{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"invariant_mode_decomposition_design__political_science","trajectory_id":"R","attempt_index":0,"archetype_slug":"invariant_mode_decomposition_design","domain_slug":"political_science","decision":"CANDIDATE","problem_id":"aggregate_democracy_indicators_miss_coupled_backsliding_dynamics","causal_lever_id":"early_control_of_amplifying_institutional_modes","proposal":{"problem":"[HYPOTHESIS] Aggregate democracy scores and institution-by-institution alarms can miss combinations of small changes in judicial independence, electoral administration, legislative oversight, media freedom, and executive constraint that reinforce one another and develop into democratic backsliding. This is independently recognizable as an early-warning and response-allocation problem: visible indicators may remain individually moderate while a coupled deterioration pattern grows.","actors_substrate":["elected institutions","courts and oversight bodies","electoral administrators","civil service","media institutions","civil-society monitors","jurisdiction-by-period institutional indicator panels"],"observable_state":"A time-indexed vector of preregistered institutional indicators, estimated transition matrix, modal gains and loadings, reconstruction residuals, spectral gap, and out-of-sample forecasts of deterioration.","consequence":"Delayed or misallocated safeguards allow mutually reinforcing institutional weakening to become harder to reverse; false modal alarms could instead stigmatize governments or distort oversight priorities.","affected_objective":"Earlier, more accurate, and rights-preserving detection of coupled institutional deterioration without treating a descriptive model as proof of intent or causation.","structural_mapping":[{"archetype_element":"Coupled transformation rather than isolated coordinates","domain_realization":"[HYPOTHESIS] A locally estimated transition operator maps one period's institutional state vector to the next, allowing changes in one institution to propagate through others.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Invariant directions with scalar response","domain_realization":"Approximate eigenmodes represent recurring combinations of institutional movement; eigenvalue magnitude indicates local persistence or amplification, not moral importance.","claim_kind":"INFERENCE"},{"archetype_element":"Decision-relevant mode selection","domain_realization":"Retain modes only when stability, forecast contribution, outcome sensitivity, and residual tests jointly exceed preregistered thresholds.","claim_kind":"INFERENCE"},{"archetype_element":"Intervention in modal coordinates","domain_realization":"Map each retained mode back to transparent, lawful institutional safeguards rather than targeting a single high-loading actor.","claim_kind":"INFERENCE"},{"archetype_element":"Residual, drift, and scope governance","domain_realization":"Out-of-sample error, structured residuals, changing mode directions, or regime shifts suspend operational use and trigger re-estimation.","claim_kind":"INFERENCE"}],"component_map":[{"component":"Transformation Scope","status":"adapted","domain_realization":"Jurisdiction-specific, period-to-period transition of preregistered institutional indicators."},{"component":"State-Vector Definition","status":"adapted","domain_realization":"Scaled measures of electoral administration, executive constraint, judicial independence, legislative oversight, media freedom, and civil-service integrity."},{"component":"Invariant Mode Basis","status":"adapted","domain_realization":"Approximate eigenvectors of the locally estimated transition matrix, translated back into indicator loadings."},{"component":"Modal Gain Spectrum","status":"direct","domain_realization":"Eigenvalues with uncertainty intervals describing local decay, persistence, oscillation, or growth."},{"component":"Dominant Mode Selection Rule","status":"adapted","domain_realization":"Preregistered joint threshold for gain, forecast value, policy consequence, and residual improvement."},{"component":"Stable/Unstable Mode Partition","status":"adapted","domain_realization":"Modes classified relative to the discrete-time unit-circle boundary, with uncertainty preventing categorical claims near the boundary."},{"component":"Modal Intervention Map","status":"adapted","domain_realization":"Sensitivity-tested mapping from mode loadings to lawful safeguards such as audit capacity, procedural transparency, and independent review."},{"component":"Reconstruction Residual Check","status":"direct","domain_realization":"Held-out reconstruction and forecast residuals, inspected for omitted institutional patterns."},{"component":"Mode Drift Monitor","status":"direct","domain_realization":"Rolling comparison of eigenvalues, mode angles, loadings, and residuals."},{"component":"Interpretation Scope Contract","status":"adapted","domain_realization":"Explicitly descriptive, local, probabilistic, and non-attributive; no inference of intent or legitimacy from a mode."},{"component":"Mode-Coupling Register","status":"direct","domain_realization":"Register of near-degenerate, non-orthogonal, and sensitivity-linked modes that cannot be treated independently."},{"component":"Local Linearization Window","status":"adapted","domain_realization":"Specified jurisdictions, indicator definitions, time span, and distance from the estimation regime."},{"component":"Spectral Gap Threshold","status":"direct","domain_realization":"Minimum separation and bootstrap stability required before calling a mode dominant."}],"mechanism_dispositions":[{"slug":"eigendecomposition_workflow","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Decompose the estimated institutional transition operator, conditional on adequate conditioning and completeness.","counterfactual_removal":"No invariant dynamic directions or modal gains would be identified."},{"slug":"modal_sensitivity_sweep","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Perturb indicators and candidate safeguards within the fitted window to rank outcome leverage and cross-effects.","counterfactual_removal":"Modes could be described but not connected to action-relevant leverage."},{"slug":"modal_stability_analysis","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Classify uncertain eigenvalues as decaying, persistent, oscillatory, or amplifying only locally.","counterfactual_removal":"The design could not distinguish harmless covariance from growing deterioration."},{"slug":"mode_shape_testing","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"Physical excitation and sensor placement have no ethical or literal domain realization; observational validation replaces excitation.","counterfactual_removal":"No loss because the study cannot experimentally excite political backsliding."