The object: a framework for formation

Ostrom was careful to distinguish frameworks, theories, and models: a framework supplies the universal elements and the metatheoretical language; theories and models make assumptions within it ({{ref:concept-iad-framework|IAD framework}}). What this work proposes is a framework — but its object is emergent governance itself, not another resource domain ({{ref:syn-emergent-governance-framework|the synthesis note}}). The distinction from the knowledge commons should be drawn precisely, because it is the paper's claim to a contribution rather than an application. IAD analyzes an action arena structured by exogenous variables — rules-in-use, attributes of the community, the biophysical and technical world — and holds those variables fixed at the time of analysis; the fixing is a footnote, an acknowledged convenience. The GKC school loosened it for constructed commons, adding the feedback loop by which a community's outputs become its own inputs ({{ref:concept-gkc-framework|GKC framework}}). This framework takes the remaining step: it makes the feedback loop the object. Formation is what happens when action situations generate their own future rules — when participants drift or deliberately shift from operational play into collective choice and occasionally into constitutional revision ({{ref:concept-rule-levels|rule levels}}). Ostrom's theory treats that level-shift as the elementary act of self-governance; here it becomes the elementary observable. On Nostr the exogeneity footnote is untenable in the most productive way: rules reshape the facility set within weeks, so the "held constant" clause fails, and what fails it is exactly the phenomenon under study.

The unit of analysis is therefore the formation episode, sitting earlier in time than the institution — the analog of the action arena, one level before rules settle. An episode has an anatomy. It opens with a shock: an exogenous disturbance the current rule configuration cannot absorb — a spam wave, a coordinated malicious influx. A response portfolio follows: uncoordinated actors across the stack deploy existing exclusion methods and mint new ones, a-priori guesses and a-posteriori attestations alike ({{ref:claim-multiplicity-stack|multiplicity}}). Adaptation registers at every level of rules at once — operational (mutes, filters, policy toggles), collective-choice (relay policy revisions, client defaults), sometimes constitutional (a NIP). And the episode closes in settlement — the cycle converges, the boundary re-expands — or in failure: a relay dies, a community disperses, a method is abandoned. Both endings are data. The failures are precisely what the survivorship-filtered record never contained, and they must be archived deliberately, or the bias the framework escapes at the field level quietly recurs at the archive level ({{ref:concept-survivorship-bias|survivorship bias}}). The {{ref:claim-expand-contract|expand–contract}} cycle is the prototype episode; in IAD terms, an episode is a linked chain of action situations whose principal outputs are rules.

The architectural section — the breakdown of emergent governance itself, as distinct from any one substrate — asks what must be true of a system for formation episodes to be observable at all. Four conditions. First, actions must be typed, public, signed, and timestamped: institutional statements with a wire format, codable into the grammar of institutions rather than reconstructed from recollection. Second, identity must be portable and separable from venue, so that exit is legible rather than silent, and boundary events are attributable — the observer can tell who left, who was filtered, and who merely moved. Third, policy and enforcement must be separately visible, so that rules-in-form and rules-in-use can be compared directly — a stated relay policy against its observed accept/reject behavior — instead of elicited after the fact ({{ref:concept-rules-in-use|rules-in-use}}). Fourth, the record must persist: on Nostr effectively irrevocably, so the history of rule changes — including the reversed and abandoned ones — is part of the archive rather than of survivors' memories ({{ref:q-advisory-deletion|advisory deletion}} is the cost of this same property). Graded against these conditions, git repositories, wikis, and DAOs are partial observatories; Nostr is, so far, the maximal one. That grading is what makes the framework portable rather than a platform study — it tells other researchers where else it applies and at what resolution — and it bounds the claim honestly: whatever coordination happens off-substrate, in the repositories and group chats, stays off-trace (the civic-sphere caveat of {{ref:src-modular-politics|Modular Politics}}), which is why the trace locates and the interviews carry.

The evaluative layer then replaces criteria that presuppose an existing institution. The 2007 criteria — sustainability, efficiency, equity, preservation — all ask how a standing commons is doing; a forming one cannot yet be scored on them. Formation-stage criteria ask instead: is boundary-setting capacity developing — after each shock, can actors still reconstitute a boundary, at falling cost ({{ref:q-boundary-sybil|boundary / Sybil}})? Are sanction signals acquiring legitimacy — do non-binding reports move behavior, do filters adopt them, do exits follow them ({{ref:q-decoupled-vs-unitary-sanctions|decoupled vs unitary}})? Is the shock cycle converging — contraction depth shrinking, recovery accelerating — or thrashing ({{ref:q-maturation-hypothesis|maturation}})? And is the option space still growing — generativity as the working definition of robustness ({{ref:q-generativity-robustness|generativity}})? Each criterion pairs with an open question in the map below; together they are the framework's answer to "evaluated against what?" for an institution that does not yet exist.

The reusable instrument that carries all of this is a coding table from event kinds to institutional statements: kind 1984 as a norm whose or-else is crowdsourced, a mute list as an operational boundary act, a relay policy document as a collective-choice output, a NIP as constitutional text ({{ref:concept-adico-grammar|ADICO grammar}}). The table is what lets someone apply the framework to their own substrate instead of reading a Nostr case study — the difference between a framework paper and a field report.

The commons school's own trajectory legitimates the whole move. Its mature statement defines commons as a mode of governance — institutionalized sharing of resources among members of a community — rather than a type of good, insists that the relevant social dilemmas are an empirical question rather than an assumed tragedy, and has begun, literally, to write "beyond design principles" ({{ref:concept-gkc-framework|GKC framework}}, {{ref:src-governing-knowledge-commons|Governing Knowledge Commons}}). The knowledge commons is not discarded in any of this; it is placed. It becomes the resource type of the demonstration case — the setting in which formation is watched — rather than the object of the theory, which is what keeps the paper out of the commons-advocacy corner its critics call romantic ({{ref:src-exchange-theory-web3|the exchange theory's jab}}) and makes its contribution an instrument rather than an endorsement.