Layer 8: Disciplinary Framework

What this means

No platform can exist without some way of responding to bad behavior. Harassment, sabotage, doxxing, coordinated manipulation, IP theft — these happen, and a platform that pretends they don't will destroy itself.

But the disciplinary system is also where most platforms go wrong in the opposite direction. They build punitive cultures that drive away good contributors. They create "enforcement officer" career paths for people who enjoy policing others. They let warnings accumulate without grace until someone's record is permanently damaged. They turn legitimate disputes into weapons by letting complaints be filed without consequence.

This Layer is designed explicitly against that failure mode — the "HOA" pattern of petty, accumulating, proceduralized enforcement. Instead, it presumes good faith, responds proportionally, sunsets its own memory, and builds in multiple protections against weaponization.

Formal Text

Article I: Guiding Principle

Discipline exists to protect the work and the community, not to enforce conformity or punish annoyance. The system presumes good faith, responds proportionally to actual harm, and sunsets its own memory so contributors are not shadowed by past mistakes forever.

Article II: Jurisdiction

2.1 Franchise-level disciplinary authority covers misconduct occurring within a single Franchise: in-Franchise harassment, quality issues within Franchise work, violations of Franchise-specific rules, canon sabotage within the Franchise.

2.2 Network-level disciplinary authority covers misconduct affecting the platform broadly: cross-Franchise harassment, manipulation of the Credit system, IP theft, coordinated bad-faith behavior, doxxing, illegal content.

2.3 All disciplinary actions at any level are appealable to the Arbiters.

2.4 Franchises handle first-instance action within their jurisdiction. Network action is reserved for cases Franchises cannot or should not handle (e.g., cross-Franchise misconduct, misconduct by Team members against their own Franchise).

Article III: The Disciplinary Ladder

Four rungs of possible response, in order of severity. Escalation is not automatic; serious offenses may skip rungs, minor issues often stop at the first rung without escalation.

Rung 1: Warning

• Private notice describing the concern
• Not publicly visible to others
• Expires and clears from record after 12 months of no further issues in the same category
• Appealable to the Arbiters even at this level

Rung 2: Credit Penalty

Two sub-types:

Prospective penalty: Reduction of future Credit accrual rate for a defined period. Used for pattern issues — persistent low-quality contribution, repeated minor misconduct.

Revocation: Specific past Credits removed. Used only when the contribution itself was fraudulent, plagiarized, gamed, or produced in bad faith.

• Credit penalties public as an action ("Penalty applied") without specific details
• Appealable to the Arbiters

Rung 3: Suspension

• Time-boxed removal of contribution rights: 30, 90, or 180 days
• Team membership preserved but inactive during suspension
• Power decays during suspension but does not drop below the floor
• Used for serious misconduct that does not rise to ban level
• Appealable to the Arbiters; appeal suspends the suspension pending ruling

Rung 4: Ban

• Permanent removal from the platform
• Reserved for severe harassment, doxxing, illegal content, systemic manipulation, repeated violations after lower rungs, threats
• Requires either two Arbiter co-signatures (for Network bans) or full Franchise Team vote (for Franchise bans)
• Banned contributor's work remains on the platform (retroactive erasure punishes readers, not the offender)
• Credits earned through work that remains are not revoked; Credits earned through the bad-faith behavior itself are
• Appealable to the Arbiters for initial review
• Restoration applications permitted after 2 years; reviewed by Arbiters

Article IV: Credit Revocation — Narrow Definition

Revocation is a narrow remedy, not general punishment.

4.1 Revocable when:

• The contribution itself was plagiarized, fraudulent, or produced in violation of Franchise policy (including undisclosed AI generation where disallowed)
• The Credits were inflated via detected gaming (sock-puppet upvotes, reciprocal rings, manipulated metrics)
• The contribution was made to manipulate the system rather than to contribute genuinely

4.2 NOT revocable for:

• Unrelated misconduct (harassment does not cost Credits for unrelated good work)
• Subsequent change in community values (work legitimate at contribution time remains legitimate)
• Unpopularity or changed taste
• Disagreement with the contributor on unrelated matters

4.3 Revocation requires specific finding that one of the Article 4.1 conditions applies, with written justification entered in the ledger.

Article V: Anti-HOA Provisions

Explicit structural guards against the failure mode of petty, accumulating enforcement.

5.1 Proportionality requirement. Every disciplinary action must cite the rung being applied and why. Skipping rungs upward requires written justification on record. Staying at lower rungs does not require justification — leniency is the default.

