Subject - Controlled
Requirement Level: Recommended
Core Element: Yes
Describes: Original Source
Subject terms describing what the resource is about, including topics, personal/corporate names, work titles, geographic information, or time periods/cultures. This element should utilize terms from a formal, controlled vocabulary: applying structured term entries provides greater consistency in search results for users. Subjects also group related resources and enable exploratory searching/browsing. For free-text keywords or terms from a local vocabulary, use the Subject – Keywords element.
|Work Title||Bible—Comic books, strips, etc.|
Mappings and Encoding
subject (for subjects relating to topics or names)
coverage (for subject terms relating to time periods, geographic locations/coordinates)
See recommended mappings for additional standards.
See the Additional Detail section for more information.
When entering multiple subjects, it is recommended to store each subject entry’s value in a separate metadata element. See the Additional Detail section for more information.
Recommended Data Entry Type:
This element should only utilize entries from a controlled vocabulary. Use the Subject - Keywords element for free-text entries.
Applying subject headings may require additional metadata and/or subject matter expertise. If you have difficulty applying controlled subject terms, onsult with a metadata specialist for assistance, or consider using the Subject – Keywords field.
Choice of authority files used for subject values will depend on the needs of the specific repository and its content. Some recommended authority sources are listed below, based on major types of subject categories. These categories generally refer to the main part of the subject, as some authorities will typically have multiple divisions based on other categories (e.g., Topic—Geographic Division—Time Period).
Topical subjects are general topics or subjects of the work being described. Selected authority files include:
- FAST (Faceted Application of Subject Terminology) from OCLC
- Library of Congress Subject Headings (LCSH)
- MeSH (Medical Subject Headings) from the National Library of Medicine
Personal/Corporate/Conference Names as Subjects
Works that are about specific people, organizations, conferences and related entities should use authorized forms of names as subject entries. Authority files for personal, corporate, and conference names include:
- Library of Congress Name Authority File (LCNAF)
- Union List of Artist Names (ULAN)
- VIAF (aggregates multiple name authority files from different countries into a single source)
Work Titles as Subjects
The titles of works may be the subject of, or represented by, other works being described. Like other controlled vocabularies for personal/corporate names, a uniform title subject entry helps to connect resources about a given work title under a common entry. Recommended authorities include:
- Library of Congress Uniform Title Authority (within LCNAF)
- Cultural Objects Name Authority (CONA) (work titles in art, architecture, and archaeology)
Geographic subjects are places that a work is about or represents. In LCSH, geographic subjects may exist as primary subject terms or may be present as refinements added to another term. Recommended authority files include:
- Getty Thesaurus of Geographic Names (TGN)
- GeoNames (also integrated within FAST subject headings)
- VIAF (aggregates multiple authorities, including name authorities)
- LCSH (primary terms or refinements on other subject terms)
- LCNAF (generally includes place-names as corporate/government jurisdictions)
TGN and GeoNames are recommended for mapping/geocoding because they include geographic coordinates, elevation, natural features and landmarks in addition to place names.
Terms for time period and culture (with associated geographic context) are structured differently across different disciplines’ controlled vocabularies.
- LCSH: Time periods in LCSH are often appended to a topical or name entry, e.g. World War, 1914-1918
- Art & Architecture Thesaurus (AAT): includes a collection of terms specific to Styles and Periods for art and architecture, which may include time periods and/or geographic/cultural context, e.g. Old Kingdom (Egyptian).
When using a term from a controlled vocabulary, make sure to preserve the entry exactly as it is presented (do not change punctuation or capitalization).
If your system or schema supports it, it is recommended to add the following attributes to provide more information about your subject entries. This extra data enhances systems’ abilities to keep the controlled terms up to date:
- Vocabulary name or code
- Vocabulary URL
- Term/value ID
- Term/value URL
The MODS metadata schema provides granular encoding options for subject categories as sub-elements. Geographic subjects, for example, can be encoded either as a basic place-name string, as cartographic/coordinate data, or as a hierarchy of placenames.
When mapping subjects to simple (unqualified) Dublin Core, note that subject terms related to topics or names should be mapped to Dublin Core’s subject element, but subjects related to geographic location or time periods map to Dublin Core’s coverage element. The ability to map to Dublin Core’s coverage element will depend upon whether or not your data entry tool can distinguish between different types of subject entries. If your data entry tool cannot, map all entries to the Dublin Core subject element instead.
Some metadata standards (FGDC, ISO19115) and local Emory systems (Emory Theses and Dissertations) require a keyword or topic to be assigned.
Emory’s ETD system requires assignment of ProQuest Subject Categories (a proprietary set of categories) to all submissions.
For best system interoperability, it is recommended to store each subject term as a separate metadata element. When separate elements are used, it’s easier for systems to separate the subject values for searching, sorting, and maintenance purposes.
<subject>Term One; Term Two; Term Three</subject>
What if I’m unable to determine what the subject matter is?
In some situations it may not be possible to determine the exact subject matter of the content. This may occur if the content is in a foreign language not easily translated by the metadata creator, or for complex, subjectively interpreted content like abstract art.
What’s the difference between a name as a subject vs. a name entered as Creator or Contributor?
If you enter a name as a subject, it means your content is about the named entity as opposed to being created by that entity.
I’m listing multiple values into a spreadsheet (or other tool). Can I put all the entries in one field/column or should I separate them into separate fields/columns?
It is strongly recommended to split out subject values into separate elements. Otherwise, the subject terms can't be separately distinguished and parsed by systems and effectively become keywords only.
Links and Resources:
Page last modified: 2015-02-16