Which Metadata Standards Should a Digital Theatre Archive Use?

There is no single metadata standard that a digital theatre archive should use. The most robust approach is a layered stack: Dublin Core or Qualified Dublin Core (DCTERMS) for broad interoperability; the Dublin Core for Performing Arts (DCPA) when the collection needs performing-arts-specific description; ISAD(G) and EAD for archival hierarchy; METS and PREMIS for packaging and preservation; and controlled vocabularies, authority files and persistent identifiers for consistency. The right combination depends on what the archive holds, who uses it, and how it will be preserved and exchanged.

Why generic metadata is not enough

Theatre archives describe more than documents. A single recording may document a performance, which realises a production, which interprets a work; that work may be a translation or adaptation; the performance may have taken place in a festival, a tour or a co-production; and the recording may have different rights from the script. Generic item-level metadata struggles to hold those relationships. The literature on performing-arts-specific metadata is relatively recent; a 2017 study of the Edinburgh Festival Fringe noted that dedicated schemas were recent and had not been explored in that context (Holland, Overture and Beginners Please!, 2017).

The DARIAH Working Group Theatralia, established in 2021, identified the need for a shared computational language for describing and representing the performing arts so that data could be exchanged and generate scholarly value (DARIAH spotlight, January 2026). That is the practical case for domain-specific metadata: it makes relationships explicit rather than leaving them in prose.

The core standards and their roles

Dublin Core and DCTERMS: baseline interoperability

Dublin Core provides 15 standardised metadata fields for a simple but effective archival framework. Qualified Dublin Core, usually called DCTERMS, adds refinements and encoding schemes for greater flexibility and customisation (Dublin Core for Performing Arts, HAL, 2025). For a digital theatre archive, Dublin Core is useful as the minimum common denominator: it supports discovery across systems and makes it easier to export records to aggregators and union catalogues. DCTERMS is better than unqualified Dublin Core for dates, rights, relations and other details that need precision.

DCPA: a performing-arts model built on Dublin Core

The Dublin Core for Performing Arts (DCPA) is a digital description language model specifically designed for performing arts archives. It aims to establish a common framework for describing performing arts collections while adhering to the FAIR principles: findable, accessible, interoperable and reusable. It was developed by the THEATRALIA Working Group (DARIAH-ERIC) during its second phase, titled Performing Arts: Navigating through Digital Knowledge Models (2024–2025), building on Dublin Core and considering Qualified Dublin Core for greater flexibility (Dublin Core for Performing Arts, HAL, 2025).

The Theatralia project reports that Dublin Core was adopted as a flexible foundation and adapted through 38 domain-specific metadata fields, while a multilingual thesaurus prototype was developed using OpenTheso with around 1,400 standardised terms (Theatralia final technical report, Annex C, 2026). For an archive that needs to distinguish works, productions, performances, agents, venues and associated materials, DCPA is the most directly relevant model among the sources discussed here. Because it is recent, institutions should check the current published documentation and any local application profile before committing to implementation.

Archival hierarchy: ISAD(G) and EAD

If the archive is arranged as a collection with fonds (the whole body of records), series, files and items, ISAD(G) principles and Encoded Archival Description (EAD) remain useful for expressing that hierarchy and publishing finding aids. EAD is not a substitute for item-level performing-arts description; it is a structure for archival arrangement. A common approach is to use EAD or ISAD(G) for the collection context and Dublin Core, DCTERMS or DCPA for the description of individual objects and performances.

Digital objects and preservation: METS and PREMIS

METS can package structural metadata for digital objects, linking a recording, its parts, captions and related files. PREMIS is used for preservation metadata: fixity, events, agents and rights information needed to manage digital objects over time. These standards matter when an archive holds video, audio, born-digital scripts or complex multi-file objects. They answer questions that discovery metadata does not: what files make up the object, what has happened to them, and what must be done to keep them usable.

