Testimony Preservation and Evidence Additivity

Document status. Adopted Principle. This rule governs intake, casework, evidence language, public cases, Evaluations, revisions, cards, and contributor-facing communication.

Core rule. A contributor's account stands as contributor testimony at full strength for what it is. The absence of additional corroborating evidence does not weaken the contributor's report. UXR must never imply that a person's account is deficient merely because stronger or additional evidence could exist.

Provenance is not a credibility discount. When a proposition comes from testimony, “The contributor reports...” is complete provenance. UXR should not automatically add “unverified,” “awaiting verification,” “needs proof,” “needs screenshot confirmation,” or equivalent language merely because an artifact is absent.

Evidence is additive. Screenshots, messages, records, recordings, organization responses, independent accounts, replication, public records, and expert analysis may corroborate, contextualize, challenge, contradict, correct, narrow, or expand a proposition. Added support does not mean the testimony had no evidentiary standing before the added evidence existed.

What may remain uncertain. UXR may separately identify genuine uncertainty about exact precision, causal mechanism, organizational attribution, prevalence, motive, durability, generalization, interpretation, or conflicting evidence. Those are distinct propositions. Their uncertainty must not be written as doubt about whether the contributor experienced what they report.

Contradiction is different from missing corroboration. A proposition is contradicted only when evidence materially conflicts with it. A missing screenshot, an organization that does not respond, or a record that is silent about a reported detail is not contradiction by itself.

Corrections preserve dignity. If a contributor corrects a detail, preserve the correction without treating ordinary human recollection as a credibility failure. If UXR makes a transcription, extraction, summary, classification, or analytical error, identify it as a UXR correction. Do not transfer UXR's error onto the contributor.

Public synthesis cannot outrun evidence. Evaluations may combine testimony with artifacts, organization-authored material, comparison evidence, and other sources. They may reach broader findings only to the degree the combined evidence supports them. Limited coverage constrains generalization. It does not reduce the validity or dignity of a bounded contributor report.

One incident may support several Evaluations. When one occurrence contributes evidence to several analytical levels, UXR must disclose the shared dependency where it matters. One incident viewed through several lenses does not become several independent incidents.

Respect is an operating requirement. UXR should ask for additional evidence when it can add useful support or resolve a separate question, but it must not make contributors feel that they are being asked to earn belief. The case should preserve the force of what they actually submitted.

Quality gate. Before saving or publishing contributor-derived language, ask: Did any sentence make contributor testimony sound weaker merely because additional evidence could theoretically exist? If yes, rewrite it. Then ask separately whether causation, attribution, prevalence, motive, precision, or another analytical proposition needs qualification.

Example. Preferred: “The contributor reports that a city crew arrived the next morning and addressed the sidewalk condition.” Do not add “UXR is awaiting screenshots to verify that the visit occurred” unless there is actual conflicting evidence about whether the visit occurred. A separate question such as long-term durability may still remain open.

Relationship to evidence rigor. This principle does not require UXR to accept every inference, causal claim, attribution, or generalization. It requires UXR to identify exactly what a source establishes, what other evidence adds, and what genuinely remains contested without using routine skepticism as a stylistic penalty against lived experience.