Skip to content
An independent guide to international theatreMany cultures. A shared stage.
The InternationalistsTHEATRE ACROSS BORDERS

A GUIDE TO THEATRE ACROSS BORDERS

Translation & Multilingual Theatre

How to Document Language Choices in Theatre Production Archives

THE SHORT ANSWER

Document language choices by separating the languages of the materials from the language of archival description, then recording each language and script with its role: source text, translation, adaptation, rehearsal, performance, surtitle, access or programming. Use controlled codes such as ISO 639 and ISO 15924, name the people and versions involved, and keep the evidence linkable. A single “multilingual” label is not enough for future restaging, translation or research.

At a glance Quick facts
Core archival distinction
Language of the material versus language of archival description.
Required description element
DACS 4.5 treats languages and scripts of the material as a required element.
Language and script codes
ISO 639 for languages and ISO 15924 for scripts, as cited in DACS 4.5.
Description-language rule
IASA 0.E prefers the language and script in which information is presented for specified areas, with fallbacks.
Digital archive priorities
Data reusability, findability and interoperability, as discussed in a 2018 article on performance documentation.
Performing-arts modelling
A 2026 CIDOC CRM SIG document discusses a Linked Art extension proposal for the performing arts.

Documenting language choices in a theatre production archive means more than noting that a production was “multilingual”. It means recording which languages and scripts appear in the materials, what role each plays, which versions circulated, who made translation and adaptation decisions, and which language the archive itself uses to describe the material. The distinction between the language of the material and the language of description is basic but often lost. DACS 4.5 treats languages and scripts of the material as a required element because they affect use, and it permits ISO 639 and ISO 15924 codes for machine processing (DACS 4.5). IASA 0.E similarly asks cataloguers to give information, in specified areas, in the language and script in which it is presented wherever possible, with defined fallbacks (IASA 0.E).

For a production archive, the practical answer is a language register: a structured record that links each language and script to a document, version, production event or access layer, together with a named role and an evidence trail. Record the source text and its language; the translation or adaptation and its language; the rehearsal language or languages; the performance language or languages; any surtitle, caption, sign-language or audio-description layer; the language of programmes and press materials; and the language of the archival description itself. Then describe the choice, not just the language.

Why language choices are archival facts

A production can use one language onstage and another in the rehearsal room, a third in the surtitle file, and a fourth in the funding application. These are not interchangeable. They have different creators, rights, audiences and evidence. The archive record needs to distinguish them. The distinction matters for restaging, translation history, accessibility, attribution of translators and dramaturgs, and research into multilingual theatre.

The source text is not the production text. A literal translation, a singing translation, a devised adaptation, a surtitle translation and a rehearsal bridge translation can all exist at once. If the archive stores only the performed script, future users may not know which text was translated, by whom, from which edition, under what permission, or whether the translation was published or commissioned for the production. Record version identifiers, dates and file relationships.

Build a language register before opening the archive box

Use a simple controlled schema. A spreadsheet can be enough at first, provided it is consistent. At minimum, record: language name; ISO 639 code; script name and ISO 15924 code where relevant; document or event; role; status or extent; source of information; rights or consent note; and link to the digital file or physical item. DACS 4.5 supports deriving language and script information from the materials themselves and notes that codes may be used for machine processing (DACS 4.5).

LayerWhat to recordExample of roleEvidence to keep
Source textLanguage and script of the edition used for translationSource language for adaptationEdition, publisher, date, translation brief
Translation or adaptationTarget language or languages, including hybrid or devised textPerformed translation, literal translation, singing translationTranslator or adaptor, version, permission record
RehearsalWorking languages of the room and any interpreting arrangementsRehearsal language, bridge languageCall sheets, notes, queries, interpreter credits
PerformanceSpoken, sung or signed languages, including code-switchingPerformance language by scene or roleScript version, run order, recordings
Surtitles and captionsSource and display languages, cue files and timingProjected translation or access layerSurtitle file, operator notes, translation version
Archival descriptionLanguage and script of the catalogue record and scope noteLanguage of descriptionRepository policy, description date

Do not let the register collapse language, dialect, accent and script into one field. DACS speaks of languages, scripts and symbol systems; IASA speaks of language and script; these are related but separate. A language may be written in more than one script. A production may use a script for visual design without spoken language. Record what you can evidence, and mark uncertainty rather than guessing.

