# Structural Dynamics + Evidence Dynamics: An Auditable Framework for Structural Change and Public-Evidence Consistency ## Interim Method Paper v0.1 **Version:** v0.1 **Date:** 2026-09-08 **Research Status:** METHOD PILOT **First Empirical Object:** Strategy Inc. (NASDAQ: MSTR) **Research Framework:** RDL x RTP x ECN x Structural Dynamics x ECL **Disclaimer:** This method paper is not investment advice, trading authorization, legal judgment or an allegation of misconduct against any entity. ## Translation Provenance - Translation Status: Canonical English Translation - Original Language: Chinese - Original Artifact: `research/Structural_Dynamics_Evidence_Dynamics_Method_Paper_v0.1_CN.md` - Original SHA-256: `fe7c8f696cf14c3f5b3b470bc21d116c304b78d7bf60f949f043bfd335b4da8b` - Translation Date: 2026-09-09 - Scientific Recalculation: No - Scientific Conclusion Change: No - Translation Scope: public canonical English surface; frozen original remains unchanged. ## Abstract This paper separates two questions that are often compressed into one chain of judgment: what is changing in the real system, and whether the public evidence about that system can coexist after normalization. The framework permanently distinguishes captured artifacts, source authorship, truth of content, structural interpretation and publication permission. Structural Dynamics studies the structure, state and transition of a real system. Evidence Dynamics studies public claims, observations, revisions and conflicts around the same entity. The Evidence Consistency Layer (ECL) is the core analytical layer inside Evidence Dynamics. It does not automatically decide who is telling the truth. It asks a narrower and auditable question: can a frozen set of public evidence coexist after entity, time, unit, scope and source-dependency corrections? The first pilot applies the method to Strategy Inc. 2026 disclosures concerning Bitcoin as primary treasury reserve asset, BTC monetization, USD Reserve, capital raising, preferred obligations and software operations. The pilot result is `HYBRID_ACCUMULATION_MONETIZATION` for Structural Dynamics, `SEMANTIC_TENSION_BUT_RECONCILABLE` for Evidence Dynamics, `NO` material inconsistency and `PARTIAL` independent-source coverage. # 1. Research Question The framework is not an automatic fact adjudicator. It is research infrastructure for changing real structures and changing public evidence. ### Structural Dynamics Structural Dynamics asks how the real system is changing. It studies state variables, transition mechanisms, persistence, coupling, constraints and regime boundaries. ### Evidence Dynamics Evidence Dynamics asks whether the public evidence about the system is mutually reconcilable. Its chain is claim capture, observation extraction, normalization, source dependency, competing hypotheses, counter-evidence search, publication gate and provenance. # 2. Why Structural Research and Evidence Research Must Remain Separate A public data object raises at least five questions: whether it was captured, who published it, whether its content is accurate, whether it can enter a scientific structural model, and whether it may be redistributed. These cannot be collapsed into a single trusted/untrusted field. A later-false public statement can still be a real historical event. Evidence Dynamics can therefore study how public narratives change, which claims are revised, removed or superseded, whether sources conflict, and whether an apparent conflict is a fact conflict, time difference, scope difference or structural change. # 3. ECL Data Objects ## 3.1 Raw Artifact A Raw Artifact answers whether a specific material object was captured or identified. It does not prove content truth. ## 3.2 Claim A Claim is a statement made by a source about an object. It must preserve speaker, time, scope, wording, unit and source dependency. The model must not rewrite a bounded claim into a stronger one. ## 3.3 Observation An Observation is a standardized observable fact or value. Claims and observations can be connected, but they are not the same object. # 4. Provenance and Freeze ECL inherits RTP provenance principles. Every material artifact should preserve source identity, locator, capture type, hash where available, timestamp and publication boundary. The first pilot showed that a file extension alone does not prove that a complete original webpage was captured. Short Strategy Investor Relations captures were therefore classified as `VERIFIED_EXTRACT`, not `FULL_RAW`. # 5. Source Independence ECL does not treat URL count as independent evidence count. Source independence is issue-specific. SEC issuer filings and Strategy IR materials have high formality and provenance, but on the question of how the entity describes its own treasury policy they are not fully independent reality-validation chains. The pilot therefore keeps `PARTIAL` independent-source coverage. # 6. Five Classes of Consistency Analysis ## 6.1 Temporal Consistency Two statements must be compared under the same time condition. Later policy changes do not automatically invalidate earlier statements. ## 6.2 Numerical Consistency Public numbers are checked with replayable formulas. Strategy pilot examples include `1,363 BTC x $59,256 ~= $80.8M`, `2,225 BTC x $60,773 ~= $135.2M` and `846,000 - 2,225 = 843,775 BTC`. Where aggregate cost-basis allocation data is missing, the system returns `INSUFFICIENT_DATA`. ## 6.3 Semantic Consistency Semantic analysis asks whether two claims actually conflict. In the Strategy pilot, "primary treasury reserve asset" is not semantically equivalent to "never sell"; therefore BTC sales alone do not prove abandonment of the Bitcoin treasury policy. ## 6.4 Cross-Source Consistency Cross-source analysis compares sources and methods only after source dependency is addressed. ## 6.5 Structural Consistency Structural consistency asks whether the public narrative is