How do you cite online museum and archive records reproducibly?

Why this question is difficult

A bare homepage link is not a reproducible citation. Collection records change, search results reorder, IIIF images may have separate rights, API responses can disappear, and an institutional label does not explain which field, page, image, or quotation supports a claim.

The Research Commons treats this as a method question rather than a shortcut to a historical conclusion. Work remains source-first, identity-aware, revision-bound, and open to correction. Every consequential result requires an attributable human reviewer, and publication remains a separate approval decision.

A reproducible method

Retain the provider, record title, stable record or archival identifier, exact URL, access date, cited field or page, quotation or faithful transcription, language, and the claim it supports. For APIs or permitted downloads, retain a content hash and original response separately from normalized research notes. Never mutate the retained source payload.

Create one machine-readable packet for the exact revision being reviewed. Preserve source URLs, visible queries, access dates, quotations or field locations, object-identity anchors, contradictions, rejected candidates, alternative hypotheses, gaps, and limitations. Hash the canonical packet so later corrections cannot silently alter the evidence reviewed.

Evidence and responsible automation

Distinguish the physical object, intellectual content, photograph, scan, metadata record, and API representation. Cite the specific layer used. Record rights and attribution separately for metadata, text, and images; public access does not imply republication permission. When a source has a persistent identifier, retain it alongside the current URL.

AI agents may assist with bounded query planning, retrieval after a human trigger, evidence extraction, citation checking, collision warnings, contradiction detection, and packet assembly. They may not establish identity, authenticity, title, wrongdoing, novelty, publication readiness, or legal conclusions. Missing evidence should produce a refusal or next research action, not invented completion.

Challenge the result before using it

Test the citation from a clean session and ask whether it resolves without private credentials, whether the quoted material is still present, whether pagination or filters are encoded, and whether the cited source actually supports the full claim. If access is restricted, document the finding aid and lawful access path without redistributing protected material.

A qualified reviewer should reproduce the most material citation, challenge identity, test the alternative hypothesis, identify source dependencies, and record corrections. The reviewer’s decision must name what is supported, disputed, unresolved, or outside scope. A polished narrative, repeated database statement, or high similarity score cannot substitute for this work.

Primary method references

• Linked Art data model — Event-centered cultural-heritage data semantics and multi-valued evidence structures.
• Getty Provenance Index — Official scope and access context for provenance, sales, dealer, and archival indexes.
• CIDOC Conceptual Reference Model — Reference model for events, actors, objects, places, timespans, and evidence-bearing relationships.

Questions and limits