Language Subject Identifier

[HughP]

Proposal

I am proposing a new (to Zotero) repeatable field, which is specifically for recording the language that a resource is about.

The need for two elements

Zotero already has a IU element (and presumably a database 'slot' or RDF element) for language. However, this 'slot' is the equivalent to the .ris identifier: LA - Language. When this element is used, it is used for the language which a resource is written in. Also equivalent to MARC 21 code point 041a.

What I am proposing is different. It is a second metadata element which is specifically for the language which the resource is about.

To facilitate the interoperability of language information linguists (and librarians) rely on the ISO 639-3 standard. (ISO 639-3 is used by Dublin Core, MARC21, Z39.53, MODS and other Metadata schema exchange systems. )

Linguist use this information quite a bit when looking for, and sharing metadata about resources -- especially resources about a particular language. See Simons (2013) slide three for how researchers use it: http://www-01.sil.org/~simonsg/local/ISO%20639-3.pdf

See the title of this work for an example of ISO 639-3 in action: http://kuscholarworks.ku.edu/dspace/handle/1808/6408

Spatial Reference in Rongga (ISO 639-3: ror), Balinese (ISO 639-3: ban), and Indonesian (ISO 639-3: ind)

Many language archives (for instance PARADISEC, Berkley, ELDP, etc ) and several resource discovery (glottolog and Open Language Archive Community (OLAC)'s search service services use an this additional element to describe and help resource users find resources.

Item record from PARADISEC collection AC1.

glottolog

Why ISO 639-3

• As of RFC 5645, RFC 5646 and BCP47 the preferred method of language Identification is through ISO 639-3. See discussion here: http://en.wikipedia.org/wiki/IETF_language_tag
• The OLAC language Extension calls for the use of ISO 639-3. http://www.language-archives.org/REC/language.html .
• The ISO 639-3 code is also designated by Dublin Core (via RFC 4646) as the way to designate "A language of the resource."
• MARC and MODS both use ISO 639-3. Because RFC 5645, RFC 5646 replace RFC 3066.
• MusicBraniz uses ISO 639-3 in the identification of its resources: http://tickets.musicbrainz.org/browse/MBS-1799

• Z39.50 uses ISO 639-3 codes via Z39.53 per section: 3.2.10.1.2 Searching for Information in a Particular Language

  To search for records in a certain language, the HumanStringLanguage attribute may be used (in conjunction with the three-character language code as the term; see Z39.53-1994).

  Why not use... in Zotero

  Keywords

  1. Keywords are less than idea because some users might apply the keyword differently ('ems','iso-639-3-ems', etc.)
  2. All keywords would need to be parsed and ISO 639-3 keywords would need to be extracted, while not impossible, it becomes difficult especially, if point number one is true. These parsers would then need to be maintained.

     The current language field option

Because this field already has a defined meaning, which is not the same as the subject language of the resource.

How am I currently getting around this?

I'm not. I am using Endnote because it has several custom fields. I can't add custom fields to Zotero. I would like to move to Zotero. I know other linguists who choose to use other citation management solutions because they can add custom fields, one of those is usually a subject language.

How might the UI work?

It could work like a keyword field as a comma separated value, but ideally it would be a single, but repeatable, select drop down menu -- however, at 7,700 codes that is a long list and unless something like Select2 or some other similar UX element were used.

[adam3smith]

I appreciate the work that you put into this (though for the most part it's better to bring up & discuss feature requests on the Zotero forums. This is an issue tracker with low participation and ill suited for discussion), however I don't think it's going to happen.

Zotero devs have said that it's possible that, at some point, there may be custom fields in Zotero (though certainly not anytime soon), but what we can absolutely not do is add specific fields for characteristics of works used in particular disciplines — There'd simply be no end to that list. Until there are more custom fields, the extra field or tags are going to be the best there is for such categories.