Cree
cre has a single ISO 639-2 form and the ISO 639-1 equivalent cr; in ISO 639-3 it is a macrolanguage covering several individually coded Cree languages.
cre looks like an ordinary language code, but behind it is a continuum of related languages spoken across a large part of Canada, from the Rocky Mountains to Labrador, with communities in northern Montana as well. That breadth is the main thing to understand before using the code.
One code in ISO 639-2, several languages in ISO 639-3
ISO 639-2 has a single entry for Cree, cre, with the ISO 639-1 equivalent cr. The Library of Congress record at http://id.loc.gov/vocabulary/iso639-2/cre carries the label Cree in both English and French.
ISO 639-3 splits the picture further. There, cre is a macrolanguage: an umbrella over several individual languages that have their own codes. Among them are Plains Cree (crk), Woods Cree (cwd) and Swampy Cree (csw), along with other varieties spoken further east. A macrolanguage code is not the same thing as a collective code such as nai. It still names one language identity for many purposes; it simply admits that the identity covers distinct varieties.
Which value goes where
| Context | Typical value |
|---|---|
| MARC 21 008/35-37 and 041 | cre |
| Dublin Core with an ISO 639-2 vocabulary | cre or its id.loc.gov URI |
HTML lang, xml:lang (BCP 47) |
cr, or the variety subtag such as crk |
| Language-technology corpora | usually the ISO 639-3 variety code |
The mismatch between the coarse library code and the finer web tags means that a digital collection pulling metadata from its catalog into a public website often has to add the variety by hand. Projects that do this keep a small mapping table from the 546 note (“In Plains Cree”) to the tag.
Scripts
Canadian Aboriginal syllabics were developed in the nineteenth century for Cree and spread to other languages, including Ojibwe and Inuktitut. Communities in the eastern and northern parts of the Cree area write mainly in syllabics, while western communities more often use Latin letters, and both are found in print. For preservation the script is as important as the language: syllabics need fonts and encodings that older systems sometimes lacked, and some legacy documents rely on pre-Unicode font hacks that render as Latin gibberish once the font is gone. Recording cr-Cans or crk-Cans makes that dependency visible. A format record for such a file should also name the font, since the language tag says what the text is meant to be and the font record says what is needed to see it.
Common slips
- Using
crein an HTMLlangattribute. It parses, but BCP 47 prefers the shortercr. - Coding Ojibwe, Innu or Naskapi material as Cree because of shared syllabics. Ojibwe has oji; Innu and Naskapi have ISO 639-3 codes of their own.
- Treating Michif, a mixed language with Cree verbs and French nouns, as Cree. It has its own ISO 639-3 code and is usually best described with a note.
The same kind of distinction comes up for Caddo and other North American languages: the more specific code, where it exists, is the one future users can actually search on.
Example in RDF
<rdf:Description rdf:about="https://example.org/periodicals/community-newsletter-1978-03.pdf"
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:dcterms="http://purl.org/dc/terms/">
<dcterms:title>Community newsletter, March 1978</dcterms:title>
<dcterms:language rdf:resource="http://id.loc.gov/vocabulary/iso639-2/cre"/>
<dcterms:language rdf:resource="http://id.loc.gov/vocabulary/iso639-2/eng"/>
<dcterms:description>Parallel text in Cree syllabics and English.</dcterms:description>
</rdf:Description>
Questions about cre
What is the language code for Cree?
In ISO 639-2 the code is cre, in ISO 639-1 it is cr. The Library of Congress publishes the three-letter code at http://id.loc.gov/vocabulary/iso639-2/cre. ISO 639-3 treats cre as a macrolanguage and gives individual Cree languages their own codes.
Which code should a web page in Plains Cree use?
BCP 47 allows either the macrolanguage subtag cr or the more specific ISO 639-3 subtag for the variety, crk for Plains Cree. The specific subtag is more informative; cr is acceptable when the variety is unknown or when software only recognizes the macrolanguage.
Can library catalogs record specific Cree varieties?
MARC 21 language fields use the MARC Code List for Languages, which follows ISO 639-2 and has only cre. Catalogs usually record the variety in a 546 language note or in subject headings, while the coded field stays cre.
Is Cree written in syllabics or Latin letters?
Both. Many Cree communities, especially in the east and north, write in Canadian Aboriginal syllabics, while others, particularly on the western plains, mainly use Latin-based spelling. BCP 47 can mark the script with Cans or Latn, for example cr-Cans.
Are Ojibwe and Innu covered by cre?
No. Ojibwe has its own ISO 639-2 code, oji. Innu (Montagnais) and Naskapi are closely related to Cree but have separate ISO 639-3 codes and are not members of the cre macrolanguage.