This theme covers four lines of work in cognitive science and philosophy of mind that v02 imports to say what kind of claim the VSM makes and where a system ends: distributed cognition (the boundary criterion behind Demonstration IIIb), the extended-mind debate (the constitutive boundary question of §2.4), predictive processing and the Markov-blanket dispute (the instrumental/constitutive binary that organises §2.2 and the theorem/hypothesis parallel of §2.3), and coordination dynamics (System Two as regime rather than structure). A VSM practitioner needs it because the paper’s central move—separating what the analyst lays over an organisation from what is claimed to be there—was made in these fields first, with named positions. I would take 5.3 before 5.1 and 5.2, because Bruineberg et al. supply the distinction every later section uses; then 5.2, 5.1, 5.4.
Prerequisites. T2.1 (research programmes, novel corroboration) and T1.2 (Beer’s recursion criterion: variety and operational closure); T3.6 (Simon) helps for 5.4.
Distributed cognition
Introduction. Distributed cognition is the claim, made by Edwin Hutchins on the basis of ethnography aboard a US Navy ship and in airline cockpits, that the unit of analysis for cognitive work is not the individual but a functional system of people, artefacts and procedures. Fixing a ship’s position is done by no one person; it is done by a chain of representational states—a bearing read from a pelorus, spoken, written, plotted—propagating and being transformed across media. The method that follows is cognitive ethnography: trace where representations go, and the boundary of the system is whatever the tracing encloses. Two results matter here. First, the properties of the functional system differ from those of its members: the cockpit “remembers” approach speeds through the layout of speed bugs and cards, not through any pilot’s memory. Second, the boundary is derived from observed propagation rather than declared in advance, which is why it can be compared against a boundary drawn by some other method. Hutchins also showed how such systems reorganise their division of labour under breakdown, with local adaptations that no one designed—a finding closer to organisation theory than to psychology.
Important authors. Edwin Hutchins, Professor Emeritus of Cognitive Science at the University of California, San Diego and a MacArthur Fellow, founded the line of work; James Hollan and David Kirsh, also at UCSD, developed it toward human–computer interaction. Karl Weick and Karlene Roberts brought the same cockpit setting into organisation studies under the heading of collective mind. The approach sits within the situated and embodied turn in cognitive science, against internalist models.
Importance for cybernetics and the VSM. Distributed cognition is the nearest thing outside cybernetics to a worked method for deciding what a system in focus is. The VSM asks the same question and answers it by variety and operational closure—does this unit produce the system’s task—but has no procedure for checking that answer against anything other than another analyst. Hutchins’s tracing method gives an independent criterion. The VSM literature has not engaged it, and Hutchins does not engage Beer; the only shared ancestry is Bateson and the cybernetic-era interest in systems of information flow.
Importance for the article. Demonstration IIIb (§8.1, §8.3) is built on it: VSM analysts partition one setting by the model’s criterion, cognitive ethnographers partition the same setting by propagation of representational state, blind to the VSM, and the partitions are compared by a preregistered similarity measure. Convergence would be a corroborated prediction of a kind the model has never produced (§8.4); localised divergence narrows the recursion criterion to what it tracks. The import is from Input from Cognitive Science §2, which argued that intra-paradigm reliability (IIIa) cannot show that a partition carves anything real, “because shared training produces shared artefacts.” A reviewer will press on two things: whether the two criteria are genuinely independent (Hutchins traces representations; Beer traces task production; both may track information flow), and whether a similarity measure between partitions can be fixed before either partition exists. Note that Hutchins (1995) is Tier C and daggered in v02; the co-author should have read the cockpit paper and the navigation chapters before review.
Sources in the reading list.
- the functional system as unit of analysis and the propagation-tracing method; read the navigation chapters and the chapter on the organisation of distributed work.
- the compact case: a two-pilot-plus-instruments system whose cognitive properties are not those of its members; the model for the IIIb comparison.
Other important sources and authors.
- Hutchins, E. (1991). Organizing work by adaptation. Organization Science, 2(1), 14–39 — how a navigation team reorganised its division of labour under a failure; the paper of Hutchins’s closest to the VSM’s subject matter.