Document decisions, credits and rights

Language choices are creative and political decisions. A director may choose to keep a repeated phrase in the source language; a translator may decide to leave a pun untranslated; a dramaturg may commission a bridge translation for rehearsal; a producer may require surtitles in a festival language. The archive should name the decision-maker and the document where the decision is visible, without turning interpretation into fact. For example, a programme note may say that the production “works between two languages”; that is evidence of the production’s self-description, not proof of every rehearsal practice.

Attribution is part of language documentation. Record translators, adaptors, interpreters, surtitle translators, captioners, sign-language interpreters, dialect coaches and dramaturgs with the roles they actually held. Avoid the generic “translated by the company” unless the company is the documented rightsholder or collective author. Keep commissioned translation agreements, permission records and credit lines with the production records where policy allows. Rights and consent rules vary by jurisdiction and repository; the archive should state which policy applies rather than implying that a general workflow settles permissions.

Rehearsal and performance evidence

Rehearsal language is often the least documented and most useful layer. A rehearsal room may use a shared language that is not the performance language; it may use interpreting, code-switching, gesture or a devised pidgin. Call sheets, rehearsal notes, annotated scripts, audio notes and video can all provide evidence. Label each item with its language and function: “rehearsal note, English, for staging”, “audio reference, Italian, pronunciation”, “translation query, French to English”. If a recording is transcribed or translated for the archive, record the transcriber or translator, the purpose and the date, and distinguish the transcript from a verbatim record.

Performance evidence has its own layers. A video may capture spoken dialogue but not the surtitle screen; a surtitle file may preserve a translation that was never published; an audience recording may include programme announcements in another language. Link these layers to the same production event, but do not merge them. The archive should make clear whether a language was spoken, sung, signed, projected, captioned, described or printed, and whether it was dominant, occasional or present only in a specific scene.

Surtitles, captions and access languages

Surtitles are a translation layer, not a subtitle track that can be inferred from the script. Record the source language and version used, the display language, the cue file, the operator or automation, and any timing constraints. The same applies to captions, audio description and sign-language interpretation. Language access can change between venues and performances; if the archive documents a tour, record each stop separately rather than assuming a single “surtitled version”.

This is also where archives can expose hidden labour. A surtitle translator may work from a rehearsal script rather than a published text; a captioner may make live editorial decisions; an interpreter may mediate a post-show talk. Name those roles and keep their files with the production record. If access files are not collected, say so in the scope note: absence is information.

Describe the archive in a declared language

Archival description has its own language choice. A repository may describe a multilingual collection in one working language, while the material remains in several. DACS notes that descriptions in United States repositories are frequently in English even when the materials are in other languages, and it asks for the language and script of the material to be recorded separately (DACS 4.5). IASA 0.E sets out preferences for using the language and script presented in specified areas, with fallbacks where the archive or cataloguing agency must choose (IASA 0.E). The practical rule is to declare both: the language of the description and the language of the described material.

In digital systems, make these fields machine-readable. Use ISO 639 for languages and ISO 15924 for scripts where appropriate, as DACS suggests. Use a controlled list for roles (source text, translation, adaptation, rehearsal, performance, surtitle, caption, programme, description) and a controlled list for extent (dominant, occasional, scene-specific, uncertain). A stable controlled list is more useful than eloquent but inconsistent prose, and it prepares the record for exchange.

Theatre and performance archives have been slow to adopt standards for data reusability, findability and interoperability, as a 2018 article comparing performance documentation with language documentation argues. The comparison is not a claim that theatre should become linguistics. It is a reminder that language documentation has developed relatively stable tools for recording languages, speakers, genres, transcripts and annotations. Theatre archives can borrow the discipline of separating layers and linking evidence. A 2026 CIDOC CRM SIG modelling document discusses a Linked Art extension proposal for the performing arts and compares models such as AusStage, Europeana Data Model and CapData Opéra; this is a live modelling discussion, not a universal requirement. For most production archives, a well-designed local register is the necessary first step; linked data can follow if resources allow.

Workflow and checklist

  1. Before rehearsal: appoint a language documentation lead, agree a register, list expected languages and scripts, and identify translation or adaptation versions and rights records.
  2. During rehearsal: label notes and recordings by language and function; record interpreters and bridge translators; keep translation queries and decisions.
  3. At production: collect script versions, surtitle and caption files, programmes, press materials and access records; note which language was used in which scene or segment.
  4. At deposit: normalise language and script codes; separate material language from description language; write a scope note explaining what is present, dominant, occasional or missing.
  5. At review: check names, roles, version links, consent and rights status; record uncertainty and any embargo.

