ANSI/NISO Z39.19 (Controlled Vocabularies and Thesauri)

ANSI/NISO Z39.19-2005 (R2010) is the American National Standard for constructing, formatting, and maintaining controlled vocabularies and thesauri. Maintained by NISO, an ANSI-accredited standards developer, it defines three term relationships — equivalence (USE/UF), hierarchical (BT/NT), and associative (RT) — that keep retrieval consistent across authority lists, taxonomies, and full thesauri.

What the standard covers

Z39.19 addresses controlled vocabularies across a spectrum of structural complexity, from simple to most elaborate: authority lists (a single approved term for each concept, no relationships), synonym rings (sets of interchangeable terms with no preferred form), taxonomies (hierarchically arranged terms), and full thesauri (hierarchical arrangement plus explicit equivalence and associative relationships). For each, it specifies rules covering term selection and form (for example, preferring noun phrases, consistent singular/plural conventions, and using a parenthetical qualifier to disambiguate a homograph), how terms should be displayed, how a vocabulary should be tested before release, and how it should be governed and updated over time as usage and the underlying subject domain change.

The three relationship types

The best-known part of Z39.19, and the part most other thesaurus/vocabulary tooling has standardized around, is its definition of three relationship types between terms:

• Equivalence (USE / UF — “Used For”): links a non-preferred synonym or near-synonym to the single preferred term that should be used for indexing. A search or browse on the non-preferred term redirects the user to the preferred one.
• Hierarchical (BT / NT — Broader Term / Narrower Term): expresses genus-species, whole-part, or instance relationships, letting a user navigate from a general concept down to more specific ones (or back up).
• Associative (RT — Related Term): connects terms that are conceptually related in a way a searcher would find useful, but that are neither synonymous nor in a hierarchical relationship with each other.

These three relationship types, and the USE/UF, BT/NT, RT abbreviations themselves, are the vocabulary most thesaurus-management software and library/information-science curricula use as the default framework for describing how any controlled vocabulary’s terms relate to one another — Z39.19 is the standard those conventions trace back to.

Where Z39.19 fits among metadata and vocabulary standards

Z39.19 is a metadata schema-adjacent standard in a specific sense: it does not define a set of descriptive fields the way a schema like Dublin Core does, but it governs the controlled-term values that fill a schema’s subject or keyword fields. It sits alongside other NISO information-organization and data-documentation standards covered elsewhere in this dictionary, including domain-specific documentation formats such as DDI Codebook (survey and social-science dataset documentation) and PREMIS (digital preservation metadata), and web-native data-catalog description via the W3C’s DCAT (Data Catalog Vocabulary) — each solves a different part of describing and organizing data, while Z39.19 specifically governs the subject-indexing vocabulary layer. Internationally, the closely related ISO 25964 (Thesauri and interoperability with other vocabularies, parts 1-2) covers similar ground with some structural differences; content produced to one is not automatically interchangeable with the other without mapping.

Why it matters for research data management

Controlled vocabularies built to Z39.19 (or an equivalent framework) underpin subject indexing in library catalogs, institutional and disciplinary repositories, and government and scientific databases. For research data management specifically, a well-governed controlled vocabulary for subject tagging or keyword assignment — rather than free-text, uncontrolled tags — is part of what makes a dataset genuinely findable: search and retrieval work reliably only when the same concept is consistently represented by the same controlled term, with documented relationships to its synonyms and related concepts, rather than left to whatever free-text keyword each depositor happens to choose.

CASRAI and the NISO standards ecosystem

CASRAI is not a steward of Z39.19 itself. It is worth noting, for context, that CASRAI does have a direct working relationship with NISO elsewhere in the standards landscape: CASRAI has stewarded the CRediT (Contributor Roles Taxonomy) taxonomy since 2014, and CRediT was formalized as the full ANSI/NISO Z39.104-2022 standard, with NISO now holding primary stewardship and CASRAI operating as a federation partner. That relationship illustrates the kind of cross-organizational standards work NISO coordinates across the research-information landscape — Z39.19 is a separate, earlier NISO standard with no CASRAI stewardship role, included here because it is the foundational reference for how the controlled vocabularies referenced throughout this dictionary and elsewhere in research data management are properly constructed.

References

• NISO, ‘ANSI/NISO Z39.19-2005 (R2010): Guidelines for the Construction, Format, and Management of Monolingual Controlled Vocabularies’ (niso.org/publications/ansiniso-z3919-2005-r2010).