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.