---
title: "Ch10 six components content"
description: "The second edition of Brain of the Firm, published in 1981, ends with a Select Bibliography in four sections. Beer titles the last of them Explicit…"
url: "https://wiki.cybernetic-futures.org/punch-and-judy/part-02-outside-the-model/ch10-six-components/ch10-six-components-content"
modified: "2026-11-12T15:26:25.895Z"
language: "en"
site: "Cybernetic Futures"
---

# Six Components, Six Literatures, No Beer

The second edition of *Brain of the Firm*, published in 1981, ends with a Select Bibliography in four sections. Beer titles the last of them Explicit References. It runs to Ashby, McCulloch, Shannon and Weaver, Forrester, Bremermann, Waddington, Thom, and *The Pauper's Cookbook*, which is there because Beer consulted it often enough to judge that it "must be of some importance," and because it gave the preface to the new edition its last line: "brains need gentle handling or they are apt to disintegrate."

It does not contain Herbert Simon, whose paper on the architecture of complexity is nineteen years old by then. It does not contain Arthur Koestler, who had named the holon fourteen years before. It does not contain Elliott Jaques, whose subject was a near neighbor of the thing the book has just spent its length drawing: levels of work, and how to tell one from another.

Nothing in the list suggests concealment. A man wrote down what he had read. It is the shelf as it stood in the room where the model was assembled, and this chapter is about the rest of it.[^ch10-opener]

Taken by itself, the list is a small piece of evidence about one man's reading. We have made more of the omission than he did. For as long as most of us have taught the model, we have presented its six components (our grouping, not Beer's "six vertical components of corporate cohesion")[^ch10-beer-six] as though they had a single origin, and we have described that origin to clients and students as a discovery rather than an arrangement. The bibliography shows in print that at least three of the people who found the same components were not on the shelf when the model was assembled. It shows nothing about whether they were right, and nothing about whether Beer would have written a different book had he read them. It shows only that the components have more than one address, and that we have been giving out one.

## The charge Part One leaves standing

Part One ended with a charge that drawing the model more carefully cannot answer: perhaps the six-part construction is only the vocabulary we inherited. A reader who has come through that hinge is owed an answer, and it cannot come from inside the vocabulary under suspicion.

The objection does not say that the Viable System Model↗[^ch10-glossary] is wrong. It says something narrower and more damaging: that the model was never chosen. It was inherited, from a community, from a reading list, from the centenary of a man's birth, and the inheritance has been mistaken for a finding. On this reading the nine chapters of Part One demonstrate nothing about the world. They demonstrate that a sufficiently general vocabulary can be applied to anything, which is a fact about vocabularies and not about organizations.

The charge is serious because it disputes no single chapter. It grants every one of them and asks the one thing none of them answers: what became of the alternatives? If nothing else was ever seriously considered, then the eight defects catalogued in Part One belong to a habit, not to a model that was tested against rivals and kept. A habit does not become a theory by being maintained for a long time.

Beer had an answer of his own, and a concession to go with it. He took the nervous system as his model because it was "a known-to-be-viable system", and held that another, "such as the amoeba", would have served: "The results are the same, as they must be if viability as such has its own laws." In *Decision and Control* he had named two rival sources, the ecological model behind laissez-faire and the neurological model behind the mandatory controller, and rejected both as "manifestly partial."[^ch10-beer-rivals] Twenty-three pages after the amoeba he wrote that "the choice of five (rather than some other number) is somewhat arbitrary." He set the standard for the whole construction in a sentence the tradition still quotes, "a model is neither true nor false: it is more or less useful", and tied it to a test the tradition does not quote: if a model's "predictive value is low, we had better start again."[^ch10-beer-arbitrary] None of this meets Part One's charge. An argument from the laws of viability, made by the author of the model, is no independent check on it, and the prediction test Beer named was never run. The outside evidence has to do the work his own argument could not.

Something did become of the alternatives. Several literatures reached pieces of this construction independently, by other doors, and one or two reached them earlier. This chapter owes that accounting: what else is on the shelf, on what evidence each neighbor stands, and what each costs the claim we have been making. Part One has already conceded the deepest of our failures, and nothing here retracts that concession. The inventory is where a defense we can actually support has to start.

