# Meridian vocabulary review method

Fictional qualitative review, 13 October 2026. Identify a specific research question, select the relevant catalog term and compare the exact supplied sources. State what the sources do and do not establish, then propose a discriminating follow-up. The five selected questions have inconclusive outcomes; none establishes a capability, origin, authenticity or scientific consensus.

The observable definitions use the directly revised V2 baseline. Their characteristics are separate from a conclusion about mechanism, cause or origin. OCS attribution and each node’s testability flag belong to that proposed origin framework. A true flag is not independent demonstration that a specific hypothesis is testable or supported. Catalog membership does not establish a finding about this case.

A standard vocabulary binding identifies the catalog, immutable revision, exact catalog bytes and selected term. It separately identifies the exact claim-history snapshot and claim, with its own assignment author, time and rationale. The two observable bindings use `about`: they associate narrow research questions about physical acceleration and comparable sensor detection with broader terms. The three origin bindings use `assessment_criterion` to declare the specific origin hypothesis examined by the evaluator. No source description is rewritten to use the catalog’s wording.

The claim’s topic states the question; its subject identifies the reviewed observation. Exact edition citations and whole-file inputs preserve the supporting material. Catalog documents are vocabulary references, not observational inputs. A missing catalog assessment is absence of a claim, not an absence finding. No parent or child inherits a selected node’s assessment.

The source set is the sourcebook and original scalar captures retained by assessments-v1. Later instrument, laboratory and sky demonstrations are separate captures and are not review inputs. Authored reviews and this method remain supporting source documents; qualitative limits and follow-up questions are reading aids in reviews.json. This vocabulary packet is updated directly to the V2 baseline; it does not perform a migration of earlier consumer data.

The standard evaluator checks exact supplied document bytes, catalog identity, term membership, claim references and binding relationships. Missing documents remain unchecked; supplied contradictory documents fail. This build does not execute suggested scientific tests, authenticate the fictional sources or compute probabilities. All catalog entries remain navigable, including those without a Meridian review.
