Governance overview

Chive uses a Wikipedia-style moderation model where the research community collaboratively manages the knowledge graph, authority records, and content policies. This decentralized approach ensures that no single entity controls the classification of scholarly work.

Governance philosophy

Chive's governance rests on three pillars:

Pillar	Description
Community-driven	Researchers propose and vote on changes
Expertise-weighted	Domain experts have greater influence in their fields
ATProto-native	Governance data lives in a dedicated PDS, ensuring portability

Key roles

Community members

Any authenticated user can:

• Propose new fields or changes to the knowledge graph
• Vote on pending proposals
• Tag eprints with user-generated tags
• Report content policy violations

Trusted editors

Community members elevated for consistent quality contributions:

• Review and approve routine proposals
• Mentor new contributors
• Weighted votes (2.0x standard)

Domain experts

Users recognized for expertise in specific fields:

• Vote on proposals in their area of expertise
• Provide specialist input on field changes
• Weighted votes (3.0x standard)

Graph editors

Users responsible for maintaining knowledge graph nodes:

• Manage knowledge graph nodes (fields, facets, authorities)
• Reconcile with external vocabularies (Wikidata, LCSH, VIAF)
• Approve node and edge changes
• Weighted votes (3.0x standard)

Administrators

Platform administrators with oversight responsibilities:

• Resolve disputes
• Manage the Governance PDS
• Veto power on proposals
• Highest vote weight (5.0x)

Governance scope

What is governed

Area	Governance mechanism
Knowledge graph nodes	Community proposals and voting
Knowledge graph edges	Community proposals and voting
Facet definitions	Proposal with lower threshold
Content policies	Administrator decisions
Tag promotion	Two-stage nomination and vote

What is not governed

Area	Rationale
User content	Lives in user PDSes; users control their data
Eprint metadata	Entered by authors; Chive only indexes
Personal tags	User-generated; no approval needed
User identity	Managed by AT Protocol DIDs

The Governance PDS

All governance data lives in a dedicated Personal Data Server:

did:plc:chive-governance

This PDS stores:

• Authority records
• Facet definitions
• Organizational records
• Reconciliation history
• Approved proposals

By storing governance data in a PDS, it remains:

• Portable (can move to different hosts)
• Verifiable (cryptographically signed)
• ATProto-native (indexed by any compliant AppView)

Proposal lifecycle

1. Draft: Proposer creates and refines the proposal
2. Discussion: Community comments and suggests changes
3. Voting: Weighted votes tallied against thresholds
4. Outcome: Approved proposals are enacted; rejected ones archived

See Proposals for details.

Voting system

Not all votes are equal. Vote weight depends on:

Role	Weight multiplier
Community member	1.0x
Trusted editor	2.0x
Graph editor	3.0x
Domain expert	3.0x
Administrator	5.0x

See Voting system for thresholds and quorum requirements.

Transparency

All governance actions are public:

• Proposals and voting records are visible to all
• Administrator decisions are logged
• Annual transparency reports detail statistics

Governance documentation

Document	Description
Voting system	Thresholds, weights, and quorum
Proposals	Types, lifecycle, and requirements
Authority control	Managing authority records
Governance PDS	Technical architecture

Getting involved

Propose a change

1. Draft your proposal with clear rationale
2. Submit via the governance interface
3. Respond to community feedback during discussion
4. Await voting outcome

Become a trusted editor

1. Contribute actively (eprints, reviews, tags)
2. Demonstrate expertise in your field
3. Apply or be nominated
4. Administrator review and approval

Next steps

• Voting system: Understand how decisions are made
• Proposals: Learn how to propose changes
• Knowledge graph: Understand what is governed