---
title: "Ch04 channels nobody specified content"
description: "In seminars and in print, we describe the algedonic↗ channel as the one that escapes ordinary hierarchy. Beer says the same: algedonic signals are “organized…"
url: "https://wiki.cybernetic-futures.org/punch-and-judy/part-01-inside-the-model/ch04-channels-nobody-specified/ch04-channels-nobody-specified-content"
modified: "2026-11-12T15:26:25.883Z"
language: "en"
site: "Cybernetic Futures"
---

# Channels Without Owners

In seminars and in print, we describe the algedonic↗ channel as the one that escapes ordinary hierarchy. Beer says the same: algedonic signals are "organized to *break through* recursivity."[^ch04-beer-break] He also says how a signal qualifies. A statistical filter watches chosen points of stability, splits off the readings that show incipient instability, and decides whether to alert System Five. His phrase for it is "a device, operating in real-time, that monitors chosen points of stability by statistical methods," and the metasystem sets its criteria.[^ch04-beer-filter] What Beer does not say is who chooses the points of stability, who may contest the filter, and who audits it, on whose behalf. Those are the questions an institution has to answer, and neither Beer nor the practice that followed him answered them. We draw this channel as a mechanism and have never specified it as an institution, and a channel without those answers has no owner anyone could name.

Beer counts six vertical channels between the operating units and their metasystem, only one of which, he complains, is usually acknowledged,[^ch04-beer-six] and he sets the algedonic channel beside them, "orthogonal to the calming system."[^ch04-beer-orth] Three of these carry this chapter: the resource bargain, audit, and the algedonic channel.

Four neighboring literatures approach the gap from outside the model. None of them was written against the VSM, and only one was written after it. Each says something specific about the relationship between a level and the level containing it: what can cross, who is entitled to send it, whose interests the crossing serves, and how large the asymmetry between the two levels may become. Taken in order they make one argument, and the last of them, about variety, is the reason the first three are more than stylistic complaints.

## Revolt, remember, and the coupling we never drew

Holling's adaptive cycle is the one member of this group that comes after Beer rather than before him: resilience in 1973, the cycle itself in 1986, panarchy proper at the turn of the century. Systems traverse exploitation, conservation, release and reorganization. The cycles nest across scales that differ in speed and extent, and the levels are coupled in both directions. In *revolt*, a fast small collapse cascades upward into a slower larger level. In *remember*, the slow large level supplies accumulated structure downward to reconstitute a small one after its collapse.[^ch04-holling]

Beer's channels have no counterpart to either. His vertical channels command, bargain, coordinate, inspect and alarm, and revolt and remember are contagion and inheritance. The algedonic signal is arguably a degenerate revolt, a pain report that bypasses levels. Remember has no counterpart at all, so the model has no account of how a destroyed unit is reconstituted out of the memory of its container.

The word *degenerate* is doing work there, and it needs care. A revolt in Holling's sense is addressed to no one. It propagates because the levels are materially coupled, and the larger level receives it whether or not it has a receiver. Our algedonic signal is addressed. In Beer's primary design nobody writes it, since a filter generates it, but it has a recipient and an obligation to respond: if the top did not respond, he says, the system would "*lose its cohesiveness*."[^ch04-beer-cohesion] Beer also has authored forms, the informal telephone call to the boss and the Chilean algedonic meter, on which a person "does not have to explain anything."[^ch04-beer-meter] The addressing makes the channel governable. It also makes the specification questions unavoidable, because an authored signal can be withheld or forged, and a filter's criterion can be set by whoever owns it. Remember is harder still. The containing system would have to hold structure on behalf of a unit that no longer exists, and be able to reissue it into a rebuilt unit with no continuity of personnel with the old one: the operating model, the local knowledge, the relationships, the standing to act.

We do not claim the model forbids this. System Four's model of the environment and System Five's account of identity might in principle be where such structure is held, and practitioners plainly do reconstitute failed units in the field. The claim is narrower and, we think, harder to dismiss: nothing in the canonical description tells us where reconstitution capacity lives, who is responsible for maintaining it, or how one would know it had decayed. Cases would settle it, where argument cannot. Take a set of documented instances in which an operational unit was destroyed and later rebuilt inside the same containing system. Ask what supplied the structure, whether the container had retained it deliberately, and whether anyone had ever been accountable for retaining it. Chapter 5 takes up the related question of what fixes the levels in the first place.

## Whose audit is it

Ostrom arrives at the same boundary from the other side, and with evidence.