compatible with broader structural change. It is not an integrity judgment. # 7. Triangulation ECL uses triangulation as a discipline, not as a mechanical three-source proof. It considers sources, methods, analysts and perspectives. Independence is issue-specific. # 8. ACH: Competing Hypotheses, Not Support Hunting The Strategy pilot froze six explanations: | ID | Hypothesis | |---|---| | H1 | Coherent Treasury Evolution | | H2 | Structural Shift to Hybrid Treasury Management | | H3 | Cash-Obligation Pressure | | H4 | Financing Flexibility Dominates | | H5 | Semantic / Narrative Lag | | H6 | Apparent Conflict Caused by Scope / Timing | ACH does not choose a winner by counting supporting evidence. It asks which hypothesis faces the most serious counter-evidence. Allowed statuses include `VIABLE`, `WEAKENED`, `REJECTED`, `UNRESOLVED` and related bounded labels. There is no final-truth hypothesis. # 9. Observed Contradiction and Potential Falsifier The pilot corrected an important method error: future conditions that could overturn a hypothesis must not be placed into current `contradicting_evidence`. A potential future formal abandonment of Bitcoin reserve policy is a falsifier condition, not observed counter-evidence. # 10. Counter-Evidence Search For every materially used hypothesis, ECL asks what evidence would make the hypothesis fail. The purpose is not to prove another hypothesis but to prevent the system from searching only for supporting material. # 11. Conflict Taxonomy ECL does not use a universal truth score. Conflicts are classified with bounded states such as temporal change, semantic tension, potential conflict, material inconsistency or insufficient data. `MATERIAL_INCONSISTENCY` must never be automatically translated into fraud, deception, insolvency or investment quality. # 12. First Method Pilot: Strategy Inc. Research ID: `ECL.COMPANY.STRATEGY_INC.2026.001`. The observation window primarily covers 2025-12-31 to 2026-08-31. The pilot does not study MSTR stock price or whether Bitcoin is a good investment. It asks whether Strategy's 2026 public disclosures concerning Bitcoin reserve policy, BTC monetization, USD Reserve, capital raising, preferred obligations and software operations can be reconciled across time, numerical, semantic and structural dimensions. # 13. Pilot Result Structural Dynamics produced `HYBRID_ACCUMULATION_MONETIZATION`. Evidence Dynamics produced `SEMANTIC_TENSION_BUT_RECONCILABLE`, with material inconsistency `NO` and independent-source coverage `PARTIAL`. The result means BTC reserve, BTC monetization, USD liquidity buffer, common/preferred capital raising, fixed cash obligations and software operations enter the same capital structure. It does not mean the company abandoned accumulation, entered a liquidity crisis, made a false narrative, or created a trading signal. # 14. Why "No Material Inconsistency Found" Is Valuable The negative result is methodologically meaningful. ECL did not manufacture a problem. After time, scope, definition, number and competing-hypothesis correction, it downgraded an apparent conflict into reconcilable semantic tension. # 15. Relationship to Structural Dynamics Structural Dynamics answers what changed in the structure. Public Evidence Research answers whether the public claims and observations about that structure can coexist. A significant structural change does not automatically mean public disclosure contains a material contradiction. # 16. Limitations ### 16.1 Independent-source coverage partial The pilot relies heavily on issuer filings and issuer-authored investor materials. This is disclosed as a source-dependency boundary. ### 16.2 ECL is not a legal judgment system ECL does not establish fraud, liability or legal breach. ### 16.3 Semantic analysis still requires human review Boundary cases require human reading of scope, time and wording. ### 16.4 A single pilot cannot prove cross-domain generalization The method remains a pilot and should not be overgeneralized. # 17. Next Research Questions Future work should test portability across assets, protocols, treasuries and institutions, improve source-independence handling, and preserve explicit boundaries between observations, tool outputs and published findings. # 18. Interim Conclusion The method establishes an auditable way to separate structural change from evidence consistency. It preserves provenance, uses competing hypotheses, searches for counter-evidence and refuses to upgrade uncertainty into accusation or investment usefulness. # Appendix A: Core Frozen Results of the First Pilot - Research ID: `ECL.COMPANY.STRATEGY_INC.2026.001` - Structural State: `HYBRID_ACCUMULATION_MONETIZATION` - Evidence State: `SEMANTIC_TENSION_BUT_RECONCILABLE` - Material Inconsistency: `NO` - Independent-source Coverage: `PARTIAL` - Research Status: `METHOD PILOT` # Appendix B: Research Evidence Basis Complete provenance, hashes, capture types, claim registries and reconciliation records remain in the frozen S6/S6.1/S6.1a evidence package and publication manifest. ## Numeric, Enum and Source Integrity Register The following material tokens are carried forward from the original artifact and must remain unchanged in interpretation: - `0.1` - `2026` - `09` - `08` - `1` - `2` - `3` - `4` - `5` - `3.1` - `3.2` - `3.3` - `S6.1a` - `6` - `6.1` - `6.2` - `1,363 BTC` - `$59,256` - `$80.8M` - `2,225 BTC` - `$60,773` - `$135.2M` - `846,000 BTC` - `843,775 BTC` - `6.3` - `6.4` - `6.5` - `7` - `8` - `H1` - `H2` - `H3` - `H4` - `H5` - `H6` - `9` - `10` - `11` - `12` - `2026.001` - `2025` - `31` - `13` - `14` - `15` - `16` - `16.1` - `16.2` - `16.3` - `16.4` - `17` - `18` - `07` - `06` - `03` - `256` Canonical states preserved: `HYBRID_ACCUMULATION_MONETIZATION`, `SEMANTIC_TENSION_BUT_RECONCILABLE`, `PARTIAL`, `NO`, `METHOD PILOT`. No source URLs, source IDs, formulas, dates or report IDs are intentionally changed by this translation.