The register of this part depends on two things. None of these literatures is an opponent. They are people who found something real without the help of our vocabulary, and the right response is credit, stated at the grain the evidence supports. And the omission on Beer's shelf is far less interesting than the omission on ours. He had a reading list and a deadline. We have had decades, indexes, search engines and a professional literature, and for most of that time, speaking for ourselves, we did not read these neighbors either.

## How the neighborhood is sorted

We sort the neighborhood by what each neighbor shares with Beer, not by discipline. Sorted that way, it falls into six groups. Five arrive at a component of the model on their own evidence, and these are less a threat than an argument. The nesting group supplies two components, nesting and criteria for a level, so five sharing groups yield six components. The sixth denies that the job the model sets itself is well-posed at all, and that group is the most likely to appear in a review. The admission rule, and what it costs, is set out in the note near the end of this chapter. Most of the groups below have a chapter of their own after this one; here they get the length an inventory should give them.

Independence, in the entries that follow, means something specific and modest: the earliest statements we could check do not cite Beer or Ashby, and no later acknowledgment of either has come to light. Absence of citation is weaker than proof of ignorance. A researcher can read without citing, and a tradition can carry an idea a long way without naming where it came from. Each finding of independence below is therefore defeasible, and the receipt at the end of this chapter prices the work that would firm those findings up or overturn some of them.

### Variety matching: Galbraith

Jay Galbraith found that organizational structure is determined by the information that must be processed under uncertainty, so that a designer has two families of moves: reduce the need for information, or increase the capacity to process it. He reached the finding through fieldwork in 1968, set it out in a book in 1973, and compressed it into a paper in 1974. The evidence base is one longitudinal study inside Boeing's commercial-airplane division. That is a narrow base for a general claim, and Galbraith did not present it as wider than it was.

For this chapter the citation record matters. The 1974 paper and the 1968 working paper cite the Carnegie tradition, and neither Ashby nor Beer.

The cost to us is precise. Variety attenuation and amplification were sufficient to describe the problem Galbraith describes, but they were not necessary to see it, because he saw it without them.[^ch10-galbraith] Chapter 12, The Variety Logic, takes the case in full, including what our search covered and what it does not.

### Nesting and levels: Simon, Koestler, Jaques

Three routes reached the nesting. Herbert Simon's near-decomposable hierarchy, in 1962, is a formal argument with a parable attached. Arthur Koestler's holon, coined in 1967 and given its propositions in the same book, is a proposition set. Elliott Jaques's time-span of discretion, developed in a program of organizational measurement from 1956 to 1989, is the only one of the three that tried to make a level measurable rather than asserted. Beer's 1981 Select Bibliography names none of them.

The three evidence bases are unlike one another, and they should be kept apart rather than added together: a formal argument and a parable, a set of propositions, and a measurement program with a mixed empirical record. The cost to us is correspondingly specific. Three independent routes found that wholes nest inside wholes, a strong result for nesting in general and one we did not need to supply. None of them delivered a ruler the others accept, so the nesting is well attested and the criterion for a level is not.[^ch10-simon][^ch10-koestler][^ch10-jaques] Chapter 11, The Nesting: Simon, Koestler, Jaques, follows each route and the state of its evidence separately.

The comparison needs Beer's own criterion, stated once, because every neighbor in this part is measured against it. Beer's recursion is a relation between viable systems, and viable means "able to maintain a separate existence." His theorem reads: "In a recursive organizational structure, any viable system contains, and is contained in, a viable system."[^ch10-beer-recursion] He fenced the term on three sides. Recursion is not "any kind of system contained inside another" but "an absolutely precise definition of viability." It is not organizational level: "If the term *recursion* is allowed to become a meaningless appendage to the term *organizational level*, then we do not have a valid model of a viable system." And the managerial parts do not count, since "Systems Two, Three, Four, and Five are not in themselves viable systems at any level of recursion." A layer of management is therefore not a recursion. Beer added that the levels "answer to mathematical criteria", and discussed them "in words and in diagrams" because his was "not a book for mathematicians."[^ch10-beer-fence]