Read her eight design principles for long-enduring common-pool resource institutions with the VSM in hand. Clearly defined boundaries. Congruence between the appropriation and provision rules and local conditions. Collective-choice arrangements, in which most of those affected by the operational rules can participate in changing them. Monitoring by monitors who "are accountable to the appropriators or are the appropriators". Graduated sanctions. Rapid, low-cost, local conflict resolution. Minimal recognition by external authorities of the right to organize. And, for resources embedded in larger systems, nested enterprises, with appropriation, provision, monitoring, enforcement, conflict resolution and governance "organized in multiple layers".[^ch04-dps]

The fourth principle is the decisive one here. It resembles the audit channel (the System Three-star of *Diagnosing the System*) and goes further than Beer did. Beer's audit is sporadic, and it reaches the operations "by agreement with System One management":[^ch04-beer-audit] a consent condition, held by the unit's managers. Ostrom's condition is held by the people whose conduct and resources are being inspected. Its monitors must be accountable to them, or be drawn from them. A channel's independence is a property of who owns it, not of the box it bypasses. The seventh principle, and Chile in September 1973, belong to the Cybersyn argument in Chapter 8. The eighth, nested enterprises, is the ladder question and belongs to Chapter 5. We note both handoffs and reopen neither.

The fourth principle transfers directly to the algedonic channel, which is why it appears here and not only in Chapter 14. If a Three-star audit is not independent simply because it routes around the line, then an algedonic signal is not a report from the affected simply because it routes around the line either. In the implementations we have seen described (we have not surveyed them systematically), the entitled sender has usually been a manager, reporting on the condition of the unit he manages. Beer's own design points the same way. The filter that decides whether pain reaches the top is designed by the metasystem,[^ch04-beer-owner] the party the signal is meant to wake. Ostrom's fourth principle exists to prevent exactly that structure. We adopted the channel with the ownership Beer gave it, and did not ask whether it was the right owner.

This evidence has a limit. Ostrom's principles were induced from common-pool resource institutions, which have appropriators with standing claims and no command channel. Firms and public agencies have both a command channel and employees who can be dismissed. Whether the ownership requirement survives that difference, or has to be rebuilt in some other form, is an open question. Chapter 14 sets out what a comparative study would have to look like.

## Who is at the boundary

Critical systems heuristics asks twelve boundary questions, each in an *is* and an *ought* mode, about who is served, who decides, what counts as knowledge and what legitimates a design; Chapter 16 sets them out in full.[^ch04-csh] Three bear on this chapter: the measure of improvement, the guarantor of success, and the tenth, which asks who is witness to the interests of those affected but not involved.

Applied to us, the questions are uncomfortable in the right way. Who names the identity in the step Beer left to the eye of the beholder?[^ch04-beer-beholder] Who fixes the recursion levels? Whose measure of improvement do the System One indices encode? And what is the assumed guarantor of success: viability itself, or the model's own theorems? Both are internal to the designer's frame, and so they can justify nothing to anyone excluded from it.

The tenth question is the one that bears on the algedonic channel. The channel has no seat for the person who is affected and was never a member. In Cybersyn, as Chapter 8 shows, an algedonic signal was more than a cry of pain; it was a transfer of control upward. The architectural consequence is that the person who bears the pain may neither author the signal nor control what the escalation does. Beer's algedonic meter is the one design in which that person authors it, and it stayed outside the architecture as we teach it.

This is Ostrom's fourth principle again, in a wider frame. Ostrom asks whether the monitor is drawn from those monitored; Ulrich asks whether anyone at all speaks for those outside the membership boundary. A specification that answered the first without the second would still leave the channel closed to the neighbor, the patient, the downstream community.

Nothing in critical systems heuristics is incompatible with the Viable System Model. The two operate at different logical levels: Ulrich interrogates the boundary, Beer designs inside it. In intent, System Five and the algedonic channel are a witness function. The model has the slot, and Beer says who should fill it: System Five "consists of many people," possibly "simply workers," in Chile *el pueblo*, and later all stakeholders.[^ch04-beer-s5] He gives no procedure for seating them, and nothing for those who are affected but outside the boundary. That defect can be fixed without refuting anything, and fixing it is the cheapest high-value reform available to us. Chapter 16 takes Ulrich's objections in full, including those that do not resolve so comfortably.

## A regulator with no stated upper bound

The three findings so far concern who may use a channel. The fourth concerns how far apart the two ends of a channel are allowed to drift.

