attempt to model VIABILITY” (DTS 1985: 2). The chart shows who reports to whom. It says nothing about how the organization keeps itself going. The third step follows directly. To understand the principles of viability, study a system already known to survive: “So if we want to understand the principles of viability, we had better use a known-to-be-viable system as a model, and that is why Part Two embarks on an account of the way the human body is organized and controlled by its nervous system” (BOTF 1981: 75). On the same page Beer insists that the model is not an analogy: “Models are more than analogies: they are meant to disclose the key structure of the system under study” (BOTF 1981: 75). An analogy would say that the firm is in some respects like a body. A model in Beer’s sense claims that the firm and the body share a structure, and that this structure is what makes both of them viable. The claim rests on a thesis of cybernetics that Beer states in the same passage. Any other viable system would have served, “such as the amoeba, or a whole animal species”. The results would be the same, “as they must be if viability as such has its own laws and enshrines its own principles (as cyberneticians contend)” (ibid.). So the choice of the human body is, on Beer’s account, a practical one. He gives two reasons. The body is “perhaps the richest and most flexible viable system of all”, and every reader has some insight into it (ibid.). The second reason is didactic. The first, on my reading, is more than that: a rich and flexible system displays more of the structure that any viable system must have. At the transition to the anatomy chapters Beer repeats the point: “There can be no interest in analogy at this point: a useful model (as already discussed) must be able to handle differences in scale, transference, workability, and appropriateness in convincing style” (BOTF 1981: 87). Scale, transference, workability and appropriateness are criteria for a model that can be put to work. A figure of speech needs none of them. A later remark in DTS, from a different book and a different context, belongs to the same line of thought: “a model is neither true nor false: it is more or less useful” (DTS 1985: 2). Read together with BOTF, it sets the standard by which Beer wants the neurophysiological model judged. Whether the firm “really is” a body is beside the question; what counts is whether the model discloses the structure that matters for viability and can be used. Beer reports how the comparison held up as he worked it out: “The more the neurophysiological model is studied in detail, the more astonishingly accurate does it seem to be in reflecting what happens in the firm” (BOTF 1981: 109). He follows this with an example at a finer grain. When an engineered control system replaces a local management centre, the unofficial connections between local management offices risk being cut; Beer sets this beside what surgeons observe when sympathetic ganglia are removed (BOTF 1981: 109–110). This is his own assessment of the fit and proves nothing by itself, but it shows how seriously he took the correspondence. Textual note: This answers the open question from chapter 1.6. The rejection of the analogy reading is in Beer’s own text (BOTF 1981: 75, 87), not only in the jacket text.

4.2 The five levels in the nervous system

Beer’s summary of Part Two lays out the argument (BOTF 1981: 75–76). It proceeds by necessity: each new level is introduced because the levels before it leave a task unsolved.

