Schneider et al. 2021 — Modular Politics: Toward a Governance Layer for Online Communities

What it is

A conceptual proposal (explicitly "a prologue," no prototype) for a composable governance layer for online communities, grounded in Ostrom's IAD, with an ontology (Platform/Instance/Org/Module/Entity/ Monitor/Permission) and an open-standard implementation strategy. First author is Schneider (not Zhang/Frey); PolicyKit appears nowhere in the paper — do not cite it to this source.

Core claims

1. 16:1 — governance tools are "remarkably impoverished, typically relying on assigning unchecked power to admins and moderators." 2. 16:2 — "Monarchic or oligarchic practices emerge in online networks more easily and frequently than liberal or democratic ones." 3. 16:2 — four design goals: Modularity, Expressiveness, Portability, Interoperability. 4. 16:2, 16:15–16:16 — chosen strategy (over cloud API and DLT): "the most flexible strategy would be to define an open standard... implemented through a variety of open-source software libraries." 5. 16:4 — decentralized/federated platforms "favor 'exit' over 'voice'" (Hirschman). 6. 16:5–16:7 — IAD chosen for its empiricism toward not-yet-theorized forms; communities formalize action situations threaded "across polycentric networks" (16:7). 7. 16:10–16:11 — Monitors: read-only composable evaluators of governance data, e.g. cross-instance censorship comparisons. 8. 16:11 — "all permissions... derive from those that the platform administrators specify at the level of the Instance." 9. 16:16–16:17 — digital-juries use case: flag → jury → removal as one binding platform pipeline. 10. 16:18 — computational governance data enables "new and more useful taxonomies of governance practices." 11. 16:19–16:20 — limitations: expressible space undefined; false universalism risk; software insufficient without a "civic sphere" ("the confluence of tools, culture, and context").

Key concepts

- {{ref:concept-implicit-feudalism|Implicit feudalism}} (16:1–16:2, ref [121]) - {{ref:concept-governance-as-software|Governance as software}} (Modules, 16:9–16:10; standard, 16:15) - {{ref:concept-iad-framework|The IAD framework}}, {{ref:concept-action-arena|Action arena}}, {{ref:concept-polycentricity|Polycentricity}} (16:5–16:7) - {{ref:concept-exit-voice|Exit, voice, and forking}} (16:4)

Hooks to the Nostr thesis

- CENTRAL: Nostr is a de-facto realization of their chosen strategy — an open standard (NIPs) implemented by independent libraries, clients, relays — but grown, not designed as a governance layer; governance accreted as NIPs (kind-1984 reports, mute lists, NIP-65). The thing they wanted, grown instead of designed. Yet Nostr breaks their ontology: no Platform/Instance root (contra 16:11 permissions derivation), permissions polycentric all the way down, and interoperability default-on where theirs is consent-gated (16:11–16:12) — {{ref:claim-no-global-bannability|No global bannability}}, {{ref:concept-polycentricity|Polycentricity}}. - Design-goals audit: modularity event kinds; expressiveness any kind mintable; portability = signed events readable everywhere; interoperability = protocol-native — {{ref:claim-multiplicity-stack|Multiplicity}}. - Monitors (16:10–16:11) exist today as WoT scorers, kind-1984 dashboards, relay-policy trackers — {{ref:claim-formation-observable|Formation is observable for the first time}}. - Digital juries decoupled: on Nostr flag, judgment, and enforcement are separate uncoordinated actors — {{ref:claim-split-sanctions|Split sanctions}}, {{ref:q-decoupled-vs-unitary-sanctions|Q: Does decoupled sanctioning govern better or worse than unitary?}}. - Exit-over-voice (16:4): Nostr makes exit near-frictionless (portable keys and graph), turning exit into operator discipline — {{ref:q-does-competition-improve-governance|Q: Does interjurisdictional competition yield better governance?}}, {{ref:concept-interjurisdictional-competition|Interjurisdictional competition}}. - Civic-sphere caveat (16:20): event streams miss the GitHub/Telegram culture layer — pairs with Schroeder's off-platform worry (1956).

Tensions / open questions

- Aspiration, not spec: no benchmarks to borrow; strengthens the "unplanned fulfillment" framing but proves nothing empirical. - Top-down residue (16:8, 16:11) sits closer to platform-hosted governance than Nostr's operator-less substrate. - DLT rejected for expressiveness costs (16:15); Nostr is the unnamed third path: open standard + cryptographic identity, no ledger. - No adversarial model (spam, sybils) — exactly where Nostr's record is richest — {{ref:q-boundary-sybil|Q: How does a commons reconstitute boundaries it cannot enforce?}}. - Ostrom nesting assumes containment (16:9); Nostr's bounded institutions overlap and cross-cut — {{ref:claim-composable-bounded-institutions|Bounded institutions compose on the open substrate}}.

Related: - {{ref:emergentgov|Index}} - {{ref:src-open-content-governance|Schroeder & Wagner 2012}} - {{ref:src-digital-constitutionalism|Alston (working paper)}} - {{ref:concept-guardianship-without-office|Guardianship without office}} - {{ref:concept-adaptive-governance|Adaptive governance as feedback architecture}} - {{ref:concept-requisite-variety|Requisite variety (Ashby's law, institutionalized)}} - {{ref:agent-open-content|Agent session}}