Common problems

“Multilingual” as the only description. It hides which languages, where, and for whom. One language field per production. It cannot represent source, rehearsal, performance and access layers. Translation credits without versions. A name without a text version does not explain what was translated. Description language left implicit. Users cannot tell whether an English scope note describes English material or translates from another language. Recordings without consent or context. A rehearsal recording may be sensitive; document permissions and access conditions. Files without links. A language register that does not connect to the item is only an inventory of labels.

The overall aim is not to exhaustively caption every utterance. It is to make the production’s language choices legible and reusable: what languages and scripts were present, in what roles, produced by whom, in which versions, and under what conditions. That is the difference between an archive that says “multilingual” and one that can support the next translation, restaging, access provision or research question.

FAQ

Is it enough to describe a production as “multilingual”?

No. “Multilingual” does not say which languages or scripts were used, in what roles, in which versions, or with what evidence. A language register should name each language and script and link it to a document, event or access layer.

Should the archive record every language spoken in rehearsal?

Record the languages that materially affect the production and its future use, with their roles and extent. Rehearsal-room bridge languages, interpreting arrangements and translation queries can be important evidence, but the register does not need to transcribe every utterance.

What language should the archival description use?

Follow the repository’s description policy and declare the language openly. DACS notes that United States repositories often describe in English while material may be in other languages; IASA 0.E sets preferences for using the language and script presented in specified areas. The key is to record the description language separately from the material language.

How should translators and adaptors be credited?

Record each person or collective with the role they actually held, the version or text they worked on, and any permission or credit-line evidence. Avoid generic company attribution unless the company is the documented author or rightsholder.

Does a theatre archive need linked data to document language choices?

Not necessarily. A consistent local register with controlled fields is the first step. Models such as the Linked Art extension proposal discussed by CIDOC CRM SIG are relevant for exchange and interoperability, but they are not a precondition for good production-level documentation.

READ FURTHER

Sources & references

  1. Language documentation: a reference point for theatre and performance archives?, International Journal of Performance Arts and Digital Media, published 2 January 2018, https://doi.org/10.1080/14794713.2018.1453242
  2. CIDOC CRM SIG, Stage CRM SIG: Modelling Questions — Towards a Shared Model for Performing Arts: A Linked Art Extension Proposal, dated 2026-03, https://cidoc-crm.org/sites/default/files/StageCRMSIG_ModellingQuestions.pdf
  3. International Association of Sound and Audiovisual Archives, 0.E. Language and script of the description, https://www.iasa-web.org/cataloguing-rules/0e-language-and-script-description
  4. Society of American Archivists, 4.5 Languages and Scripts of the Material (Required), Describing Archives: A Content Standard, https://saa-ts-dacs.github.io/dacs/06_part_I/05_chapter_04/05_languages_and_scripts_of_the_material.html

THE LANGUAGE OF PRACTICE

Related terms

Language register
A structured list recording each language or script used in a production and the role, version, source and evidence attached to it.
Language of description
The language in which archival metadata, scope notes and finding aids are written, as distinct from the languages of the materials described.
Source text
The text from which a translation or adaptation begins; its language, edition and version should be documented.
Surtitle file
A cue-based translation or caption file projected during performance, with its own source language, display language, timing and operator context.
ISO 639
A standard set of codes for representing names of languages in metadata systems.
ISO 15924
A standard set of codes for representing names of scripts in metadata systems.

A SHARED BODY OF KNOWLEDGE

Something to add or correct? Help us keep the conversation informed.

Corrections & updates

Leave a Reply

Your email address will not be published. Required fields are marked *

THE CONVERSATION CONTINUES

Keep a window open to the world.

Fresh perspectives and practical reading, straight to your feed reader.

Subscribe via RSS

FOLLOW YOUR CURIOSITY

Find your next idea.

Search the journal, practical guides, and pages.

THE PRACTICAL TOOLKIT / 01

Your first conversations.

Use these prompts to begin planning with your collaborators. Your progress is saved in this browser.

A starting point for discussion. Adapt it to your project and check local requirements with the relevant organisations.

THE PRACTICAL TOOLKIT / 02

A time to come together.

Choose a date and time in UTC to see when your collaborators would meet. Times account for daylight saving on your chosen date.

Shaded times fall outside 09:00–18:00 local time. Agree working hours and access needs with your team.

BEHIND THE IMAGES

Photography credits

Editorial stock photographs illustrate themes; they do not document the productions discussed on this site. Used under the Pexels licence.