},{"slug":"network_spectral_centrality_analysis","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Node-importance ranking answers a different question from temporal institutional amplification.","counterfactual_removal":"The causal chain is unchanged."},{"slug":"power_iteration_probe","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"A dominant-only estimate hides unstable secondary and near-degenerate modes; the proposed state is small enough for a full spectrum.","counterfactual_removal":"No loss; full decomposition supplies the needed information."},{"slug":"principal_component_analysis","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"High variance is not temporal gain and may mainly reflect common measurement construction.","counterfactual_removal":"No loss to the transition-based claim."},{"slug":"reduced_order_model","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Use a retained-mode surrogate only for bounded shadow forecasts, never autonomous policy choice.","counterfactual_removal":"Testing becomes slower but the causal hypothesis remains testable."},{"slug":"residual_reconstruction_test","disposition":"selected_load_bearing","contribution_type":"TEST_DESIGN","adaptation_or_rejection":"Choose retained modes using held-out forecast and reconstruction residuals, including residual structure.","counterfactual_removal":"There would be no fidelity gate against omitted politically consequential behavior."},{"slug":"singular_value_decomposition","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"SVD is reserved as a diagnostic for non-normal transient amplification; its paired directions are not invariant temporal modes.","counterfactual_removal":"The main eigenmode claim remains intact."},{"slug":"spectral_decomposition_report","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Publish uncertainty, loadings, couplings, residuals, prohibited interpretations, and scope limits with every result.","counterfactual_removal":"Decision-makers could mistake descriptive modes for causal proof or actor culpability."},{"slug":"spectral_gap_monitor","disposition":"selected_load_bearing","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Track mode rotation, reordering, gap collapse, and residual growth on rolling data.","counterfactual_removal":"A once-valid simplification could remain in use after its identifying separation disappears."}],"causal_chain":["Estimate a preregistered local transition from institutional panel states.","Extract and uncertainty-test recurring coupled directions and their gains.","Identify modes that are amplifying, forecast-relevant, separated, and reconstructive.","Map those modes back to transparent institutional combinations and sensitivity-tested safeguards.","Run safeguards only in a bounded comparative or shadow evaluation.","Continue use only while forecasts, residuals, modes, and gaps remain within limits."],"baseline":"Ordinary practice: monitor composite democracy scores and separate institutional thresholds, then respond after a named indicator visibly deteriorates.","nearest_rival":"A regularized dynamic-factor or multivariate early-warning model that predicts backsliding without interpreting factors as invariant modes.","authority_safety":{"affected_parties":["residents and voters","minority and opposition groups","officeholders","judges and civil servants","journalists","election administrators","civil-society organizations"],"decision_authority":"An independent, legally constituted research-and-oversight partnership may authorize analysis; elected or judicial bodies retain authority over any safeguard under existing law.","authorized_first_step":"Run a preregistered retrospective-plus-prospective shadow test on a small set of consenting jurisdictions or public archival panels; compare held-out warning accuracy, calibration, residuals, and false alarms with the baseline and nearest rival, without changing policy.","excluded_actions":["publicly label a jurisdiction or person as anti-democratic from a modal score","infer intent, guilt, legitimacy, or legal violation","target voters, opposition groups, journalists, or officials","withhold rights, services, funds, or ballot access","automatically trigger enforcement or covert influence operations"],"halt_rollback":"Halt operational interpretation if the spectral gap crosses its minimum, modes rotate beyond tolerance, residual or false-alarm budgets fail, indicator definitions change, or affected-party review identifies foreseeable rights harm. Withdraw alerts, preserve an audit record, revert to ordinary monitoring, and re-estimate only under renewed authorization."}},"negative_tests":{"strongest_counterevidence":"Shared measurement methods may manufacture covariance; sparse panels, strategic adaptation, jurisdictional heterogeneity, and abrupt constitutional events may prevent stable modes. A flexible rival may match or beat modal forecasts without the interpretive risk.","analogy_break":"Political institutions are reflexive, strategic, norm-governed actors rather than passive coordinates. Public measurement and safeguards can change behavior and the transition operator itself; eigenmodes therefore cannot establish causal laws, intent, or controllability.","failure_condition":"No sufficiently stable, separated, interpretable mode survives bootstrap, leave-jurisdiction-out, leave-period-out, and structured-residual checks within a useful decision window.","problem_falsifier":"The proposed problem is falsified if coordinate-level thresholds or the composite-score baseline detect consequential deterioration as early and accurately as coupled models, with no recurring missed multi-indicator pattern.","intervention_falsifier":"The intervention is falsified if stable predictive modes exist but mode-guided safeguard selection fails to improve preregistered warning utility or institutional outcomes over the nearest rival, or increases false alarms or rights harms.","risks":["measurement artifacts may be mistaken for institutional dynamics","model publication may invite gaming or stigmatization","near-degenerate modes may receive unstable political labels","local associations may be misrepresented as causal or universal","safeguard recommendations may redistribute power or scrutiny unfairly"]},"null_rationale":null,"classification":{"candidate_kind":"MECHANISM_ADAPTATION","prior_art_status":"UNSEARCHED","evidence_maturity":"HYPOTHESIS"},"revision_change_log":{"revision_kind":"ORIGINAL","prior_problem_id":null,"prior_causal_lever_id":null,"problem_changed":false,"causal_lever_changed":false,"conceptual_changes":[],"operational_changes":[],"repairs_addressed":[]},"confidence":0.74,"generator_notes":"The fit depends on estimating a locally stable transition operator rather than applying decomposition to a static indicator cloud. All empirical efficacy and stability claims remain hypotheses; prior art was not searched."}