5.2 Warning sunset. Warnings expire after 12 months of clean conduct in the same category. Contributors cannot be warning-stacked across a career. A contributor with two warnings 3 years apart, no other issues, is treated as someone with no active warnings.

5.3 Frivolous-complaint handling. Complaints found by the Arbiters to be in bad faith result in disciplinary action against the complainant. This prevents weaponization of the complaint system.

5.4 No enforcement quotas. Team members performing moderation work are not evaluated or Credited based on volume of enforcement actions. Moderation Credits come from sustained good judgment, not action count. This kills the "enforcement officer" incentive structure.

5.5 Cooldown on repeat complaints. If a contributor has been complained about and cleared (no action taken) within the last 60 days, new complaints about the same person require two-moderator review before any disciplinary action. Prevents pile-on campaigns.

5.6 Public enforcement report, not case-by-case public shaming. Monthly transparency report shows aggregate patterns (counts by rung, category, trends) without naming individuals for lower-rung actions. Specific names are public only for suspensions and bans.

5.7 Restoration path. Suspended contributors return automatically at end of term — no re-earning of rights. Banned contributors may apply for restoration after 2 years.

5.8 No retroactive rule changes. Disciplinary standards are applied as they existed at the time of the alleged misconduct. Rules changed after the fact do not apply to prior conduct.

Article VI: Transparency

6.1 Public: The fact that an action occurred, the rung applied, and the general category (harassment, quality, manipulation, etc.). No personal details.

6.2 Private: Specifics of the complaint, evidence, back-and-forth, identity of complainant unless they choose to be named.

6.3 Complainant anonymity: Complainants are never named publicly by the system unless they choose to be.

6.4 Aggregate patterns public: Monthly enforcement report shows system-wide patterns without identifying anyone below suspension level.

6.5 Suspensions and bans: Named publicly, with rung and category.

Article VII: Appeals

7.1 All disciplinary actions, at every rung, are appealable to the Arbiters.

7.2 Appeal timing: contributor files appeal within 30 days of action.

7.3 Arbiter review timing: Arbiters rule within 14 days of receipt.

7.4 Appeal outcomes:

• Uphold: Action stands
• Reduce: Action lowered one or more rungs
• Vacate: Action removed from record entirely, as if it had not occurred

7.5 Vacated actions are genuinely cleared, not marked "vacated." The record shows no disciplinary history for the vacated matter.

7.6 Appeal outcomes are published in aggregate (same standards as original actions).

7.7 Repeated appeals for the same action: only one appeal per action, except on declared procedural error.

Article VIII: Dispute-Gated Content Monetization

Related to the dispute-handling mechanisms of Layer 9 (Commons/Home) and Layer 10 (Fact Lifecycle).

8.1 Content with active, unresolved disputes regarding its canonical grounding cannot generate monetization revenue during the dispute period.

8.2 Upon resolution, monetization activates (or reactivates) retroactively to the dispute-filing date if resolved in the creator's favor, or from resolution forward if the creator must modify the content to proceed.

8.3 Article-level demonetization is the unit of effect. A disputed claim within an article demonetizes the whole article until resolved or removed. Creators may resolve by removing the disputed claim, marking it as interpretation, or winning the dispute.

8.4 Disputes filed by accounts with a history of rejected disputes require Arbiter screening before they trigger demonetization. This prevents weaponization of disputes as income-denial tactics.

8.5 No clawback of revenue earned before a dispute was filed. Dispute effects are forward-only from filing date.

8.6 Grace period for Fact graph changes: when Franchise Teams amend Facts, content referencing the old Facts is flagged for update but not automatically demonetized. Creators have 30 days to update or adjust before dispute-gating applies.

Article IX: Moderator Conduct

9.1 Moderators and Team members performing enforcement are themselves subject to the disciplinary framework. Abuse of enforcement authority is misconduct.

9.2 A pattern of overturned disciplinary actions by a given Team member may trigger review of that Team member's judgment and, if warranted, removal from moderation duties.

9.3 Moderators cannot discipline contributors with whom they have undisclosed conflicts of interest (personal, financial, or ongoing dispute). Conflicts require recusal and handoff to another Team member.

Article X: Emergency Situations

10.1 In imminent harm situations (active harassment, doxxing in progress, illegal content being shared), immediate removal of harmful content and temporary contributor suspension may be taken without following the full process.

10.2 Emergency actions must be reviewed by Arbiters within 72 hours.

10.3 If the emergency basis is not found to be legitimate, affected contributors are restored and the responsible moderator is subject to review per Article IX.