# SE-OPEN-CIV-001 — Structural Decision Boundary Protocol v0.3

Date: 2026-10-13
Status: PROPOSED_METHOD / NOT EMPIRICALLY VALIDATED
Scope: methodological stress test, NOT political monitoring, moral ranking, collapse forecasting or advice to preserve incumbent power.
Parent: https://structurevidence.org/cases/civilization-under-irreversible-time/

## 1. Research objective

Test whether an evidence-state workflow can reduce the cost of knowing which structural constraints were observable at a decision point, which were not, and whether any newly verified fact changes a feasible decision. It does not assume visibility changes behavior.

Primary question: Under bounded knowledge, irreversible time, conflicting interests and changing constraints, can a decision maker better distinguish (a) an unpleasant choice, (b) a deteriorating path, (c) a viable structural alternative, and (d) an already closed option?

## 2. Separate five claim classes

- OBSERVATION: a dated event, source, measurement or institution.
- INTERPRETATION: a mechanism inferred from observations.
- COUNTERFACTUAL: what could have happened under another action.
- NORMATIVE: whose interest ought to take priority.
- PREDICTION: a future outcome under specific conditions.

None may silently upgrade to another. Multiple agreeing interpretations are not independent observed facts.

## 3. Interests are actor- and time-indexed

Record actor_id, stated_goal, revealed_incentive (if evidence supports), authority_to_act, veto_power, affected_groups, current_benefits, future_costs, distribution, time_horizon and missing_interests. Never conflate the survival of the ruler, regime, state, population and cultural capabilities.

## 4. Three clocks and uncertainty

Define distributions or qualitative intervals, not fictitious precise numbers:

- T_fail: first time a named critical function falls below a defined threshold.
- T_effect: time until an intervention demonstrably improves that function.
- T_restore: time until depleted capacity is restored.

A negative estimated window for a particular intervention does NOT prove the system has no other viable intervention. Policy may change T_fail; external shocks can change every clock. Mark UNKNOWN when historical or present evidence cannot support calibration.

## 5. Adversarial hypotheses and disconfirmation

H1 (proposed): Among cases with severe structural stress, earlier recognition of constraints is associated with a wider feasible action set.
Against: accurate recognition may coexist with vetoes, violence, absent capacity, or unavoidable external shocks.
Disconfirming observation: independently verified early recognition with no usable action despite adequate decision capacity.

H2 (proposed): Adaptive institutional changes sometimes extend the time before critical-function failure.
Against: reform can undermine confidence, increase conflict or arrive too slowly.
Disconfirming observation: reform with demonstrably worse function outcomes after controlling as far as possible for contemporaneous shocks.

H3 (proposed): Buffer capacity can make reforms feasible under shocks.
Against: reserves carry costs, can be captured, and may be misallocated.
Disconfirming observation: comparable cases in which holding buffers materially worsened the overall viable action set.

H4 (proposed): A provenance-rich evidence process can identify when a previously reasonable claim becomes stale.
Against: evidence systems can overfit available public records or create performative certainty.
Disconfirming observation: material state changes routinely missed at the declared review scope or false reopen alerts not distinguished from verified changes.

H5 (proposed): Future-oriented rules can reduce irreversible harm without unjustifiably depriving present people.
Against: present elites can instrumentalize unborn persons' interests.
Disconfirming observation: rules presented as future protection that systematically shift disproportionate costs onto excluded present populations with no demonstrated risk reduction.

## 6. Comparison cohorts

C1 FAILURE: Late Ming / Chongzhen, 1627–1644 — multicausal stress; no claim of deterministic collapse.
C2 RECOVERY: Sweden banking crisis, 1990–1996 — bank resolution and political coordination; external setting and social costs retained.
C3 RECONSTITUTION: Denmark state bankruptcy, 1813, and monetary institutional change, 1818 onward — state and monetary system changed; no immediate full recovery asserted.

These are contrasting illustrations, not matched controls. A causal comparison requires specifying outcome functions, baseline social institutions, competing risks, source comparability and plausible alternatives. Do not score or rank political leaders. New comparators require a reasoned admission decision, not a search for examples confirming a thesis.