- Hollan, J., Hutchins, E., & Kirsh, D. (2000). Distributed cognition: Toward a new foundation for human-computer interaction research. ACM Transactions on Computer-Human Interaction, 7(2), 174–196 — the programme stated as a method, with its unit of analysis and its ethnographic commitments.
- Weick, K. E., & Roberts, K. H. (1993). Collective mind in organizations: Heedful interrelating on flight decks. Administrative Science Quarterly, 38(3), 357–381 — the organisation-studies counterpart, on aircraft-carrier decks; useful when a reviewer asks whether organisation theory has its own version.
- Kirsh, D., & Maglio, P. (1994). On distinguishing epistemic from pragmatic action. Cognitive Science, 18(4), 513–549 — actions taken to change the world versus actions taken to change one’s own information state; relevant to what “propagation of representational state” counts.
- Rogers, Y., & Ellis, J. (1994). Distributed cognition: An alternative framework for analysing and explaining collaborative working. Journal of Information Technology, 9(2), 119–128 — an early statement of the framework for work analysis, with its limits.
The extended mind and the coupling–constitution debate
Introduction. Clark and Chalmers argued that when an external resource plays the functional role that internal memory would—Otto’s notebook standing in for Inga’s biological memory—it is part of the cognitive process, and that whether something is cognitive should not depend on its being inside the skull (the parity principle). The thesis provoked a long argument whose two standard objections transfer directly to the VSM. Adams and Aizawa named the coupling–constitution fallacy: from the fact that X is causally coupled to a cognitive process it does not follow that X is part of it, and without a “mark of the cognitive” the extended thesis cannot tell the two apart. Rupert distinguished extended cognition (external resources are constituents) from embedded cognition (cognition depends heavily on the environment without being constituted by it) and argued that the weaker hypothesis explains the same facts at lower cost, because external resources differ in fine-grained functional profile from internal ones and so do not form a useful kind with them. The debate’s lasting product is a set of criteria for constitution—parity, functional integration, mutual manipulability—each of which has been tested against cases.
Important authors. Andy Clark (University of Sussex, previously Edinburgh) and David Chalmers (New York University) wrote the 1998 paper; Clark’s Supersizing the Mind (2008) is the full defence. Frederick Adams (University of Delaware) and Kenneth Aizawa (Rutgers University–Newark) are the principal critics; Robert Rupert (University of Colorado Boulder) developed the systems-based, organism-centred alternative. Richard Menary edited the standard collection.
Importance for cybernetics and the VSM. The VSM’s recursion criterion—a unit is a System One if it produces the system’s task, under variety and operational closure—is a constitution criterion, and it is of exactly the kind Adams and Aizawa argue is too permissive: coupling is ubiquitous and closure can be drawn at many grains. The VSM has no mark of the systemic that would distinguish a supplier coupled to a firm from a supplier that is a System One of it. Cybernetics has discussed boundaries mainly through Ulrich’s critical systems heuristics, which asks a different question. The two questions have not been separated in the VSM literature.
Importance for the article. §2.4 states the split: Ulrich asks a normative question (whose interests, who may challenge the boundary) and extended-mind philosophy asks a constitutive one (when is a component part of the system rather than coupled to it); “a boundary can be normatively impeccable and constitutively arbitrary, or the reverse,” so two receipts are required, and Demonstration III addresses the constitutive one. The import is from Input from Cognitive Science §3, which found v01 treating Ulrich’s question as the only boundary issue. Reviewers may press on three points: v02 cites all three sources as Tier C and daggered, so any characterisation of the coupling–constitution objection must be checked against the texts; the analogy runs from individual cognition to organisations, and the co-author should be able to say why the objection survives the change of scale; and a reviewer sympathetic to Rupert will ask what the VSM’s “embedded” alternative would be—a firm that depends on its suppliers without their being operational units—and whether the model can express it.
Sources in the reading list.
- the parity principle and the Otto case; the origin of the constitutive boundary question.
- the coupling–constitution fallacy and the demand for a mark of the cognitive; the template for “no mark of the systemic.”