Set against that criterion, the three routes reach objects that are related to Beer's and not the same. Simon's blocks are defined by rates of interaction and need not be viable. Koestler's holon is any part that is also a whole, an organ as much as a firm. Jaques's strata are layers of managerial roles, the reading of level that Beer ruled out. Where their boundaries coincide with his, the accurate description is convergence on a related but not identical object, and whether they coincide is an empirical question that Chapter 11 prices. Koestler's holon later went into factories as holonic manufacturing, whose production record Chapter 15 examines.[^ch10-holonic]

### Cross-scale coupling: Holling and Gunderson

The panarchy literature describes adaptive cycles coupled across scales through two named and unlike connections, *revolt* and *remember*. Holling's resilience work begins in 1973; the cross-scale account is developed in 2001 and 2002. The dates shape the claim. This group arrives after Beer, so it is no evidence of priority over him. It is evidence of independent rediscovery, a weaker result and still a real one, and the reason the priced receipt at the end of this chapter treats convergence and priority as different findings. The evidence base is an ecological synthesis together with a later review literature.

It costs us a component we draw only in part. Beer's recursion does couple levels in more than one way. Planning, which he called the "glue" of organizational cohesion, runs between recursions; algedonic signals, "alone of all the signals in the enterprise, are organized to break through recursivity"; and System Two and Three-Star "require cross-recursion coordination of a delicate kind."[^ch10-beer-coupling] None of these couplings depends on phase, and none of them corresponds to *remember*.[^ch10-panarchy] Chapter 11 takes the panarchy record as its fourth entry, and Chapter 23 carries the coupling claim forward.

### Bounded autonomy: Trist, Herbst, Endenburg

Between 1951 and 1976 the sociotechnical tradition produced semi-autonomous groups and the principle of minimum critical specification. Gerard Endenburg's double-linked circles descend through Kees Boeke's consent principle of 1945. Neither lineage establishes a dependency on Beer, in either direction, and neither claims one. The evidence base is field studies, design principles, and an adoption record that neither community has adequately counted. The missing count is as much our failure as theirs, and we are in a position to remedy it.

This is the most practical entry in the inventory. Beer supplied a rule and mechanisms of his own. The metasystem "should make only that degree of intervention that is required to maintain cohesiveness", and the resource bargain, an audit run "by agreement with System One management" and the sorting of constraints into mandatory and anti-oscillatory are the means.[^ch10-beer-autonomy] What we most often supply in practice is a placement: a named box on a diagram. What neither Beer nor we supply is a way to compute the rule, or evidence that the mechanisms bound autonomy as intended. These traditions supply design rules and working mechanisms at that point.[^ch10-sociotechnical][^ch10-endenburg] Chapter 13, The Autonomy Design, sets the rules and mechanisms beside what we offer.

### Multi-level viability: Ostrom

Elinor Ostrom's eight design principles describe institutions that persist across generations under conditions theory predicted would destroy them. The coding instrument appeared in 1989 and the synthesis in 1990, and the convergence is independent of Beer though not of Ashby: the Ostroms' multiple levels of rules come, largely through Vincent Ostrom, from Ashby's hierarchy of feedback mechanisms (Chapter 14). Her eighth principle nests enterprises in layers without requiring each layer to be viable, so what she reached is nested governance: related to Beer's recursion and not the same claim. Chapter 14 draws the consequence. The evidence base is a coded case corpus with a published instrument, and the instrument deserves our attention. Because it was published, other researchers could code further cases, disagree in a form that could be checked, and record cases where the principles did not hold.

That is where the cost falls, and it falls on us rather than on the model. This literature measured failures and published disconfirmations. Ours did not.[^ch10-ostrom] Chapter 14, The Evidence: Ostrom, takes the comparison seriously, including what a comparable instrument for our own claims would have to look like.

