*Replication Project* (credibility / falsification)

“If X is true, then Y must fail under Z.” This guide specifies the minimal steps required to independently test whether persistent constraints on agency reduce recoverability over time, as stated by SDTL. It is designed to be observer-invariant, domain-agnostic, and non-interpretive. 1. Scope and Applicability This replication procedure applies to regulated systems where: agents or components can maneuver, constraints can persist over time, and recoverability can be observed or proxied. Examples: markets institutions conflicts organizations social systems historical cases (via proxy evidence) This guide does not apply to: purely random systems, unregulated physical phenomena (e.g., earthquakes), systems without agency or maneuverability. 2. Required Inputs Replicators must provide: A. Agency Constraint Proxy Any observable indicator that reflects restriction of maneuverability, such as: volatility regimes, regulatory burden, coercion levels, institutional rigidity, loss of optionality. The proxy does not need to be metric; ordinal suffices. B. Time Series A chronological sequence of observations with: consistent resolution (daily, monthly, yearly), no requirement for high frequency, clearly defined time index. C. Recoverability Proxy An observable measure of a system’s ability to return to prior states, such as: time to recover peak capacity, drawdown magnitude, loss of functional diversity, duration of impaired operation. 3. IMI Scoring (Ordinal) Each observation is assigned an Intrinsic Maneuverability Index (IMI) score: Guidance – high flexibility, voluntary coordination Influence – incentives dominate Indifference – signals ignored, drift begins Manipulation – asymmetric pressure emerges Obstruction – options blocked Compel – forced behavior, minimal choice Scoring must be: ordinal only (no interpolation), applied consistently, justified by observable evidence. 4. Persistence Detection Define a persistence window � appropriate to the system’s clock speed. Compute: proportion of time IMI ≥ Manipulation within window �. Persistence is established when: IMI ≥ Manipulation is sustained beyond noise thresholds. 5. Trajectory and Recoverability Test For each persistence onset: Mark T₁ (drift onset) Mark T₂ (conditioning / sustained constraint) Measure recoverability proxy forward in time Expected outcome under SDTL: 6. Optional Aggregation (CPT) If multiple systems or lenses are available: compute local trajectories independently, normalize trajectories, aggregate to assess alignment. CPT is derived, not required for single-system replication. 7. Falsification Criteria SDTL is falsified if replicators observe: sustained high constraint with no loss of recoverability, or improved recoverability under persistent constraint, or random timing unrelated to constraint persistence. Negative results are valid and should be reported. 8. Reporting Requirements A valid replication report includes: system description, proxies used, IMI scoring rationale, persistence window, recoverability results, any deviations from this guide. No interpretation beyond results is required. 9. Interpretation Boundary This guide: does not infer intent, does not assign moral judgment, does not predict specific events. It tests a necessary condition, not a sufficient cause. Replication Status Independent replication encouraged Method invariant under refinement Law unchanged by measurement improvements One-line summary (for OSF) This guide specifies how to empirically test whether persistent constraints on agency reduce system recoverability, independent of domain or interpretation. Any documented counterexample that violates the constraint should be reported as a falsification attempt.