- extended versus embedded cognition and the argument from theoretical cost; the position a sceptical reviewer will hold.
Other important sources and authors.
- Clark, A. (2008). Supersizing the Mind: Embodiment, Action, and Cognitive Extension. Oxford University Press — the full defence, with replies to Adams–Aizawa and Rupert.
- Menary, R. (Ed.). (2010). The Extended Mind. MIT Press — the standard collection, including Clark’s reply to critics and Sutton’s complementarity principle.
- Rupert, R. D. (2009). Cognitive Systems and the Extended Mind. Oxford University Press — the systems-based account of what bounds a cognitive system, developed at length.
- Adams, F., & Aizawa, K. (2001). The bounds of cognition. Philosophical Psychology, 14(1), 43–64 — the first statement of the coupling–constitution objection, shorter than the book.
- Kaplan, D. M. (2012). How to demarcate the boundaries of cognition. Biology & Philosophy, 27(4), 545–570 — applies Craver’s mutual-manipulability criterion for constitutive relevance to the extended-mind dispute; the most operational boundary criterion in the debate and a candidate for the IIIb similarity question.
- Theiner, G., Allen, C., & Goldstone, R. L. (2010). Recognizing group cognition. Cognitive Systems Research, 11(4), 378–395 — moves the constitution question from individual-plus-tool to groups, the scale the VSM operates at.
Predictive processing, the free-energy principle and Markov blankets
Introduction. Predictive processing (PP) holds that the brain is a hierarchical generative model that continually predicts its sensory input and propagates only prediction errors upward, with precision-weighting deciding which errors are allowed to drive revision. Perception is inference; action is the other way of reducing error, by changing the world so that it matches the prediction. Friston’s free-energy principle (FEP) generalises this: any system that persists in a changing environment must, in the long run, appear to minimise a bound on surprise, variational free energy, and its boundary is formalised as a Markov blanket—a set of states that renders internal and external states conditionally independent. Bruineberg and colleagues traced the blanket from Pearl’s Bayesian networks, where it is a formal device inside a model, to active inference, where it is asserted to be the real boundary of an organism, and showed that the two uses—”Pearl blankets” and “Friston blankets”—have been run together. The first is well supported and does little metaphysical work; the second does the work but needs premises the mathematics cannot supply. A parallel caution applies to the FEP itself: as a variational principle it is close to analytic, and the empirical content lies in the auxiliary claims.
Important authors. Karl Friston (Wellcome Centre for Human Neuroimaging, University College London; FRS) originated the FEP and active inference. Andy Clark (Sussex) and Jakob Hohwy (Monash University) are the philosophers most responsible for predictive processing’s uptake. Jelle Bruineberg (Macquarie University), Krzysztof Dołęga, Joe Dewhurst and Manuel Baltieri wrote the Behavioral and Brain Sciences critique. Rao and Ballard’s 1999 model is the computational origin of cortical predictive coding.
Importance for cybernetics and the VSM. PP descends from Ashby and Conant–Ashby: a regulator that models what it regulates. Active inference reproduces several VSM functions from one principle—prediction-error minimisation across a hierarchy of timescales resembles the Three–Four homeostat, precision-weighting resembles variety attenuation, and “change the world” gives the niche-construction move Beer lacks (T8.5). There is no separate coordination organ in the FEP. If it can generate System Two’s function from the principle that generates Three’s and Four’s, the five functions are a coarse partition of one process rather than a minimal set (Applying new lenses §III). The VSM literature has not engaged PP; the FEP community, by contrast, absorbed the blanket critique in its most visible venue and continued.
Importance for the article. Three places. §2.2 adopts Pearl versus Friston blankets as the organising binary—instrumental versus constitutive readings of the five systems—and states that all three v01 demonstrations tried to test constitutive claims with evidence that licenses only instrumental ones; the rivals in every later receipt are written so the two readings predict different observations (from Input from Cognitive Science §1). §2.3 uses the FEP as the parallel case of a formally true core surrounded by empirical claims that borrow its authority, so that the theorem/operationalisation problem in Demonstration II (§7) is not parochial (Applying new lenses §III). §13.3 cites the FEP community’s absorption of Bruineberg et al. as a precedent for corrigibility without a register. Reviewers will press on whether the instrumental/constitutive binary is exhaustive, on whether a receipt can really keep the two readings apart in practice, and on the unification claim: if PP subsumes S2–S4, what is the discriminating observation? Note that Clark (2013, 2016) are Tier C and not cited in v02’s text; Bruineberg et al. is Tier A.