### The group that denies the job is well-posed

This group will actually appear in a review, and we have engaged it least. Niklas Luhmann's social systems are operationally closed and constituted by communication, which makes an organization something that cannot be steered from outside at all.[^ch10-luhmann] Assemblage theory, in Manuel DeLanda's rendering, insists that a component can be detached from one whole and plugged into another without ceasing to be itself.[^ch10-delanda] Actor-network theory abolishes levels entirely.[^ch10-ant] Werner Ulrich's critical heuristics asks, of any systems study, whose boundary judgments are in force and who is affected but not consulted.[^ch10-ulrich] Michael Jackson's classification places the VSM among the functionalist methodologies, useful where interests are unitary and untrustworthy where they are coercive.[^ch10-jackson] Hayek's argument that no regulator can assemble the knowledge dispersed among the regulated is older than most of it, and Chapter 16 finds one layer of it, the measure problem, still unanswered.[^ch10-hayek] And process philosophy, from Whitehead to Robert Cooper, holds that an entity is an artifact of a description that moves more slowly than the thing described.[^ch10-process] Chapter 16, The Objections That Land, states each objection at the strength its author would want and asks which of them land.

## What the convergence establishes

The five sharing groups answer the charge of arbitrariness, and the form of that answer is easy to overstate. It is not that the VSM is confirmed by the company it keeps; resembling its neighbors makes no construction true. It is that every one of its components has been reached independently by someone who had not read Beer: variety matching by Galbraith, nesting by Simon and Koestler, criteria for managerial strata by Jaques, cross-scale coupling by Holling, bounded autonomy by Trist, nested governance by Ostrom. Four of those entries are convergence on a related but not identical object. The nesting of Simon and Koestler, Jaques's strata and Ostrom's nested enterprises do not require each nested layer to be a viable system, as Beer's recursion does, and Holling's coupling carries a phase that Beer's couplings lack. A construction assembled from parts that six other literatures found, or found versions of, on their own may still be wrong. It is not arbitrary.

Inside that answer sits an asymmetry we have not always observed. Where a neighbor arrived earlier, the accurate word is priority, and priority belongs to them. Where a neighbor arrived later by an unrelated route, the accurate word is convergence, which strengthens the case for the component and settles nothing about who found it first. We have at times used the second sort of evidence as though it were the first, and the panarchy entry is where the temptation is strongest.

The answer has a price, and Part Two hands it to Part Four. If each component has an independent address, we cannot say that Beer discovered the components, and we have said it, in teaching and in print, for a long time. What survives is narrower: these components were assembled into one construction, with couplings specified between them, and nothing else on this shelf provides that assembly. It is an assembly claim, not a discovery claim. Chapter 23, The Narrow Claim, states it precisely and tests whether it holds. This chapter makes the credit real and hands the narrowed claim forward without inflating it on the way.

:::: {.how-we-looked}
**How we looked**

We assembled the neighborhood above by asking what each candidate shares with Beer rather than which discipline it belongs to, which is why the groups cut across management studies, ecology, sociology, philosophy and economics. A candidate was admitted if it makes a claim about the same object, the persistence of a bounded system under complexity it cannot absorb, and left out if it only shares vocabulary. That rule is ours, and it is contestable: it admits Friston[^ch10-friston] and Ostrom and excludes most of the strategy literature.

Two disclosures. We are inside the community this chapter is about, and we are not neutral about the outcome. And this is a shelf we can reach. The works here are in English or in translation into it, and a literature that never crossed over would be invisible from where we stand, which is exactly how a literature that does not exist would look. The Select Bibliography check was repeated on 23 September 2026 against a copy of the 1981 edition; the cross-literature checks were last repeated on 12 August 2026.
::::

## The priced receipt

**Receipt — completeness and priority audit of the six components.** Commission an audit that identifies, for each component, the earliest independent statement, and checks the citations in both directions. One researcher, six bounded searches, a preregistered evidence table, and external spot-checking: approximately fifteen working days. The evidence table records for each entry whether the object reached is Beer's (a viable system within a viable system) or a related one; Simon and Koestler, Jaques, Ostrom and Holling are entered as related, so a negative result on them narrows the claim about nesting, not about recursion. A negative result — that one or more components cannot be shown to have an independent lineage — would narrow the non-arbitrariness claim to the components that survive, and we would print the narrowed version. Holling is already the warning: convergence after Beer is evidence of independent rediscovery, not priority.