Controlled vocabularies, authorities and linked data

Names, genres, roles, techniques, venues and organisations should be controlled. Authority files record preferred and variant forms of names; controlled vocabularies make terms consistent; persistent identifiers allow records to be linked. The multilingual thesaurus prototype reported by Theatralia shows why this matters for cross-border performance: terms for roles, forms and techniques need consistent multilingual equivalents, not just ad hoc translation (Theatralia final technical report, Annex C, 2026). Where an archive needs to exchange event and object relationships across institutions, a linked-data ontology such as CIDOC CRM or Linked Art may be appropriate, but that is a design choice rather than a baseline requirement.

A practical stack by need

Documenting works, productions and performances

A workable data model separates entities that theatre practice often blends:

• Work: the play, score, libretto, devised concept or other creative entity, including original title, language and authorship.
• Translation or adaptation: the translator, adaptor, version, language and relationship to the source work.
• Production: the company, director, designers, production dates, venue and season or festival context.
• Performance: the individual date and time, cast and crew, language of performance, surtitle or caption language, and recording status.
• Item: the script, poster, programme, photograph, video, audio file or other object, with identifier, format, provenance, rights and preservation information.
• Agent and venue: people, companies, venues and festivals, recorded in authority files with variant names and persistent identifiers.

The relationships matter as much as the fields. A recording documents a performance; a performance realises a production; a production interprets a work; a translation is a version of a work. Recording those links in structured fields makes it possible to answer questions such as which productions of a translated play were performed in which languages, or which recordings exist for a particular tour.

Multilingual and cross-border metadata

For theatre across borders, metadata should distinguish the original title from translated, adapted and transliterated titles; the language of the source text from the language of performance; and the company’s country from the presentation venue’s country. Record translators, adaptors, dramaturgs and surtitle or caption makers as agents with defined roles. Use language tags and separate fields rather than a single free-text note such as translated into English. For touring productions, record the festival, season or venue context separately from the production’s origin, and keep rights and access conditions attached to the relevant recording or document rather than only to the collection as a whole.

Rights, consent and access

Rights metadata should be specific. A script, a translation, a filmed recording, a production photograph and a programme may have different rights holders and permissions. Record rights statements, access conditions, consent agreements and any restrictions at the appropriate level, and link them to the agents and dates involved. DCTERMS provides rights fields for discovery; PREMIS provides a more detailed preservation-oriented rights model. Neither replaces legal advice or a properly reviewed agreement. Where recordings involve audience members, contractors or community collaborators, consent and cultural context need to be documented alongside technical metadata.

Quality checks before implementation

• Model the archive’s entities and relationships before choosing a schema.
• Use Dublin Core as a baseline and DCTERMS refinements where precision is needed.
• Adopt DCPA, or a documented local profile of it, when performing-arts detail is required.
• Use ISAD(G) or EAD only where collection hierarchy and finding aids need them.
• Add METS and PREMIS for complex digital objects and long-term preservation.
• Control names, venues, genres and roles with authority files and vocabularies, recording the source and version of each term list.
• Record language, translation and rights relationships explicitly.
• Document mappings between local fields and any shared standard, and test exports to aggregators or preservation systems.

Conclusion

A digital theatre archive should not ask which single metadata standard will solve everything. It should ask which standards together can describe its objects, events, agents and rights with enough precision while remaining exchangeable. In practice, Dublin Core and DCTERMS provide the interoperability floor; DCPA provides performing-arts specificity; ISAD(G) and EAD provide archival structure; METS and PREMIS support digital objects and preservation; and controlled vocabularies and authority files provide consistency. The Theatralia work makes the domain case clearly: performing arts data needs a shared descriptive language, and a multilingual thesaurus is part of that infrastructure (Theatralia final technical report, Annex C, 2026). A small archive can start with a documented Dublin Core and DCTERMS profile, then add DCPA fields and controlled vocabularies as resources allow. A larger institution can integrate the full stack, but only if the local workflow and preservation plan make it sustainable.