# Comparative Survey Index — Auditable Cross-Reference

**Status:** v0.1 · Normative for research traceability · Companion to `01-national-constitutions.md`, `02-nonstate-systems.md`, `03-synthesis.md`

This index is the audit trail for the kernel. Every system we surveyed gets a stable ID, a list of the meta-rules we extracted from it, the failure modes it exhibited or guarded against, and the kernel clauses that the lesson landed in. The claim "the kernel is derived from 20+ real systems" should be checkable line by line from this file. If a kernel clause cannot be traced to at least one row here, that clause owes us a justification in its annotation.

**26 systems surveyed:** 12 national/constitutional (N-01 … N-12), 14 non-state (S-01 … S-14).

---

## 1. Kernel article key

Cross-references below use the article numbering of `kernel/constitution.md` v0.1:

| Art | Title |
|---|---|
| I | Purpose, Scope, and Supremacy of Invariants |
| II | Citizenship and Suffrage |
| III | Amendment Process and Versioning |
| IV | Quorum, Deliberation, and Voting |
| V | Invariants |
| VI | The Right to Fork |
| VII | Userland Modules and Parameterization |
| VIII | Emergency Powers |
| IX | Interpretation and Dispute Resolution |
| X | Ratification, Releases, and the Public Ledger |

---

## 2. Meta-rule taxonomy (M-codes)

Recurring meta-rules found across surveyed systems. "Adopted in" lists where each landed in the kernel.

| ID | Meta-rule | Description | Adopted in |
|---|---|---|---|
| M-01 | Tiered supermajority | Change difficulty scales with blast radius of the change | Art III (semver thresholds: patch/minor/major) |
| M-02 | Mandatory deliberation period | A forced delay between proposal and ratification vote (review periods, FCPs, timelocks) | Art III (review period floor); Art IV |
| M-03 | Double-gate ratification | Two distinct bodies, or two votes separated in time, must both pass a change | Art III (kernel-major: vote + cooling-off confirmation); Art VII (modules may add gates) |
| M-04 | Eternity clause / unamendable core | A small set of provisions placed beyond amendment | Art I, Art V (invariants) |
| M-05 | Quorum doctrine | Minimum participation for decisions to bind, with anti-deadlock relief | Art IV (base/escalated quorum, adaptive decay) |
| M-06 | Equal suffrage | One member, one vote; no capital or status weighting | Art II; Art V (invariant: equal suffrage) |
| M-07 | Codified exit | Legitimate, rule-governed secession/fork/ragequit with defined asset treatment | Art VI |
| M-08 | Mandatory expiry of delegated power | Terms, sunsets, and renewals instead of indefinite grants | Art VIII (auto-sunset); Art VII (role terms are required parameters) |
| M-09 | Sortition for interpretive bodies | Random selection to resist capture of referees | Art IX (sortition as default jury selection) |
| M-10 | Proposal/ratification separation | Whoever drafts cannot alone ratify; whoever ratifies cannot alone draft | Art III |
| M-11 | Single canonical versioned text | One consolidated authoritative document, not scattered conventions | Art I; Art X |
| M-12 | Hierarchy of norms | Ordinary rules reviewable against constitutional rules | Art I (supremacy ordering); Art VII (userland subordinate to kernel) |

---

## 3. Failure-mode taxonomy (F-codes)

Recurring failure modes, each with the kernel countermeasure. These codes also seed the adversarial test suite (Milestone 2): every F-code must have at least one scenario family.