[^ch10-galbraith]: Jay R. Galbraith, *Designing Complex Organizations* (Addison-Wesley, 1973), and "Organization Design: An Information Processing View," *Interfaces* 4, no. 3 (1974): 28–36. The airplane division is *Environmental and Technological Determinants of Organization Design: A Case Study*, Working Paper 352-68 (Sloan School of Management, MIT, 1968). The working paper names the site as "the Commercial Airplane Division of the Boeing Company" in its abstract, as "a longitudinal case study". That he arrived at it without Ashby is a negative finding; Chapter 12 states what was searched and what the finding does not cover. For the 1973 book, a search of the full-text index of the Internet Archive's scan (29 September 2026) finds neither Ashby nor Beer, nor Wiener, Shannon or *cybernetic*; the reference list has not been read page by page.

[^ch10-friston]: Karl Friston, "The Free-Energy Principle: A Unified Brain Theory?," *Nature Reviews Neuroscience* 11, no. 2 (2010): 127–38, https://doi.org/10.1038/nrn2787; Karl Friston, Thomas FitzGerald, Francesco Rigoli, Philipp Schwartenbeck, and Giovanni Pezzulo, "Active Inference: A Process Theory," *Neural Computation* 29, no. 1 (2017): 1–49; Thomas Parr, Giovanni Pezzulo, and Karl J. Friston, *Active Inference: The Free Energy Principle in Mind, Brain, and Behavior* (MIT Press, 2022). The technical criticism of the principle, which is substantial, is in Chapter 12.

[^ch10-simon]: Herbert A. Simon, "The Architecture of Complexity," *Proceedings of the American Philosophical Society* 106, no. 6 (1962): 467–82. Simon's own retraction of the watchmakers' arithmetic two paragraphs after he gives it, and the near-decomposability definition that does the theoretical work, are quoted in Chapter 11, The Nesting.

[^ch10-koestler]: Arthur Koestler, *The Ghost in the Machine* (Hutchinson, 1967), 48, where the holon is coined, and Appendix I, "General Properties of Open Hierarchical Systems (O.H.S.)," 341–48, where the propositions first appear. The propositions usually quoted come from Koestler's 1969 restatement in the Alpbach appendix, "Some General Properties of Self-Regulating Open Hierarchic Order"; Chapter 11 records which wording we were able to verify and which we were not.

[^ch10-holonic]: Hendrik Van Brussel, Jo Wyns, Paul Valckenaers, Luc Bongaerts, and Patrick Peeters, "Reference Architecture for Holonic Manufacturing Systems: PROSA," *Computers in Industry* 37, no. 3 (1998): 255–74, for the reference architecture. The count of systems that have run in day-to-day production is ours and it is a count of a negative; the three sources it rests on, and what would refute it, are given in Chapter 15.

[^ch10-jaques]: Elliott Jaques, *Measurement of Responsibility: A Study of Work, Payment, and Individual Capacity* (Tavistock Publications, 1956), and *Requisite Organization: The CEO's Guide to Creative Structure and Leadership* (Cason Hall, 1989). The empirical record is mixed and is set out in Chapter 11, including the replications that do not support the criterion.

[^ch10-panarchy]: Lance H. Gunderson and C. S. Holling, eds., *Panarchy: Understanding Transformations in Human and Natural Systems* (Island Press, 2002); C. S. Holling, "Resilience and Stability of Ecological Systems," *Annual Review of Ecology and Systematics* 4 (1973): 1–23, https://doi.org/10.1146/annurev.es.04.110173.000245, where the resilience work begins; C. S. Holling, "Understanding the Complexity of Economic, Ecological, and Social Systems," *Ecosystems* 4, no. 5 (2001): 390–405. How much of the panarchy literature actually uses the cross-scale couplings, rather than the adaptive cycle alone, is quantified in Chapter 11.

