What Part of the Organisation Does a Person Think For?

ThisThis cluster opened by asking whether an organisation can know something no individual member knows, and answered yes, in at least the distributed-processing and emergent-coordination senses — a ship's position fix, a hospital ward's composite picture of a patient. That answer treats the members doing the knowing as largely interchangeable: any sufficiently competent person could, in principle, be standing where the bearing-taker or the plotter is standing. In practice they can't, and the reason why is worth taking seriously on its own terms. A collective doesn't think merely because several people think within it. It thinks through the differentiated, interacting roles by which one person's activity becomes the group's attention, memory, interpretation, and action — and those roles are not job titles so much as cognitive functions, enacted socially rather than performed by one mind alone.

Five Things Easily Collapsed Into "Role" FoundationalKnowledge that endures for decades — core principles

AA 2009 study of online community behaviour by this site's own author and colleagues, Folksonomological Reification, treats the roles a person takes on in a community as outward expressions of strands of identity — borrowing from identity theory's account of a self built from the meanings a person attaches to the different parts they play across a differentiated social life, and from the idea of a digital persona: the aggregate model other people and systems build of someone from their visible digital traces, assembled strand by strand¹. Sharpening that account further than the paper itself does, in a distinction this page is drawing rather than one the paper makes in quite this form, "role" is doing at least five different jobs that are worth keeping separate. A position is a formally recognised place in a structure: principal investigator, reviewer, committee chair. An assigned role is a function someone is expected to perform, whether or not they actually do. An enacted role is the recurrent pattern of behaviour actually visible in practice — someone formally a researcher who recurrently acts as an explorer of other fields' ideas. An attributed role is the part other members believe someone plays, which matters because trust and attention follow perceived roles rather than necessarily formal ones. An identity strand is the meaning the person themselves attaches to occupying that role. What someone is appointed to do, what they actually do, what others think they do, and what they understand themselves to be doing need not coincide — and a good deal of organisational friction lives in the gaps between those four.

An updated version of the activity-theory relationship model in Folksonomological Reification's Figure 7. Roles sit under Rules and Division of Labour, not as a direct property of a person's identities — and, as the original figure makes a point of showing, there is no arrow running straight from Identities to Community. Without a shared tool mediating between them, there is no community for a role to be enacted in at all.the tool is the community not just the people

A Taxonomy of Roles, by Cognitive Function FoundationalKnowledge that endures for decades — core principles

FolksonomologicalFolksonomological Reification catalogues a working taxonomy of roles observed in online communities: the Explorer, who brings material from outside the group into its shared attention; the Writer (and the Diarist, a specific kind of writer); the Reviewer, who bridges explorers and writers by taking existing content and analysing it; the Classifier and the Editor, who produce and extend the group's shared categories; the Translator, who carries material across a language or representational boundary; the Ambassador, who connects material from one community to another that might find it useful; the Socialite and the Reader, who contribute and consume conversational content respectively; and the Troll and Flamer, inherited from Usenet, the first of which the paper is careful to note was not always considered purely negative — a deliberate provocateur, not only a disruptor. Composite roles bundle several of these into one stable configuration: a Technology Steward combines exploring, reviewing, leading, and mentoring, directed at selecting and supporting a specific technology for the community's benefit; an Ethnographer is a reviewer whose object is the community's own tools and practices rather than its content. Reading this list by the cognitive job each role is doing, rather than by the paper's own flat list — again, this regrouping is this page's own move, not the paper's structure — the roles cluster into bringing material in (Explorer), turning it into something inspectable (Writer, Diarist), relating it to what the group already holds (Reviewer), stabilising how it's found again (Classifier, Editor), carrying it across a boundary (Translator, Ambassador), and watching how the group itself is working (Ethnographer).roles are social, not psychological

An updated version of Figure 8's illustration of roles and identities: the same person carries different identity strands into different communities, enacting a different role — drawn from a different part of this page's taxonomy — in each.

How a Role Becomes Real: the Amanda and Bob Mechanism FoundationalKnowledge that endures for decades — core principles

TheThe paper's own central example shows how an enacted role turns into an attributed one. Amanda shares something she's found with her network, S. Bob, a member of S, re-shares it with his own network, T, which doesn't overlap with S. If Bob re-shares without attributing Amanda, the members of T learn about the resource and about Bob's interest in it, but nothing about Amanda — and if the resource turns out to be dull, that reflects on Bob alone. If Bob attributes the find to Amanda, members of T learn about the resource, about Amanda's interests, about Bob's interest in Amanda's interests, and — cumulatively, over many such exchanges — start to trust Amanda's judgement specifically in her role as an explorer, without ever having interacted with her directly. The mechanism generalises past this one example into a reinforcing loop: behaviour becomes visible, visibility earns attribution, attribution builds trust, trust creates opportunity, opportunity invites more of the same behaviour, and repeated behaviour eventually becomes part of how the person understands their own identity. A role is not assigned once and then simply held. It's built, transaction by transaction, out of exactly the kind of attribution Bob either does or doesn't make. A later, more formal piece of work on this site treats trust between autonomous parties as exactly this kind of asymmetric, context-dependent, continuously-updated quantity rather than a flat reputation score — the Amanda-and-Bob mechanism above is an earlier, informal sketch of the same intuition, worked out before that formal calculus existed rather than derived from it.a reputation is a map of trust, not a fact

An updated version of Figures 9 and 10's two attribution cases. Only in the second case does the wider network T have any way of learning that Amanda, not just Bob, is worth trusting in the explorer role — attribution is what lets an enacted role become an attributed one outside a person's own immediate network.