Ashby's law states a minimum variety for a regulator.[^ch04-ashby] Nothing in the sources we searched for this chapter, Ashby's *Introduction to Cybernetics* and five of Beer's books, states a maximum compatible with that regulator remaining auditable by the regulated.[^ch04-nomax]

Beer comes closest. The metasystem "should make only that degree of intervention that is required to maintain cohesiveness," and freedom is "in principle a computable function of systemic purpose as perceived."[^ch04-beer-min] Metasystemic effort beyond what cohesion needs is, in his diagnosis, a pathology.[^ch04-beer-estab] That is a ceiling on intervention, set by cohesion and purpose. It is not a ceiling on the variety the regulator may hold, and Beer gives no procedure by which the regulated could check either one.

We flag this as a scoped negative finding. It describes those texts and is not the result of a systematic literature search; a single citation we have missed would overturn it, and we would welcome one. Chapter 3 sets out what Ashby's law does and does not say. The absence matters here because our other three findings all assume that the regulated can, in principle, check the channel that speaks about them. A regulator whose variety may grow without stated bound can satisfy the law and still be unauditable in practice. The ownership requirement and the witness seat then become formalities. If there is an upper bound, we should be able to say what it is a function of.

The transducer, the mechanism that decides which differences survive passage through a channel, is the part of the variety logic its rediscoverers in Chapter 12 left alone. Part Four returns to it, in Chapter 25. A specification of who may send an algedonic signal is incomplete without an account of what the channel does to the signal in transit. Beer states the requirement: "the variety of the transducer must be at least equivalent to the variety of the channel," and channel capacity and transduction capacity are two different things.[^ch04-beer-trans] He gives no way to measure either at a real boundary.

## What it would cost to find out

**Receipt — red-team the algedonic channel.** The reform this chapter argues for begins with a document we have not written. Write a qualifying-signal specification naming entitled senders, thresholds, evidence, routing, response time, appeal and audit owner. Then give a red team the same operating data and ask it to manufacture a signal that passes while misrepresenting the condition. Price: one facilitated design session and one red-team day. If false signals pass trivially, the negative result is publishable: the channel transfers control without a defensible rule for when transfer is warranted. Chapter 9 explains why so few tests of this kind exist in our literature, and the explanation is not that they are expensive.

[^ch04-holling]: C. S. Holling, "Resilience and Stability of Ecological Systems," *Annual Review of Ecology and Systematics* 4 (1973): 1–23; C. S. Holling, "The Resilience of Terrestrial Ecosystems: Local Surprise and Global Change," in *Sustainable Development of the Biosphere*, ed. William C. Clark and R. E. Munn (Cambridge University Press, 1986), 292–317, where the adaptive cycle first appears; Lance H. Gunderson and C. S. Holling, eds., *Panarchy: Understanding Transformations in Human and Natural Systems* (Island Press, 2002), and C. S. Holling, "Understanding the Complexity of Economic, Ecological, and Social Systems," *Ecosystems* 4, no. 5 (2001): 390–405, for revolt and remember.

[^ch04-dps]: Elinor Ostrom, *Governing the Commons: The Evolution of Institutions for Collective Action* (Cambridge University Press, 1990), 90 (table 3.1). Principle 4 reads in full: "Monitors, who actively audit CPR conditions and appropriator behavior, are accountable to the appropriators or are the appropriators." Principle 8 reads: "Appropriation, provision, monitoring, enforcement, conflict resolution, and governance activities are organized in multiple layers of nested enterprises," and Ostrom states it only "for CPRs that are parts of larger systems."

[^ch04-csh]: Werner Ulrich, *Critical Heuristics of Social Planning: A New Approach to Practical Philosophy* (Paul Haupt, 1983); Werner Ulrich and Martin Reynolds, "Critical Systems Heuristics," in *Systems Approaches to Managing Change: A Practical Guide*, ed. Martin Reynolds and Sue Holwell (Springer, 2010), 243–92. The ordering used here is Ulrich's own: the twelve questions fall in four groups of three — sources of motivation, of power, of knowledge, of legitimation — and each is asked in an *is* and an *ought* mode. The measure of improvement is question 3, the guarantor question 9, and the witness to those affected but not involved question 10, in the numbering of Ulrich's checklist of boundary questions as reproduced in Werner Ulrich, "A Brief Introduction to Critical Systems Heuristics (CSH)" (ECOSENSUS project, Open University, 2005), 11 (fig. 4), which he dates to 1987.

