Published products can be related to languages two main ways.

They can be written in one or more languages, and/or translated from one or more languages, and/or augmented with different languages (so the main text is in one language, while the abstract or subtitles or audio track can be in one or more others). These are the languages of content.

On the other hand, they can be on the subject of languages – either of languages in general, or of one or more specific languages or a group of languages.

Because the ONIX standard allows all of these facets to be represented, Consonance allows languages to be incorporated in your metadata in all of these ways.

If your book is about languages, then use the subject area to represent this. BIC, BISAC, and Thema all contain codes for this purpose.

These cover subjects such as Language and Linguistics, Sign Language, Multi-language Dictionaries, as well as comprehensive lists of languages themselves.

There may also be qualifiers that are relevant, either directly or indirectly.

• Educational qualifiers: for example TOEFL
• Special purpose or interest qualifiers: for example in Thema’s 5PB: Relating to Ethnic Minorities and Groups series
• Historical qualifiers, for language subjects relating to a particular time period
• Geographical qualifiers, for languages relating to particular countries or areas of the world.

Where it is desirable to specify details of the language of content, we pretty much have you covered.

On the metadata page, under Languages of content, add multiple languages and describe the role that they have for the work.

Primarily, this is the Language of text. However, where multiple languages are used and one or more are the translations of others, then you can represent this using the Original language in a multilingual edition and Translated language in a multilingual edition codes.

Translated, or not

There is a subtle difference here between having a book in which the text is in two languages, and a book in which one language is translated into the other. If translation of text is not a function of the book’s content, and text is available in two languages, then use 01: Language of text for both.

National variants in language can also be represented by adding a relevant country code, hence you can express the concept of Swiss German, Canadian French, French French, British English, etc.. We do not validate this relationship, so you are free to specify Finnish Mongolian if you feel that it is necessary to do so.

Be aware of the presence of obscure, artificial, historical, or group languages in the code list, and be sure not to accidentally choose enm: English, Middle (1100-1500) if you mean to choose the standard eng: English.

If your work has no linguistic content, for example in the case of a series of recordings of steam engines, or an illustrated book with no captions or text whatsoever, the code zxx: No linguistic content is available.

Finally, express the concept that your work has been translated into the current language from a different work entirely. In the Publication section, under Inheritance, you can identify the publisher on an original language work that was used as the basis for your translation, and the language of the original work. For example, you might state that your publication of Miss Smilla’s Feeling for Snow was translated from Danish. In such cases, as in multilingual editions with translators, consider adding the translator as a contributor also, of course.