[^ch10-sociotechnical]: E. L. Trist and K. W. Bamforth, "Some Social and Psychological Consequences of the Longwall Method of Coal-Getting," *Human Relations* 4, no. 1 (1951): 3–38; F. E. Emery and E. L. Trist, "The Causal Texture of Organizational Environments," *Human Relations* 18, no. 1 (1965): 21–32; P. G. Herbst, *Socio-Technical Design: Strategies in Multidisciplinary Research* (Tavistock Publications, 1974), for "Designing with Minimal Critical Specifications"; Albert Cherns, "The Principles of Sociotechnical Design," *Human Relations* 29, no. 8 (1976): 783–92. Attributing minimum critical specification to the tradition is right at this grain; Chapter 13, The Autonomy Design, narrows it to Herbst and records Cherns's own disclaimer of originality.

[^ch10-endenburg]: Gerard Endenburg, *Sociocracy: The Organization of Decision-Making; "No Objection" as the Principle of Sociocracy* (Eburon, 1998); Kees Boeke, "Sociocracy: Democracy as It Might Be" (May 1945), English text at https://www.sociocracy.info/sociocracy-democracy-kees-boeke/, for the consent principle (original place of publication not identified); A. Georges L. Romme and Gerard Endenburg, "Construction Principles and Design Rules in the Case of Circular Design," *Organization Science* 17, no. 2 (2006): 287–97. Chapter 13 reports that neither the sociocratic movement nor this community publishes an adoption count that would allow a comparison, so both sides of it are unavailable.

[^ch10-ostrom]: Elinor Ostrom, *Governing the Commons: The Evolution of Institutions for Collective Action* (Cambridge University Press, 1990), the eight design principles at 90; Elinor Ostrom, Arun Agrawal, William Blomquist, Edella Schlager, and S. Y. Tang, *Common-Pool Resources (CPR) Coding Manual* (Workshop in Political Theory and Policy Analysis, Indiana University, 1989), for the instrument. The reviews that test the principles against coded cases, and what they found, are in Chapter 14.

[^ch10-luhmann]: Niklas Luhmann, *Social Systems*, trans. John Bednarz Jr. with Dirk Baecker (Stanford University Press, 1995), and *Organization and Decision*, ed. Dirk Baecker, trans. Rhodes Barrett (Cambridge University Press, 2018). Chapter 16, The Objections That Land, states the objection at the strength Luhmann would want and says what survives it.

[^ch10-delanda]: Manuel DeLanda, *A New Philosophy of Society: Assemblage Theory and Social Complexity* (Continuum, 2006), 10–11, for relations of exteriority. Chapter 16 takes the argument up, together with Beer's own statements on detachment.

[^ch10-ant]: Bruno Latour, *Reassembling the Social: An Introduction to Actor-Network-Theory* (Oxford University Press, 2005), Part II, "How to Keep the Social Flat," from 165, with "keep the social domain completely flat" at 171 and "the Flat-Earthers of social theory" at 172. The application to Cyberstride and the operations room — which is ours and not Latour's — is in Chapter 16.

[^ch10-ulrich]: Werner Ulrich, *Critical Heuristics of Social Planning: A New Approach to Practical Philosophy* (Paul Haupt, 1983), and "Critical Heuristics of Social Systems Design," *European Journal of Operational Research* 31, no. 3 (1987): 276–83. His 1981 critique of the Chilean project, "A Critique of Pure Cybernetic Reason: The Chilean Experience with Cybernetics," *Journal of Applied Systems Analysis* 8 (1981): 33–59, is read from the source in Chapter 16, and it is the hardest thing in this book to answer.

[^ch10-jackson]: Michael C. Jackson and Paul Keys, "Towards a System of Systems Methodologies," *Journal of the Operational Research Society* 35, no. 6 (1984): 473–86; Michael C. Jackson, *Systems Thinking: Creative Holism for Managers* (John Wiley & Sons, 2003). What the 1984 grid does and does not classify — coercive contexts are named there and deliberately left outside it — is set out in Chapter 16.

[^ch10-hayek]: F. A. Hayek, "The Use of Knowledge in Society," *American Economic Review* 35, no. 4 (1945): 519–30. Chapter 16 separates the objection into layers and says which of them the model can meet and which it cannot.