[^ch04-ashby]: W. Ross Ashby, *An Introduction to Cybernetics* (Chapman & Hall, 1956), 206–7 (sections 11/6–11/7): if the variety in the outcomes is to be reduced to an assigned number, "R's variety must be increased to at least the appropriate minimum." The one maximum Ashby states runs the other way. A regulator fixed in its channel capacity meets "an absolute limit to the amount of regulation (or control) that can be achieved by R" (245, section 13/2): a ceiling on what the regulator can do, not on how much variety it may hold. Chapter 3 sets out the law's conditions.

[^ch04-beer-break]: Stafford Beer, *The Heart of Enterprise* (John Wiley & Sons, 1979), 407, where algedonic signals, "alone of all the signals in the enterprise," break through the recursions.

[^ch04-beer-filter]: Beer, *Heart of Enterprise*, 386 for the quotation, and 275, 381–82; Stafford Beer, *Brain of the Firm*, 2nd ed. (John Wiley & Sons, 1981), 178–79: "Statistical criteria must be established to ensure that ascending algedonic information … is not simply absorbed"; Stafford Beer, *Diagnosing the System for Organizations* (John Wiley & Sons, 1985), 133, where the signal passes through "its own algedonic filter" on the way to System Five.

[^ch04-beer-six]: Beer, *Heart of Enterprise*, 142–43: "These six vertical loops (of which only one, the command loop, is customarily acknowledged …)"; Beer, *Diagnosing the System*, 81–82, for the six: corporate intervention, the resource bargain, operational linkages, environmental intersects, System Two, and audit; and 68 on System Two: "Its function is not to command, but to damp oscillations."

[^ch04-beer-orth]: Beer, *Heart of Enterprise*, 406.

[^ch04-beer-cohesion]: Beer, *Heart of Enterprise*, 407.

[^ch04-beer-meter]: Beer, *Diagnosing the System*, 133, for the informal route; Beer, *Brain of the Firm*, 283–84, for the meter.

[^ch04-beer-audit]: Beer, *Diagnosing the System*, 82. The label "3*" for the audit channel is from this book; in *The Heart of Enterprise* the same label names a model inside System Five (352).

[^ch04-beer-owner]: Beer, *Heart of Enterprise*, 386, where the limits of each point of stability are "determined by the criteria laid down by the metasystem"; and 382, where the resolution filter on a manager's input "has been designed by the metasystem of which our manager is a member." The resolution filter "does not know whether John or Mary is operating that machine today"; the manager's perceptual filter, by contrast, is designed by the manager himself.

[^ch04-beer-s5]: Beer, *Heart of Enterprise*, 263–64; Beer, *Brain of the Firm*, 258, and 288: "System Five within a people's government cannot be an élite ruling class"; Stafford Beer, *Beyond Dispute: The Invention of Team Syntegrity* (John Wiley & Sons, 1994), 149: "All stakeholders are represented in System Five."

[^ch04-nomax]: The texts searched were Ashby, *Introduction to Cybernetics*, read in full for this question, and five of Beer's books, searched in full text for any stated upper limit on a regulator's variety: *Decision and Control* (John Wiley & Sons, 1966); *Platform for Change* (John Wiley & Sons, 1975); *The Heart of Enterprise*; *Brain of the Firm*, 2nd ed.; and *Diagnosing the System*. The nearest statements are limits of capacity, not of auditability. Beer writes that the maximum variety a senior management "can handle for each subsidiary is their own total capacity divided by the number of subsidiaries in System One" (*Diagnosing the System*, 51), and he reports Bremermann's physical limit on the information-processing capacity of matter (*Brain of the Firm*, 48). The neighboring literatures of this chapter (Holling, Ostrom, Ulrich) were not searched for the question.

[^ch04-beer-min]: Beer, *Heart of Enterprise*, 157–58.

[^ch04-beer-estab]: Beer, *Heart of Enterprise*, 411–12, on the metasystem that maintains itself as "The Establishment." The Law of Cohesion (355) is stated as an equality, not as a minimum.

[^ch04-beer-beholder]: Beer, *Heart of Enterprise*, 9: "we simply cannot attribute purposes, or even boundaries, to systems as if these were objective facts of nature. The facts about the system are in the eye of the beholder."

[^ch04-beer-trans]: Beer, *Heart of Enterprise*, 101; Beer, *Diagnosing the System*, 47, where "channel capacity and transduction capacity … are not at all the same thing."