| ID | Failure mode | Description | Kernel countermeasure |
|---|---|---|---|
| F-01 | Ossification | Amendment so hard that change routes through reinterpretation instead | Art III: calibrated tiers; patch-level changes deliberately cheap |
| F-02 | Entrenchment ratchet | Incumbents use a temporary majority to lower future checks on themselves | Art V: equal suffrage and amendment-process floors are invariant; Art III: changes to thresholds are always kernel-major |
| F-03 | Emergency-power ratchet | Emergency powers normalized, renewed indefinitely | Art VIII: hard auto-sunset, capped duration, rising renewal threshold, non-suspendable scopes |
| F-04 | Interpreter capture | The referee (court, ArbCom, council) packed or captured | Art IX: sortition default, short terms, appeal to full vote |
| F-05 | Plutocratic capture | Capital-weighted votes, vote buying, whale dominance | Art V: equal-suffrage invariant; Art II: personhood-based citizenship |
| F-06 | Participation collapse | Quorum failure produces deadlock, or apathy produces minority rule | Art IV: adaptive quorum decay with floor; Art X: public latency metrics |
| F-07 | Ambiguity exploit | Vague text or unwritten convention weaponized in bad faith | Art I: canonical text; Art IX: rulings recorded on ledger as precedent; test suite gates wording |
| F-08 | Majority-faction extraction | 51% drains the commons or suppresses a minority within the rules | Art V: worst-off (dignity-floor) invariant; Art VI: credible exit disciplines majorities |
| F-09 | Exit suppression | No legitimate fork path; dissent escalates to schism or capture | Art VI: fork right with defined notice, faction minimum ceiling, asset rules |
| F-10 | Sybil / onboarding attack | Membership flooded immediately before a vote | Art II: onboarding waiting period; Art III: suffrage snapshot at proposal time |
| F-11 | Key-person risk | BDFL/founder burnout, capture, or contested succession | Art VII: roles live in userland with mandatory term parameters; Art IX: no role is final interpreter of the kernel |
| F-12 | Process-speed mismatch | Governance slower than the attack (flash-loan votes) or so heavy it is abandoned | Art III: review periods as social timelocks; Art I: kernel minimalism caps process overhead |

---

## 4. System index — national / constitutional (N-codes)

Surveyed in detail in `01-national-constitutions.md`.

| ID | System | Period | Meta-rules extracted | Failure modes observed | Kernel clauses informed |
|---|---|---|---|---|---|
| N-01 | United States | 1789– | M-01, M-03, M-10, M-12 | F-01 (zero amendments in 50 yrs; change routed through courts), F-04 | Art III (thresholds set *below* US Art V difficulty), Art IX |
| N-02 | United Kingdom | uncodified | (absence of M-11 as the lesson); parliamentary sovereignty | F-07 (2019 prorogation; convention ambiguity), F-02 (no entrenchment of anything) | Art I (canonical text required), Art V (invariants exist because the UK has none) |
| N-03 | Germany (Basic Law) | 1949– | M-04 (Art 79(3) eternity clause), M-01 (⅔ both chambers), M-12 | F-04 (managed via staggered court terms — design lesson, not failure) | Art V (invariant set modeled on the eternity clause), Art IX |
| N-04 | Switzerland | 1848/1999– | M-03 (double majority: people + cantons), M-05, open citizen initiative | F-08 (initiatives targeting minorities, e.g., 2009 minaret ban — referendum without an invariant gate) | Art III (citizen proposal right **plus** Art V invariant gate on every amendment) |
| N-05 | France (Fifth Republic) | 1958– | M-01, dual amendment paths (Congress ⅗ or referendum) | F-03 (Art 16; 2015–17 état d'urgence normalization), F-02 (executive dominance) | Art VIII (auto-sunset, renewal thresholds) |
| N-06 | India | 1950– | M-01 (tiered amendment categories), M-12; basic-structure doctrine = M-04 improvised by judges | F-03 (1975–77 Emergency), F-04/F-01 interplay | Art V (explicit invariants beat judicially improvised ones), Art VIII |
| N-07 | South Africa | 1996– | M-01 (75% for founding provisions, ⅔ elsewhere), M-12 (court certified the text itself) | F-05 (state-capture era: patronage vs. institutions) | Art III (tiered thresholds), Art X (everything on a public ledger) |
| N-08 | New Zealand | uncodified+ | M-04 partial (entrenched electoral provisions only); constitution mostly ordinary statute | F-07 (convention reliance) | Art VII (substantive rules as cheap-to-change userland; only the kernel entrenched) |
| N-09 | Japan | 1947– | M-01 + M-03 (⅔ both houses + referendum) | F-01 extreme (zero amendments ever; Art 9 governed by reinterpretation) | Art III (threshold calibration), Art IX (interpretations recorded, versioned) |
| N-10 | Iceland (2011 crowdsourced draft) | 2010–13 | M-09-adjacent (randomly invited national forum), open drafting, M-11 | Pipeline without a binding ratification gate → text died in parliament (root cause of F-01-by-default) | Art III (end-to-end pipeline ends in a binding vote gate), Art X |
| N-11 | Chile (constitutional processes) | 2021–23 | Elected convention, ⅔ internal supermajority, parity rules | Two failed plebiscites: maximal substantive drafts overload ratification | Art I + Art VII (kernel minimalism; substance pushed to userland) |
| N-12 | Hungary | 2011– | (negative exemplar) supermajority alone, without invariants | F-02 exemplar (⅔ used to lower future checks), F-04 (court packing), cardinal-law entrenchment of policy | Art V (invariants bind even supermajorities), Art IX (sortition selection), Art III (entrenching *policy* at high tiers is structurally discouraged: substance is userland) |