  • Systems One, Two and Three are “concerned, automatically, with the regulation of internal stability” (BOTF 1981: 76). Their model is the autonomic nervous system, which the summary calls “the autonomous nervous system” (ibid.): “Three vital systems are identified as prerequisites of all autonomous control” (ibid.).
  • System Four is needed because the organism “needs also to maintain dynamic equilibrium with the external world”, and because there “must be systems for arousal and adaptation” (ibid.).
  • System Five is “the level of conscious direction (attributable to the board of the firm and to the cerebral cortex in the brain)” (ibid.). The first three systems keep the organism stable from the inside, without conscious attention. Early in Chapter Eight, on autonomy, Beer says what such a level is: “Both systems operate autonomic control, which is to say a level of management which does not involve conscious direction by the organism as a whole” (BOTF 1981: 103). Its function, “from the point of view of the whole organism, whether body or firm”, is “essentially to maintain a stable internal environment” (ibid.). Chapter 5 takes up the argument about autonomy that Beer builds on this. Internal stability is not enough, because the organism lives in an environment that changes. System Four is therefore introduced as the part that faces outward and prepares adaptation. System Five is the level at which the whole organism directs itself. Beer places it in “the ‘thinking chamber’ of System Five”, from which a “vertical command structure” originates “intentions (as distinct from reflex responses)” (BOTF 1981: 135). In the body of the text Beer assigns the systems to anatomical structures. The assignments below are Beer’s own statements; the connecting remarks between them are mine.
  • Sympathetic and parasympathetic system: The autonomic control of the organs sits in the thoracolumbar outflow of the spinal cord, “the sympathetic nervous system” (BOTF 1981: 104). The sympathetic trunks link the peripheral ganglia; “these paravertebral ganglionic chains […] really govern the stability of the internal environment” (BOTF 1981: 108). In the same passage Beer names their function: they are “the feedback regulators and integrators” (ibid.). “If the sympathetic system is, as it were, a middle management function, then the parasympathetic system is a senior management function” (BOTF 1981: 110).
  • System Two as spinal cord: In the managerial application Beer calls System Two “the analogue of the spinal cord” (BOTF 1981: 124). The sympathetic trunk “links together the vertebral nodes of System Two” (BOTF 1981: 129). Each division’s control centre is a “vertebra” of System Two: “it is the System One command centre itself” (BOTF 1981: 125).
  • System Three and the brain stem: System Three is “a naturally sleepy or inattentive controller. It is wholly alert downwards […]. But its own major controllers of corporate, that is upward, awareness are essentially inhibitors” (BOTF 1981: 140). Beer links this to the sleep centres in the pons and medulla (ibid.).
  • Transition to System Four: “The hypothalamus is the floor of the third ventricle, and very much the link between the third and fourth systems of the model under discussion. It is the highest element in System Three, or the lowest element in System Four” (BOTF 1981: 141). The arousal mechanism is the ascending reticular formation (BOTF 1981: 95, 141).
  • System Four as “the biggest switch”: System Four is “the great linking mechanism between volitional and autonomic control. It is the biggest switch of the whole organization” (BOTF 1981: 135).
  • System Five as cortex: The direction of the enterprise, with its foresight, “is the thinking part of the whole organization. This, for the body, is the cortex itself — and for the firm it is the senior management” (BOTF 1981: 201). Two features of this list deserve comment. Beer’s wording in the managerial chapters is not always free of the language of analogy. He calls System Two “the analogue of the spinal cord”, and he introduces the sympathetic system “as it were” as middle management. The rejection of analogy in 4.1 concerns the status of the model as a whole; it does not stop him from using comparative language when he maps single parts. And the boundaries between the systems are not sharp in the anatomy. The hypothalamus is “the highest element in System Three, or the lowest element in System Four”. Beer does not hide this. He uses it to show where one system hands over to the next. The bibliography of BOTF follows the same assignment. Literature on the autonomic nervous system is listed “On Autonomics (Systems One-Two-Three)”, on the reticular formation “On ‘arousal’ mechanisms (System Four)”, and on the cortex “On the Cortex (System Five)” (BOTF 1981: 407).

4.3 What changes in the later books

Only BOTF derives the model from the nervous system. The Heart of Enterprise (1979) develops the same model again from variety analysis, without the neurophysiology (see chapter 3). Diagnosing the System for Organizations (1985) presents it as a working method. That the same five systems can be reached by a second route, without reference to the body, fits Beer’s claim in BOTF that the results “are the same” whichever viable system serves as the model (BOTF 1981: 75). On my reading, it also means that the model does not stand or fall with every neurophysiological detail: the neurophysiology is one of two derivations. Two differences between the books matter for this text:

  • Status of System Two. In BOTF: “System Two is the metasystem subsuming all Systems One. Throughout this book its existence has been diagrammatically indicated by the tall thin rectangular box drawn round the column of boxes which are themselves the System One” (BOTF 1981: 172). In HOE, by contrast, System Two is “a service to System One” with “the highly specific function of damping oscillations, and nothing else” (HOE 1979: 177). DTS calls it “the viable system’s anti-oscillatory device for System One” (DTS 1985: 66). The function, anti-oscillation, stays the same; its placing in relation to the metasystem shifts. In BOTF, where System Two is modelled on the spinal cord and its sympathetic trunks, it is a structure that runs through the whole column of divisions. In HOE, where the derivation starts from variety and oscillation, it is defined by the single task it performs.
  • Wording of the Recursive System Theorem. BOTF: “If a viable system contains a viable system, then the organizational structure must be recursive” (BOTF 1981: 228). HOE: “In a recursive organizational structure, any viable system contains, and is contained in, a viable system” (HOE 1979: 118). Beer himself calls the HOE version “an alternative version […] from the opposite angle” (ibid.). The practical consequence for the reader of this book: when a question concerns the derivation of the five systems from biology, BOTF is the source, with its spinal cord, brain stem and cortex. When it concerns the definition of a system’s function, HOE and DTS give the later and in some cases narrower formulation, and the differences should be stated rather than smoothed over. The Recursive System Theorem, in both its wordings, is the subject of chapter 6.