Sources in the reading list.
- Pearl versus Friston blankets, inference with a model versus within a model; the source of §2.2’s binary, read in full including the commentaries and response.
- the hierarchical prediction machine and its difficulties in one article; read for how precision-weighting and the action route are defined.
- the book-length version; read the chapters on action and on the embodied, situated agent for the “change the world” route.
Other important sources and authors.
- Friston, K. (2010). The free-energy principle: a unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138 — the standard short statement of the FEP from its author.
- Friston, K. (2013). Life as we know it. Journal of the Royal Society Interface, 10(86), 20130475 — where the Markov blanket is made the boundary of a living system; the paper the Friston-blanket reading rests on.
- Hohwy, J. (2013). The Predictive Mind. Oxford University Press — the other philosophical synthesis, more internalist than Clark’s; useful for the contrast.
- Rao, R. P. N., & Ballard, D. H. (1999). Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects. Nature Neuroscience, 2(1), 79–87 — the computational model that PP generalises.
- Raja, V., Valluri, D., Baggs, E., Chemero, A., & Anderson, M. L. (2021). The Markov blanket trick: On the scope of the free energy principle and active inference. Physics of Life Reviews, 39, 49–72 — a second critique of the blanket’s scope, complementary to Bruineberg et al.
- Seth, A. K. (2015). The cybernetic Bayesian brain: From interoceptive inference to sensorimotor contingencies. In T. Metzinger & J. M. Windt (Eds.), Open MIND. MIND Group — draws the line from Ashby and Conant–Ashby to predictive processing explicitly.
Coordination dynamics and metastability
Introduction. Coordination dynamics studies how the components of a system—fingers, limbs, neural populations, people—form and dissolve patterns of coordinated behaviour, using the tools of nonlinear dynamics and Haken’s synergetics. Its founding result is the Haken–Kelso–Bunz model: two fingers oscillating in anti-phase switch abruptly to in-phase as frequency rises, a phase transition described by a single collective variable, relative phase, with no controller choosing the switch. Metastability is the regime that follows when coupling is weak relative to the difference between the components’ intrinsic frequencies: the system has no stable attractor, but retains “ghosts” of them, so that tendencies toward integration (synchrony) and segregation (independence) coexist and coordination patterns form transiently and dissolve. Tognoli and Kelso argue that the brain lives in this regime, that it explains flexible switching among functional networks without a switching organ, and that it can be detected in recordings as dwelling near, rather than at, phase-locked states. The general lesson: coordination is a property of coupling and dynamics, not of a module.
Important authors. J. A. Scott Kelso (Center for Complex Systems and Brain Sciences, Florida Atlantic University; also Ulster University) founded coordination dynamics; Hermann Haken (University of Stuttgart) supplied synergetics and co-authored the 1985 model; Emmanuelle Tognoli (Florida Atlantic University) is Kelso’s main collaborator on brain coordination. Karl Friston’s early work on transients and metastability, and Gustavo Deco and Morten Kringelbach’s whole-brain models, carry the concept into computational neuroscience.
Importance for cybernetics and the VSM. Beer gives coordination a box: System Two, the anti-oscillatory function among operational units, taught as a structure (schedules, standards, sympathetic ganglia). Coordination dynamics says the brain achieves coordination through phase relationships and metastable regimes, not through a coordination organ. That supports reading System Two as a dynamical regime of the System One assembly rather than a component of it, and it explains why a cross-sectional instrument would fail to resolve it: a regime is a property of time series, not of a snapshot. The VSM literature has not used this work, though Beer’s own description of System Two as damping is already dynamical in content.