---

## 5. System index — non-state (S-codes)

Surveyed in detail in `02-nonstate-systems.md`.

| ID | System | Type | Meta-rules extracted | Failure modes observed | Kernel clauses informed |
|---|---|---|---|---|---|
| S-01 | Debian Constitution | OSS project | M-06 (one developer one vote), M-01 (3:1 for constitutional changes), M-05, Condorcet ballots, M-10 (GR process) | F-06 (declining GR participation), F-11 (DPL burnout) | Art II, Art IV (voting method as a parameter, not kernel text) |
| S-02 | Apache Software Foundation | OSS foundation | Lazy consensus; justified-veto (consent-flavored decision rule); PMC layering | F-11, F-07 (veto-scope ambiguity) | Art IV (consent method option), Art IX (blocks must be reasoned and recorded) |
| S-03 | Python (PEP 13) | OSS project | M-08 (annual steering-council terms), M-10 (PEP author ≠ ratifier); BDFL→council succession | F-11 exemplar (2018 BDFL resignation under PEP 572 strain) | Art VII (roles are userland with mandatory term parameters; succession is a tested path, not a crisis) |
| S-04 | Rust (RFC process / governance council) | OSS project | M-02 (Final Comment Period), M-10, team delegation | F-07/F-11 (2021 mod-team resignation: no clear final authority; fixed by RFC 3392) | Art IX (an explicit final interpreter must exist), Art I (supremacy ordering) |
| S-05 | Wikipedia (English) | Commons community | Five pillars ≈ proto-invariants (M-04 analog), M-12-ish policy hierarchy, elected ArbCom | F-07 (wikilawyering), F-06 (shrinking admin corps), F-08 (RfC brigading) | Art V (small invariant set), Art II + Art III (suffrage snapshot vs. brigading) |
| S-06 | IETF | Standards body | Rough consensus + running code; NomCom random selection (M-09); layered appeal chain | F-07 ("rough consensus" is a judgment call), F-10 (open participation packing, managed socially) | Art IX (appeal chain), Art II (kernel requires countable membership precisely because IETF-style vagueness doesn't transplant) |
| S-07 | Mondragon | Worker co-op federation | M-06 (one worker one vote), general assembly, M-01 for bylaws, wage-ratio caps ≈ invariant | F-05 pressure (non-member subsidiary workers dilute the principle) | Art V (invariants as non-negotiable floors), Art II (citizenship boundary must be explicit) |
| S-08 | Haudenosaunee Great Law of Peace | Confederal (oral) | M-03 (consensus across nations), recall power held by clan mothers, M-02 deliberation norms | F-07 variant (oral transmission → divergent readings) | Art X (single canonical versioned text), Art III (repeal/recall symmetric with enactment) |
| S-09 | Robert's Rules of Order | Procedural canon | M-01 (thresholds tiered by motion class), M-05 (quorum doctrine), appealable rulings (M-12 analog) | F-07 (procedure weaponized), F-12 (motion fatigue) | Art IV (minimal procedural surface in kernel; complexity pushed to userland) |
| S-10 | MakerDAO | DAO (DeFi) | M-02 as timelock (Governance Security Module), emergency shutdown ≈ collective exit | F-05 exemplar (token-weighted whales), F-06 (low turnout), F-12 (flash-loan-era governance attacks in DeFi) | Art V (equal-suffrage invariant), Art III (review period as a social timelock) |
| S-11 | MolochDAO | DAO (grants) | M-07 exemplar (ragequit: codified exit with pro-rata assets) | F-10 (gated entry centralizes membership control), small-N collusion | Art VI (pro-rata exit default is directly Moloch-derived) |
| S-12 | Nouns DAO | DAO (treasury) | M-07 (fork mechanism added 2023 under raid pressure) | F-05/F-08 (buy-in raiders; 2023 fork drained a large treasury share — exit terms interact badly with capital-weighted entry) | Art VI (fork parameters: faction minimum, asset rules), Art V (equal suffrage removes the buy-in raid vector) |
| S-13 | Optimism Collective | DAO (bicameral) | M-03 (Token House + Citizens' House), iterated "Working Constitution" releases (M-11) | F-05 partial (token house side); experiment in progress | Art IV/VII (multi-gate composition in userland), Art X (versioned releases of the governing text — closest living analog to our model) |
| S-14 | Holacracy | Org-governance framework | M-11 (explicit written org constitution), governance/operations separation (M-12 analog), integrative consent process | F-12 exemplar (process overhead → abandonment, e.g., Medium 2016), F-07 | Art I (kernel minimalism: ~10 articles is a hard scope discipline learned here) |

