Cataloguing is the process of building the intellectual record for each object: what it is, what it is made of, who made it, what it means, and how it connects to other objects and people. Where accessioning proves ownership, cataloguing creates the knowledge that makes a collection usable for research, exhibition and access.

Good catalogue records are structured (data in defined fields, not free-text soup), consistent (controlled terminology) and interoperable (able to be shared and aggregated). That is achieved through standards for data structure, data content and data value.

Data structure standards

These define which fields a record contains. For museum objects the leading choices are:

  • Spectrum - the underlying UK/international units of information for object records.
  • CIDOC CRM (ISO 21127) - an event-centred ontology for integrating heritage data across domains.
  • LIDO - an XML schema for harvesting and exchanging object records.
  • Dublin Core - a lightweight cross-domain set useful for aggregation and web exposure.

Controlled vocabulary and authorities

Consistency comes from controlled value standards rather than free typing. Widely used international authorities include the Getty vocabularies - AAT (Art & Architecture Thesaurus), TGN (geographic names) and ULAN (artist/agent names) - alongside Iconclass for subject classification. Recording persistent identifiers (Getty IDs, Wikidata Q-numbers) makes records linkable.

In practice. Separate the fact from its source and its certainty. A dating field should carry the date, the basis for it, and a confidence flag. Records that record "1650, attributed" as a single string cannot later be corrected or queried reliably.

Digital surrogates and images

Catalogue records increasingly bind to digital images and media. Deliver zoomable, shareable images through the IIIF APIs (Image and Presentation), and embed technical and rights metadata using standards such as PREMIS and Dublin Core so surrogates remain usable and correctly attributed over time.

Levels of cataloguing

Practitioners distinguish an inventory-level record (minimum data to account for an object) from a full catalogue record. Backlogs are managed by cataloguing to inventory level first, then enhancing - never leaving objects with no record at all.

Key takeaways

  • Use recognised structure standards (Spectrum, CIDOC CRM, LIDO, Dublin Core) so records are shareable.
  • Control your values with authorities like the Getty vocabularies and record persistent identifiers.
  • Bind records to IIIF-served images and capture certainty and source, not just facts.

Check your understanding

Last modified: Saturday, 1 August 2026, 10:37 AM