## 7. Minimum record per historical proposition

claim_id | case_id | actor | event_at | published_at | known_at (if demonstrable) | observation_cutoff | source | source_accessibility | exact_supported_statement | contrary_record | state | search_scope | limitations | confidence_basis | next_observable | reopen_trigger

Allowed state labels: SUPPORTED, OBSERVED, UNKNOWN, NOT_ESTABLISHED (bounded search only), CONTRADICTED (explicit sourced contradiction). Missing evidence is not nonexistence. A timestamp assigned retrospectively is not a demonstrated contemporaneous knowledge state.

## 8. Decision pressure and structural traps

For each candidate policy path identify:
1. Named critical function and minimum viable threshold;
2. Resources, financing and coercive authority needed;
3. Who pays now and who benefits later;
4. Implementation delay, coordination delay and recovery delay;
5. Vetoes and commitment/credibility constraints;
6. Accessible alternative paths;
7. Ex ante observable evidence versus ex post knowledge;
8. Potential irreversible damage and who bears it.

A structural trap is ESTABLISHED only if source-backed constraints rule out the defined action set in the defined time horizon; otherwise it remains a HYPOTHESIS. A single failed strategy cannot establish that all strategies were infeasible.

## 9. SE-specific validation gates

G0: QUESTION_FROZEN — exact hypothesis and subject stated.
G1: ACTOR_BOUNDARY — all materially affected actor categories disclosed or gaps recorded.
G2: SOURCE_PROVENANCE — dated retrievable references, access limits, rejected sources and review cutoff recorded.
G3: TEMPORAL_INTEGRITY — no hindsight facts masquerading as ex ante information.
G4: ALTERNATIVE_PATH — at least one countervailing mechanism and one case that could falsify the preferred interpretation.
G5: POWER_BOUNDARY — authority, veto rights, coercion and distribution explicitly mapped or marked UNKNOWN.
G6: CLAIM_SCOPE — observation, explanation, counterfactual, value judgment and prediction separated.
G7: OUTCOME_FEEDBACK — what new evidence would change the state is specified.
G8: PUBLICATION_BOUNDARY — no unwarranted causal proof or policy endorsement.
G9: SE_SELF-TEST — show a limitation or failure mode of StructEvidence itself.

Missing evidence for a gate => HOLD / UNKNOWN rather than invented PASS.

## 10. Comparative case notes / evidence anchors

Sweden: IMF reports 1992 guarantees and a Bank Support Authority, with restructuring and later exit from emergency guarantee measures. Sources:
- https://www.imf.org/external/pubs/ft/issues/issues14/index.htm
- https://www.elibrary.imf.org/view/journals/002/2009/247/article-A001-en.xml
- https://www.elibrary.imf.org/view/journals/002/2002/161/article-A001-en.xml

Denmark: Danmarks Nationalbank traces its origin to 1818 following wartime monetary instability and 1813 bankruptcy:
- https://www.nationalbanken.dk/en/about-us/danmarks-nationalbank-s-history-and-monetary-history
- https://www.nationalbanken.dk/en/about-us/the-law

Late Ming drought records (one evidence class, not proof of a singular collapse cause):
- https://cp.copernicus.org/articles/20/2287/2024/

## 11. SE value claim and honesty boundary

Potential value: lower cost of evidence acquisition, provenance review, temporal comparison and challenge tracking for decisions.
NOT ESTABLISHED: that users change decisions; that their outcomes improve; that SE can predict collapses; that all structural risks are observable; that access to evidence defeats incumbent incentives.
A genuinely negative result (e.g. insufficient historical record; no timely observable warning; no distinguishable decision path) is a successful research stop point and must remain publicly citable.

## 12. Change policy

Append only: do not retroactively overwrite the v0.1 essay or v0.2 adversarial addendum. Introduce new hypotheses with dated revisions, source changes and a disconfirmation ledger. No automated tracking of political events, external outreach, commercial claims, or deployment outside this research case is authorized.