Collection is useful only when it is designed around the decision and governed by what each source can support.

Consequential questions rarely yield to one source class. The relevant record may include public filings, open-source material, commercial data and signals, client-provided material, prior work, and human-sourced context. These materials differ in authority, recency, completeness, and interpretive risk. Bringing them together can strengthen a decision view; flattening them into one undifferentiated file can weaken it.

Source governance preserves the relationship among information, origin, confidence, relevance, and use. It asks what kind of source something is, what it establishes, what it merely suggests, where its boundary lies, and how it should affect the decision.

Collection begins with a decision question

More material is not automatically better. Collection should begin with the decision to be supported, the uncertainties that matter, and the time and legal boundaries of the engagement.

That frame determines where to look and when to stop. A counterparty question may require corporate, regulatory, dispute, commercial, and reputational context. A legislative question may require formal procedure, stakeholder positions, issue history, and public signals. A litigation question may require pleadings, testimony, chronology, commercial exposure, and focused contextual inquiry.

The objective is not exhaustive capture. It is sufficient, relevant coverage of the factors capable of changing judgment.

Different sources do different work

A public record may be authoritative for a filing date but silent on practical capacity. A commercial signal may show activity or exposure while leaving causation uncertain. A news report may identify an issue that still requires corroboration. Client-provided context may reveal the operative decision frame but must remain distinguishable from independently sourced material. Human-sourced context may explain motive, constraint, and practical capacity; its basis, specificity, recency, incentives, and relationship to other sources determine how it should be weighted.

A source can be strong for one proposition and weak for another. It may establish that an event occurred without explaining why. It may show an actor’s formal position without revealing what the actor can do. It may support a timeline without supporting a recommendation.

Those distinctions should survive collection and ingestion. Material can be normalized enough to compare without losing the boundaries that make comparison honest.

Provenance and confidence keep analysis reviewable

Source governance separates observation, inference, and judgment.

Observation is what a source can fairly support. Inference is what may reasonably follow when the observation is read with other evidence. Judgment is the conclusion reached after weighing relevance, uncertainty, context, and decision consequence.

The transitions among those stages deserve attention. A polished conclusion can rest on a weak inference; a frequently repeated claim can still come from one uncertain origin. A current source may conflict with an authoritative but dated record. The process should make those conditions visible rather than smoothing them away.

Confidence is therefore not a stylistic tone. It is an assessment of how strongly the available source base supports a proposition and how vulnerable that proposition is to new evidence.

Integration should preserve difference

Multi-source work becomes valuable when separate materials can be related without being confused. Dates, actors, institutions, statements, transactions, relationships, and events may be placed into a common analytical structure while retaining source type, date, limits, and confidence.

That structure permits convergence and contradiction to be seen. Several independent sources may reinforce an assessment. A conflict may expose error, timing, bias, or a question worth developing. A missing source may itself define the next collection priority.

The public account of this discipline should remain high-level. Its value does not depend on disclosing tools, architecture, inventories, collection targets, or proprietary workflows. The essential point is simpler: heterogeneous material becomes useful when its provenance and limits remain accountable to the decision.

Source governance keeps collection honest. The next note turns to the source class most capable of explaining how the formal record operates in practice: human-sourced context.