Importance for the article. The import is from Variety and Channels Now §9 (“coordination without a coordinator”), which read metastability alongside the feed-forward-loop result and concluded that “coordination lives in topology plus dynamics, not in a module.” v02 does not cite Tognoli and Kelso by name; the idea enters in three places. §7.1 states IIb as a dynamical, subgraph claim—System Two’s damping is a consequence of the One–Two–Three topology, measurable as asymmetric escalation latency. §9.1 (IVd) asks whether the same assembly carries information about the future that no component does. §5.3–5.5 record that the confirmatory factor analysis could not separate Three-star from Three, which the critique explained as the failure of a cross-sectional instrument to resolve a regime. The co-author should be ready to say that “System Two is a regime” is a rival reading of the same evidence as “System Two is a structure,” and that IIb and IVd are the discriminators. A reviewer will ask what organisational observable corresponds to relative phase, and whether a dynamical reading of System Two is a narrowing or an abandonment of Beer’s claim.
Sources in the reading list.
- metastability defined, contrasted with stable synchrony, and its detection in neural data; read for the argument that coordination needs no coordinator.
Other important sources and authors.
- Haken, H., Kelso, J. A. S., & Bunz, H. (1985). A theoretical model of phase transitions in human hand movements. Biological Cybernetics, 51(5), 347–356 — the founding model; note the venue.
- Kelso, J. A. S. (1995). Dynamic Patterns: The Self-Organization of Brain and Behavior. MIT Press — the programme in book form, with the concepts of collective variable and control parameter.
- Kelso, J. A. S. (2012). Multistability and metastability: understanding dynamic coordination in the brain. Philosophical Transactions of the Royal Society B, 367(1591), 906–918 — the clearest short statement of the metastable regime and why it differs from multistability.
- Bressler, S. L., & Kelso, J. A. S. (2001). Cortical coordination dynamics and cognition. Trends in Cognitive Sciences, 5(1), 26–36 — coordination among cortical areas as transient phase relations; the bridge from motor to cognitive coordination.
- Deco, G., & Kringelbach, M. L. (2016). Metastability and coherence: Extending the communication through coherence hypothesis using a whole-brain computational perspective. Trends in Neurosciences, 39(3), 125–135 — metastability in whole-brain models; useful for how it is measured.
- Haken, H. (1983). Synergetics: An Introduction (3rd ed.). Springer — the physics behind order parameters and the slaving principle, which Simon’s timescale separation (T3.6) parallels.
What you should be able to say after this theme
- The VSM’s central equivocation is between an instrumental reading (the five systems are a construct laid over an organisation, judged by usefulness and reliability) and a constitutive reading (the functions are really there and really necessary, judged by truth), and Bruineberg et al. diagnosed the same slide for Markov blankets in active inference.
- Evidence that practitioners find the vocabulary useful licenses the instrumental reading only; v01’s demonstrations tested constitutive claims with such evidence, and every v02 receipt states rivals so that the two readings predict different observations.
- The free-energy principle, like Ashby’s law, has a near-analytic core surrounded by empirical claims that borrow its authority, so the theorem/operationalisation problem in Demonstration II is a general property of principle-based systems theories, not a VSM defect.
- If active inference can derive System Two’s function from the same principle that yields Three and Four, the five functions may be a coarse partition of one process; that is a candidate receipt, not a settled objection.
- There are two boundary questions: Ulrich’s normative one and the extended-mind constitutive one; they are independent, the VSM has answered only the first, and Demonstration III addresses the second.
- The VSM has no mark of the systemic: its criterion of operational closure cannot distinguish a unit coupled to the firm from a unit that is part of it, which is the coupling–constitution objection transposed.
- Distributed cognition derives system boundaries from observed propagation of representational state, giving Demonstration IIIb an independent partition against which VSM recursion analysis can converge or diverge; convergence would be the model’s first corroborated prediction of that kind.
- Coordination dynamics shows coordination arising from coupling and metastable dynamics without a coordinating organ, which supports reading System Two as a regime rather than a structure and explains why Schwaninger and Scheef’s cross-sectional instrument could not separate Three-star from Three.