What the Collective Knows About Who Knows What FoundationalKnowledge that endures for decades — core principles

DanielDaniel Wegner's account of transactive memory gives this mechanism its organisational payoff: a group doesn't need every member to know everything, it needs a reasonably accurate, distributed answer to who in the group is likely to know what, so that a question can be routed to the right person rather than answered from scratch by whoever happens to be asked first². This cluster's page on shared mental models already covers what lets two experienced teammates coordinate in few words; transactive memory is the organisational-scale cousin of that same resource, applied to expertise rather than to a shared task model. Read together with the role distinctions above — again, this combination is this page's own extension, not a move either source makes on its own — an organisation's transactive memory isn't really indexed by topic alone. "Pat knows about X" is a flatter claim than what the attribution mechanism actually builds, which is closer to "in context C, Pat is recurrently trusted to perform function R in relation to X" — a highly trusted Explorer in one domain can be a poor Classifier, or an excellent Reviewer inside one community and an unreliable Translator across its boundary, and a transactive memory worth the name tracks the role as well as the topic.

When the Role Ecology Fails Applied / MethodologicalKnowledge with a 5–10 year half-life — stable practice

NoneNone of this is the 2009 paper's own argument — it's what falls out of applying its machinery to the organisational-cognition question this cluster has been asking throughout, and it sharpens this cluster's page on field-level blind spots directly. A collective blind spot can result from several distinct failures in its role ecology, not just one. Role absence is the simplest: no one is performing the exploratory or translational function that would bring an unfamiliar possibility into view, so nothing rejects it — it simply never gets proposed. Role capture happens when a function exists but is monopolised by people who share the same training, network, and assumptions, so the function runs without ever actually testing the group's blind spots against anything genuinely outside them. Role misattribution happens when the community's attributed model of who performs a function drifts from who actually does — trust and attention keep flowing to a formal position after the person who actually does the exploring or translating has gone unrecognised. Role overload happens when the person expected to explore, translate, and challenge is also expected to deliver the core operational work, and the boundary-spanning activity is the first thing displaced under pressure. Role homogenisation is the subtlest: a field full of excellent reviewers, all trained to apply the same standard, can be less cognitively capable collectively than one with a more uneven mix of explorers, translators, and critics, because reviewing well by a shared standard is exactly the function that can't notice a possibility the standard itself excludes.

Strongest Objection: Naming a Role Doesn't Guarantee It's Filled Honestly FoundationalKnowledge that endures for decades — core principles

AA reasonable objection to all of this is that it relabels ordinary social facts — some people notice things first, some people are good at explaining them, everyone already knew that — in unnecessary jargon. The response is the same one the 2009 paper makes about its own project: naming the roles explicitly is what lets a community compare, discuss, and deliberately design for them, rather than leaving the division of cognitive labour to accumulate as an unexamined accident of who happened to be in the room. A sharper version of the objection survives that response. An Ethnographer role watching how a community works is still a member of that community, situated inside whatever assumptions the community already holds — reflexivity can expose an assumption that was previously invisible, but it can't guarantee an assumption-free vantage point to expose it from, which is the same limit this cluster's page on field-level blind spots already runs into from a different direction.

Provisional Conclusion FoundationalKnowledge that endures for decades — core principles

AA person participates in an organisation's collective cognition not simply by contributing knowledge, but by enacting one or more differentiated roles through which the collective discovers, interprets, classifies, translates, and acts on what its members individually notice — and those roles are built, not assigned, through a real mechanism of behaviour, attribution, and trust that this page has tried to state precisely rather than leave as a vague appeal to "team dynamics." An organisation's cognitive capacity depends not only on how capable its individual members are, but on whether its role ecology is diverse, correctly attributed, and revisable — which gives a shared paradigm a second description worth holding alongside the first: not only a field's shared mental model, but its role ecology, the set of cognitive performances the paradigm makes recognisable, legitimate, and worth rewarding.

Questions for Further Thought

• Can an organisation notice that a role is structurally missing, rather than merely under-performed — and if the previous section's objection holds, can it notice this from inside its own role ecology at all?
• Does the Amanda-and-Bob attribution mechanism scale past small, overlapping personal networks, or does it break down once a community is too large for any one member to track who's attributing whom?
• If role homogenisation is a real risk, is there a reliable way to tell a healthily diverse role ecology from one that merely looks diverse because its members hold different job titles while actually all performing the same underlying cognitive function?

Related Topics

• Can an Organisation Know Something No One in It Knows? — the distributed-cognition claim this page explains the differentiated mechanism behind.
• Communities of Practice: Learning by Belonging, Not by Being Told — legitimate peripheral participation as the route by which a newcomer grows into one of this page's enacted roles.
• Shared Mental Models: Why Experienced Teams Need Fewer Words — the task-coordination resource this page's transactive memory is the expertise-routing cousin of.
• Can a Whole Research Field Have a Blind Spot? — the field-scale question this page's role-ecology failures are a mechanism for.
• Inside the Rhizome: Four Social Groupings, and How Each Should Count — a related but differently-cut taxonomy, sorting a network's members by trust calculus and turnover rather than by cognitive function.
• A Distributed Embedding Calculus of Trust — the later, formal treatment of the same asymmetric, evidence-updated trust this page's Amanda-and-Bob mechanism sketches informally.

Further Reading

• Parslow, P., Lundqvist, K. Ø., Porter-Daniels, E., Hussey, R., & Williams, S. (2009). Folksonomological Reification. University of Reading. Not formally peer-reviewed; released instead for open, crowd-based review through a personal learning network.
• Wegner, D. M. (1987). Transactive memory: A contemporary analysis of the group mind. In B. Mullen & G. R. Goethals (Eds.), Theories of Group Behavior (pp. 185–208). Springer-Verlag.

References