PRESS / NO. 002

No. 002 · Translation report

44 language options.Translation is still in progress.

The language menu is longer than the list of translated questionnaires. Here is where things stand, what the numbers mean, and how to help.

OpenJung · Translated from Chinese

A fox, rabbit, otter and capybara compare proof sheets and arrange printing blocks in a woodland print workshop.
Comparing the proofs, one sheet at a time.FIG. 01

We want people to answer in a language they are comfortable with. Adding a language to the menu is only part of that work: the buttons, questions, answer labels and explanations all need attention.

The coverage report

Snapshot

Language and regional options
44
Complete questionnaire versions
5
Answer-label versions
6

These are counts from the website’s source files, not visitor statistics. Simplified and Traditional Chinese are separate options in this count. A language option does not mean every page has been translated.

The full 32-question test has complete text in English, Simplified Chinese, Traditional Chinese, Japanese and Korean. Other interface languages currently use English questions. Malay has translated answer labels, but not a complete Malay questionnaire.

“Filled” counts non-empty entries against the English interface dictionary. It includes entries that still contain English. It does not measure the whole website: page-specific dictionaries, Press articles, type descriptions, mobile and MCP content are outside this interface count.

“Same as English” flags exact matches for checking, after an allowlist for names, URLs and references. Some matches are legitimate; different wording can still be wrong. “Complete text” requires the title and both answer poles for all 32 questions. None of these counts establishes translation quality or assessment validity.

View all 44 language options

On a narrow screen, scroll the table sideways to see all columns.

Interface entries and complete question text, before English fallback
Language option Interface entries filled Same as English Questions
English en 1250 / 1250 0 32 / 32 Complete text
Simplified Chinese zh 1248 / 1250 51 32 / 32 Complete text
Japanese ja 1250 / 1250 0 32 / 32 Complete text
Korean ko 1233 / 1250 69 32 / 32 Complete text
Traditional Chinese zh-tw 1248 / 1250 87 32 / 32 Complete text
Malay ms 1160 / 1250 314 0 / 32 English fallback
German de 1160 / 1250 286 0 / 32 English fallback
French fr 1160 / 1250 308 0 / 32 English fallback
Spanish es 1160 / 1250 277 0 / 32 English fallback
Arabic ar 1160 / 1250 542 0 / 32 English fallback
Hebrew he 1160 / 1250 542 0 / 32 English fallback
Russian ru 1160 / 1250 273 0 / 32 English fallback
Brazilian Portuguese pt-br 1160 / 1250 272 0 / 32 English fallback
Indonesian id 1160 / 1250 279 0 / 32 English fallback
Vietnamese vi 1160 / 1250 280 0 / 32 English fallback
Thai th 1160 / 1250 279 0 / 32 English fallback
Turkish tr 1160 / 1250 283 0 / 32 English fallback
Italian it 1160 / 1250 277 0 / 32 English fallback
Polish pl 1160 / 1250 280 0 / 32 English fallback
Dutch nl 1160 / 1250 287 0 / 32 English fallback
Hindi hi 1160 / 1250 521 0 / 32 English fallback
Bangla bn 1160 / 1250 616 0 / 32 English fallback
Filipino fil 1160 / 1250 8 0 / 32 English fallback
Ukrainian uk 1160 / 1250 524 0 / 32 English fallback
Swahili sw 1160 / 1250 606 0 / 32 English fallback
Czech cs 1160 / 1250 57 0 / 32 English fallback
Romanian ro 1160 / 1250 52 0 / 32 English fallback
Hungarian hu 1160 / 1250 359 0 / 32 English fallback
Slovak sk 1160 / 1250 598 0 / 32 English fallback
Greek el 1160 / 1250 406 0 / 32 English fallback
Swedish sv 1160 / 1250 68 0 / 32 English fallback
Norwegian no 1160 / 1250 55 0 / 32 English fallback
Danish da 1160 / 1250 74 0 / 32 English fallback
Finnish fi 1160 / 1250 347 0 / 32 English fallback
Burmese my 1160 / 1250 801 0 / 32 English fallback
Khmer km 1160 / 1250 47 0 / 32 English fallback
Lao lo 1160 / 1250 708 0 / 32 English fallback
Sinhala si 1160 / 1250 732 0 / 32 English fallback
Tamil ta 1160 / 1250 609 0 / 32 English fallback
Amharic am 1160 / 1250 715 0 / 32 English fallback
Hausa ha 1160 / 1250 200 0 / 32 English fallback
Yoruba yo 1160 / 1250 443 0 / 32 English fallback
Zulu zu 1160 / 1250 351 0 / 32 English fallback
Igbo ig 1160 / 1250 548 0 / 32 English fallback
Download CSV ↓ Download JSON ↓ Question bank at the recorded revision ↗

What counts cannot tell us

“Sometimes” and “always” are not interchangeable. A sentence can read smoothly while changing the strength of a statement. The two ends of a question, negation and the wording of answer options need to be checked separately.

AI can help draft and compare wording, but it cannot certify a translation. Documented human review and language-specific validation are still missing for the current questionnaire versions. Even the five complete versions should not be mistaken for validated translations.

Work already under way

The test now identifies the interface language, question language and answer-label language separately. It tells readers when the questions will appear in English.

Automated checks compare the declared question languages with every question in the bank. Dictionary reports also identify missing entries, missing placeholders and wording identical to English. These catch specific problems, not every mistranslation.

For Press, language switching keeps the article and section. The five editions are checked for narrow layouts, keyboard navigation and text overflow. Clear fallback notices, readable fonts and right-to-left layout matter alongside the words themselves.

Animal friends welcome a hedgehog carrying a bundle of papers at a woodland correspondence table.
There is room at the table for another reader.FIG. 02

A better way to contribute

The aim is to reach readers in every language we can, while improving the translations already available. We plan to make contributing easier, with smaller tasks and a visible review trail. The process below is a proposal, not a translation platform that has already launched; there is no published rollout date yet.

  1. Small, contextual tasks

    Offer a short passage or a few related strings, with the source, where they appear and any length constraints. Helping should not require taking responsibility for an entire language.

  2. Explain the proposed wording

    Keep the original, current translation, suggestion and reason together. Maintain shared terminology, and record disagreements rather than silently replacing one choice with another.

  3. A separate review step

    Ask another reader fluent in the language to review the change. Questionnaire edits need additional checks on meaning, answer poles and scoring direction, with version changes recorded where appropriate.

  4. Visible status and credit

    Distinguish drafts, work awaiting review and reviewed wording. Record who checked what and when; credit contributors with their agreement. Wording review and language-specific validation must remain separate.

If you would like to help

For now, send a suggestion through the public GitHub issue tracker. Include the language, public page address, original wording, proposed wording and a short explanation. For a question, its number is enough; no answer history is needed.

GitHub issues are public. Do not include private result links, answers or personal information.

Suggest a translation change

Thanks to everyone who points out an unclear sentence or offers a better translation. Those specific suggestions give us something we can check and improve.

NO. 001 / We made a free personality test ↗