---

## 6. Reverse cross-reference — kernel article → evidentiary base

| Kernel article | Primary systems informing it |
|---|---|
| Art I (Purpose, Supremacy) | N-02, N-11, S-04, S-14 |
| Art II (Citizenship & Suffrage) | S-01, S-06, S-07, S-05, F-10 cases |
| Art III (Amendment & Versioning) | N-01, N-04, N-09, N-10, N-07, S-04, S-10, S-13 |
| Art IV (Quorum & Voting) | S-01, S-02, S-09, S-10, N-04 |
| Art V (Invariants) | N-03, N-06, N-12, N-04, S-05, S-07, S-10, plus `04-moral-convergence.md` |
| Art VI (Right to Fork) | S-11, S-12, N-02 (negative: no exit doctrine), F-09 cases |
| Art VII (Userland Modules) | N-08, N-11, S-03, S-13, S-14 |
| Art VIII (Emergency Powers) | N-05, N-06, N-12, S-10 (emergency shutdown) |
| Art IX (Interpretation) | N-01, N-03, N-12, S-02, S-04, S-06, S-05 |
| Art X (Releases & Ledger) | N-10, S-08, S-13, N-07 |

Every kernel article traces to ≥3 surveyed systems. Every F-code traces to ≥1 system and ≥1 kernel countermeasure. Gaps found during audit should be filed as issues against this index, not silently patched.

---

## 7. Methodology and audit notes

1. **Source granularity.** Detailed claims, citations, and the reasoning for each extraction live in `01-national-constitutions.md` (N-codes) and `02-nonstate-systems.md` (S-codes); `03-synthesis.md` derives the taxonomies. This file is deliberately a flat, checkable map.
2. **Selection criteria.** Systems were chosen to maximize variance on four axes: scale (household-analog to nation), formality (oral → uncodified → codified), suffrage basis (person, worker, token, contributor), and observed stress (each system has at least one documented stress event or a documented design response to one).
3. **Negative exemplars are evidence.** N-12 (Hungary) and the F-05 DAO cases are surveyed as attack documentation, not as models. The kernel treats them the way a security team treats a published CVE.
4. **Test-suite linkage (forward reference).** Milestone 2's adversarial scenarios will carry F-code labels from §3 of this file. Coverage is defined as: every F-code has ≥1 failing-by-construction scenario against a known-bad wording and ≥1 passing scenario against kernel v0.1 text.