4.4 Control physiology and the figures of the derivation

Chapter Seven of BOTF, “Control physiology” (BOTF 1981: 89–102), comes before the autonomic chapters on which 4.2 draws. It describes the somatic nervous system and supplies the vocabulary the later chapters use. The first element is a pair of axes. A slice across the spinal cord shows an afferent–efferent mechanism: “Inputs come in at the back of the tube, and outputs leave from the front” (BOTF 1981: 90). Reflexes work across the cord at each level, so Beer speaks “of a lateral command axis”; bundles of nerves also run up and down “the central command axis” (ibid.). “Therefore we have an essentially two-dimensional system. This is one organizational secret of the body’s ability to run affairs — a particular organ for example — autonomously (working laterally), and also to integrate that local activity into an organic balance (working vertically)” (ibid.). The second element is the work of the “specialized computers” that lie on the vertical axis: a slice through the cord, a swelling on the brain stem, a lobe of the cortex (BOTF 1981: 93). Each has three tasks, which Beer lists “in the form of instructions” (BOTF 1981: 95). The first is to recognize data on which command action should be taken, take that action, “and send on the original information, suitably modified”. The second is to filter at that level, “compressing, facilitating, inhibiting the ascending path”. The third is to “Store a record of these transactions, in case details have to be retrieved” (ibid.). A filter, in Beer’s definition, is “a many-one variety reducer” (BOTF 1981: 93). The arousal mechanism is such a filter. Its anatomical name is the ascending reticular formation, and “this filter is one-way” (BOTF 1981: 94–95). The third element is the number five. “We are confronting what seems to be a five-level hierarchy of systems contained within a major computer configuration. I say ‘seems to be’ because the choice of five (rather than some other number) is somewhat arbitrary” (BOTF 1981: 98). Five levels are distinguished “because of the major functional differences involved, and we do not want to differentiate any further than we must”; all five lie on the vertical command axis, “and they model the somatic nervous system of the body” (ibid.). Figure 14, captioned “Exploded diagram of the brain (compare Figure 13) showing classification as five-tier hierarchy”, labels them Control Echelons I to V: the spinal vertebral level; the spinal cord; mesencephalon, pons, medulla and cerebellum; diencephalon, basal ganglia and third ventricle; the cerebral cortex (BOTF 1981: 99). Figure 13 shows the general layout of the brain (BOTF 1981: 91). From the fourth echelon Beer draws a lesson for management. Everything the cortex receives is sorted and switched through the diencephalon and basal ganglia. “In orthodox management talk this level has nothing to do with command. In our model it has everything to do with command: it lies smack on the vertical axis itself” (BOTF 1981: 98). He rejects the line/staff dichotomy as “chimerical” and wants staff advisers to accept “their role as commanders-one-down overtly” (ibid.). Chapter Seven speaks of control echelons and levels; it does not use the names System One to System Five. The names appear in the summary of Part Two (BOTF 1981: 76) and, in the managerial application, from Chapter Nine onward (e.g. BOTF 1981: 124). On my reading, the echelons of Figure 14 are the first, somatic version of the five-tier hierarchy, and the model’s Systems One to Three take their form only when the autonomic system is added in Chapters Eight and Nine. Beer’s figures trace that step. Figure 22, “The automatic system of a firm having subsidiaries A, B, C, and D” (the caption reads “automatic”), shows the boxes 5, 4 and 3, the column of controllers 1A to 1D, and the outside world (BOTF 1981: 130). Figure 23 shows the “Two dimensions of neurophysiological control: the main vertical command system (somatic) and the sympathetic and para-sympathetic systems (automatic)”, and Beer asks the reader to compare it with Figure 22 (BOTF 1981: 131). Figure 27, printed without a caption, is the model itself (BOTF 1981: 157). Its labels name the corporate management (Board Level, Development Directorate, Operations Directorate), the Corporate Regulatory Centre, the Divisions with their Divisional Directorates, and the Divisional Regulatory Centres (ibid.). Beer ties it back to the derivation: “It may be compared at once with its neurophysiological counterpart in Figure 23, and with the idealized (content-less) chart at which we arrived in Figure 22” (BOTF 1981: 156). Figure 27 is the figure chapter 6 refers to when Beer reproduces the model within its own units (BOTF 1981: 228).

References

  • MS — Beer, Stafford (1967). Management Science: The Business Use of Operations Research. Garden City, NY: Doubleday, in association with Aldus Books, London.
  • HOE — Beer, Stafford (1979). The Heart of Enterprise. Chichester: John Wiley & Sons.
  • BOTF — Beer, Stafford (1981). Brain of the Firm. 2nd ed. Chichester: John Wiley & Sons. (1st ed. London: Allen Lane, 1972.)
  • DTS — Beer, Stafford (1985). Diagnosing the System for Organizations. Chichester: John Wiley & Sons.
  • BD — Beer, Stafford (1994). Beyond Dispute: The Invention of Team Syntegrity. Chichester: John Wiley & Sons.