The Contributor Roles Taxonomy (CRediT) — formalized as ANSI/NISO Z39.104-2022 — is officially published in English only. Over the past two years, a community-led translation project has been closing that gap at scale: as of this writing, human-reviewed CRediT translations exist in more than 35 languages, maintained in machine-readable form so publishing systems can use them directly. For journals editing, submitting, or indexing outside the English-language mainstream, that milestone changes what “adopting CRediT” practically requires.
The milestone: translations now cover most of a journal’s likely author languages
CRediT’s 14 contributor roles and their definitions became a full ANSI/NISO standard in 2022, but like other NISO standards, the authoritative text is English-only. That is a real adoption barrier for journals whose editorial boards, authors, or readerships work primarily in another language, particularly outside North America and Western Europe. Starting with a hackathon at the Big Team Science 2023 conference, a volunteer initiative called the Contributorship Collaboration set out to translate the taxonomy’s role names and definitions into other languages while keeping the underlying vocabulary identical to the English original. That project’s own tracking page now lists translations completed or in progress for more than 35 languages, stored as machine-readable files in a public GitHub repository so journal management systems, author-facing forms, and publisher workflows can pull them in directly rather than each publisher commissioning its own translation from scratch.
What NISO does and doesn’t do here
It’s worth being precise about governance, because it’s easy to over-read “35 languages” as an official NISO release. It isn’t. NISO’s own CRediT materials are explicit that the organization does not produce or formally endorse specific translations of its standards — the ANSI/NISO Z39.104-2022 text itself remains the sole normative version, in English. The translation effort is a separate, community-run undertaking that maps onto the same 14 roles and definitions rather than replacing or altering them. That distinction matters for anyone citing a translated version in an editorial policy: it is a usage aid, not a substitute normative text.
How the translations are produced
The Contributorship Collaboration’s published process requires at least two people fluent in the target language to work on each translation — commonly a draft translation followed by an independent back-translation or cross-check pass into English, to surface discrepancies before a version is marked complete. Translator names and the translation procedure used are recorded alongside each language version for transparency, and the project has documented its methodology in a preprint that received review through MetaROR, a Publish-Review-Curate metascience review platform. This is a meaningfully more rigorous process than an uncredited machine translation, but it is still a volunteer, non-normative effort — quality and completeness can vary by language, and not every one of the 35-plus languages has necessarily been reviewed by the same number of translators.
What actually changes for a non-English-language journal
For an editor deciding whether this milestone is relevant, the practical implications are narrower than “CRediT is now available in your language” might suggest:
- Author-facing text can be localized without changing the metadata. A journal can present role labels and definitions to authors in their own language during submission, while the contributor-role values recorded in the manuscript’s metadata, JATS XML, and any data sent to Crossref or similar registries still map back to the same 14 canonical English-language role identifiers. That consistency is what makes CRediT statements interoperable across journals and databases in the first place — see the 14 CRediT roles for the canonical list.
- Lower per-journal translation burden. Regional and non-English-language publishers no longer need to commission their own ad hoc translation of role definitions (or rely on an inconsistent one) — they can adopt an existing, dual-reviewed version instead, which also reduces the risk of semantic drift between journals that publish in the same language but translated the taxonomy independently.
- Easier onboarding for authors who don’t work primarily in English. Author confusion about what a role like “Formal analysis” or “Project administration” actually covers is a real friction point in CRediT adoption; localized, reviewed definitions reduce miscategorized contributions at the point of submission rather than requiring later correction.
- It does not create a separate taxonomy. A translated CRediT statement should still resolve to the same role set documented on CASRAI’s CRediT roles reference — journals integrating a translation into a submission system should confirm the underlying role identifiers, not just the display labels, are unchanged.
What to check before a journal adopts a translated version
Because these translations are not NISO-issued, an editorial office adopting one should treat it the way it would any community-maintained reference tool: check who reviewed it and how many fluent reviewers were involved, confirm the language is actually one of the completed (not still in-progress) entries, and verify that the version maps cleanly onto the unchanged 14-role English original before wiring it into a submission system or house style guide. CASRAI maintains a running reference of translation status by language at CRediT translations, which is a useful starting point alongside the Contributorship Collaboration’s own repository.
Frequently asked questions
Is a CRediT translation an official NISO document?
No. ANSI/NISO Z39.104-2022 is published in English only, and NISO does not produce or formally endorse specific-language translations. Community translations are a usage aid that maps onto the same standard, not an alternate normative text.
Does using a translated CRediT statement change what metadata gets sent to Crossref, PubMed, or ORCID?
No, provided the translation is implemented correctly. The role identifiers exchanged in JATS XML and registry metadata should remain the same 14 canonical English-language values regardless of what language is shown to the author or reader on the page.
How is translation quality checked?
The main community effort (the Contributorship Collaboration) requires at least two fluent speakers per language, typically a draft plus an independent back-translation or cross-check, with translator names and procedure documented for each completed language.
Where can a journal find an existing translation?
The Contributorship Collaboration publishes completed and in-progress translations in machine-readable form in a public repository, with a human-readable version on its project site. CASRAI’s own translation status reference tracks language coverage as well.
What if a journal’s language isn’t covered yet?
The project has an open process for contributing new translations, following the same dual-translator review approach used for existing languages — coverage has grown from a handful of languages at its 2023 origin to 35-plus, and continues to expand.
For the underlying standard itself, see ANSI/NISO Z39.104-2022 and the 14 CRediT roles. For how CRediT relates to a similarly named taxonomy, see CRediT vs. NISO Contributor Roles Taxonomy. Journals crediting translators for actual manuscript translation work (a distinct question from translating the taxonomy) should see Translator contribution.
