Preserving living Assyrian language with consent

Documentation starts with people

A living-language project should ask speakers how they identify their community, language and home variety. Cambridge's NENA work shows why such distinctions matter: its records span many Christian and Jewish local dialects with distinct linguistic histories. A file called 'Assyrian voice' without locality, date or speaker permissions is not an adequate scholarly record.

Record what the speaker authorised

Written consent should specify whether an interview may be streamed, downloaded, transcribed, translated, quoted, used for teaching or included in research datasets. Those permissions are distinct, and people may have good reasons to restrict a family story, voice or image. Avoid creating public synthetic voice models from recordings or treating public listening access as unrestricted redistribution rights.

Preserve context, not just sound

For a recording, retain a stable identifier, approximate date, language variety, locality where consent permits, recording environment, interviewer, transcription or translation credits and any redactions. The Cambridge databases illustrate how descriptive metadata helps listeners interpret variation. A future public archive should distinguish original recording, edited clip, transcription and editorial gloss as separate artefacts.

Enable correction and future review

Transcriptions and transliterations can contain errors or differ under valid conventions. Speakers, descendants and qualified linguists should have a clear way to request corrections, contextual limits or withdrawal where terms allow. A responsible archive keeps an internal revision trail, credits named reviewers when consented to, and never marks material as verified just because it has been uploaded. Current submissions in this project remain download-only.

Describe real uses before recording

A speaker may agree to a story being used in a language class without agreeing to unrestricted web publication, commercial reuse or the release of identifying family details. Intended uses need ordinary-language explanation rather than one vague permission.

A practical record distinguishes attribution, variety, date and access conditions. Private contacts need not be public. Context and agreed limits strengthen the educational value because later users know what the recording represents and how they are permitted to use it.

Teach through an authorised segment

A short agreed example can explain meaning and grammar without exposing sensitive third-party details. A contributor's agreement does not automatically resolve the privacy of everyone mentioned in a story.

Preserve corrections too. A speaker may clarify a word, correct a transcript or change public attribution preferences. Documenting that change improves reliability. Ethics supports usable language material rather than functioning as a separate checklist added after publication.