[^ch10-process]: Alfred North Whitehead, *Process and Reality: An Essay in Cosmology* (Macmillan, 1929), for the position, and *Science and the Modern World* (Macmillan, 1925), where the fallacy of misplaced concreteness is named. The management-studies form, Robert Cooper's among it, is cited in Chapter 16, whose note on Tsoukas and Chia records that they are not Whiteheadians.

[^ch10-opener]: Stafford Beer, *Brain of the Firm*, 2nd ed. (John Wiley & Sons, 1981), Select Bibliography, 405–10, with the Explicit References at 410; Beer's reason for the cookery book is given in the Preface to the Second Edition, xii. The cookery book is Jocasta Innes, *The Pauper's Cookbook* (Penguin Books, 1971). The dates are Herbert A. Simon, "The Architecture of Complexity," *Proceedings of the American Philosophical Society* 106, no. 6 (1962): 467–82, and Arthur Koestler, *The Ghost in the Machine* (Hutchinson, 1967). This is a check of a printed list and is as reliable as such a check can be; it is not a claim that Beer nowhere mentions these men in his running text, although a search of the scanned text of this edition (23 September 2026) found no mention of any of the three, nor of the holon, and the separate list of references for Part Four (397) does not name them either. Chapter 11, The Nesting, uses the same list as evidence and does not repeat the description.

[^ch10-glossary]: Terms marked ↗ are defined in the Glossary at the back of the book. The mark is set on the first appearance of each term in each chapter, and not afterwards.

[^ch10-beer-six]: Stafford Beer, *The Heart of Enterprise* (John Wiley & Sons, 1979), 217, the First Axiom of Management, which balances horizontal variety against "the six vertical components of corporate cohesion." The six components of this chapter are the authors' grouping of design ideas, not Beer's.

[^ch10-beer-rivals]: Stafford Beer, *Decision and Control* (John Wiley & Sons, 1966), 374–75, paginated after the Stafford Beer Classic Library reprint (John Wiley & Sons, 1994): "the doctrinaire laissez-faire man is backing a model of the ecological system, and the doctrinaire controller is adopting a model from the neurological system"; "we want neither of these models; because they are manifestly partial." Beer's alternative is "a model taken from an organic system": "Only a complex creature, such as a man himself, provides an adequate model" (375).

[^ch10-beer-arbitrary]: Stafford Beer, *Brain of the Firm*, 2nd ed. (John Wiley & Sons, 1981), 75 ("a known-to-be-viable system"; the amoeba) and 98 ("somewhat arbitrary"); Stafford Beer, *Diagnosing the System for Organizations* (John Wiley & Sons, 1985), 2 ("So remember: / a model is neither true nor false: / it is more or less useful."); Stafford Beer, *Decision and Control* (John Wiley & Sons, 1966), 113, where the measure of utility "may be accomplished by straightforward testing of the predictive value of the scientific model. If this predictive value is low, we had better start again." *Decision and Control* is paginated after the Stafford Beer Classic Library reprint (John Wiley & Sons, 1994), which reproduces the 1966 text.

[^ch10-beer-recursion]: Beer, *Diagnosing the System*, 1, where "viable" is defined with the *Oxford English Dictionary*; Beer, *Heart of Enterprise*, 118, for the Recursive System Theorem, restated at 308 and in the glossary of rules at 566.

[^ch10-beer-fence]: Beer, *Diagnosing the System*, 2; Beer, *Heart of Enterprise*, 308 (organizational level, "mathematical criteria", "since this is not a book for mathematicians, we have to discuss the issues in words and in diagrams") and 411 (Systems Two to Five). Chapters 11 and 14 compare their neighbors with this criterion and do not restate it.

[^ch10-beer-coupling]: Beer, *Heart of Enterprise*, 336, for planning as the "glue" of organizational cohesion, and 407, for the algedonic signal; Beer, *Diagnosing the System*, 135.

[^ch10-beer-autonomy]: Beer, *Heart of Enterprise*, 158, for the rule, and 180–81, for the red and green paper; Beer, *Diagnosing the System*, 38–39, for the resource bargain, and 82, for the audit.