RFC 1766
This document is obsolete. Please refer to RFC 3066 and RFC 3282.






Network Working Group                                      H. Alvestrand
Request for Comments: 1766                                       UNINETT
Category: Standards Track                                     March 1995


                Tags for the Identification of Languages

Status of this Memo



   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Abstract



   This document describes a language tag for use in cases where it is
   desired to indicate the language used in an information object.

   It also defines a Content-language: header, for use in the case where
   one desires to indicate the language of something that has RFC-822-
   like headers, like MIME body parts or Web documents, and a new
   parameter to the Multipart/Alternative type, to aid in the usage of
   the Content-Language: header.

1.  Introduction



   There are a number of languages spoken by human beings in this world.

   A great number of these people would prefer to have information
   presented in a language that they understand.

   In some contexts, it is possible to have information in more than one
   language, or it might be possible to provide tools for assisting in
   the understanding of a language (like dictionaries).

   A prerequisite for any such function is a means of labelling the
   information content with an identifier for the language in which is
   is written.

   In the tradition of solving only problems that we think we
   understand, this document specifies an identifier mechanism, and one
   possible use for it.







Alvestrand                                                      [Page 1]

RFC 1766                      Language Tag                    March 1995


2.  The Language tag



   The language tag is composed of 1 or more parts: A primary language
   tag and a (possibly empty) series of subtags.

   The syntax of this tag in RFC-822 EBNF is:

    Language-Tag = Primary-tag *( "-" Subtag )
    Primary-tag = 1*8ALPHA
    Subtag = 1*8ALPHA

   Whitespace is not allowed within the tag.

   All tags are to be treated as case insensitive; there exist
   conventions for capitalization of some of them, but these should not
   be taken to carry meaning.

   The namespace of language tags is administered by the IANA according
   to the rules in section 5 of this document.

   The following registrations are predefined:

   In the primary language tag:

    -    All 2-letter tags are interpreted according to ISO standard
         639, "Code for the representation of names of languages" [ISO
         639].

    -    The value "i" is reserved for IANA-defined registrations

    -    The value "x" is reserved for private use. Subtags of "x"
         will not be registered by the IANA.

    -    Other values cannot be assigned except by updating this
         standard.

   The reason for reserving all other tags is to be open towards new
   revisions of ISO 639; the use of "i" and "x" is the minimum we can do
   here to be able to extend the mechanism to meet our requirements.

   In the first subtag:

    -    All 2-letter codes are interpreted as ISO 3166 alpha-2
         country codes denoting the area in which the language is
         used.

    -    Codes of 3 to 8 letters may be registered with the IANA by
         anyone who feels a need for it, according to the rules in



Alvestrand                                                      [Page 2]

RFC 1766                      Language Tag                    March 1995


         chapter 5 of this document.

   The information in the subtag may for instance be:

    -    Country identification, such as en-US (this usage is
         described in ISO 639)

    -    Dialect or variant information, such as no-nynorsk or en-
         cockney

    -    Languages not listed in ISO 639 that are not variants of
         any listed language, which can be registered with the i-
         prefix, such as i-cherokee

    -    Script variations, such as az-arabic and az-cyrillic

   In the second and subsequent subtag, any value can be registered.

   NOTE: The ISO 639/ISO 3166 convention is that language names are
   written in lower case, while country codes are written in upper case.
   This convention is recommended, but not enforced; the tags are case
   insensitive.

   NOTE: ISO 639 defines a registration authority for additions to and
   changes in the list of languages in ISO 639. This authority is:

         International Information Centre for Terminology (Infoterm)
         P.O. Box 130
         A-1021 Wien
         Austria
         Phone: +43 1  26 75 35 Ext. 312
         Fax:   +43 1 216 32 72

   The following codes have been added in 1989 (nothing later): ug
   (Uigur), iu (Inuktitut, also called Eskimo), za (Zhuang), he (Hebrew,
   replacing iw), yi (Yiddish, replacing ji), and id (Indonesian,
   replacing in).

   NOTE: The registration agency for ISO 3166 (country codes) is:

         ISO 3166 Maintenance Agency Secretariat
         c/o DIN Deutches Institut fuer Normung
         Burggrafenstrasse 6
         Postfach 1107
         D-10787 Berlin
         Germany
         Phone: +49 30 26 01 320
         Fax:   +49 30 26 01 231



Alvestrand                                                      [Page 3]

RFC 1766                      Language Tag                    March 1995


   The country codes AA, QM-QZ, XA-XZ and ZZ are reserved by ISO 3166 as
   user-assigned codes.

2.1.  Meaning of the language tag



   The language tag always defines a language as spoken (or written) by
   human beings for communication of information to other human beings.
   Computer languages are explicitly excluded.

   There is no guaranteed relationship between languages whose tags
   start out with the same series of subtags; especially, they are NOT
   guraranteed to be mutually comprehensible, although this will
   sometimes be the case.

   Applications should always treat language tags as a single token; the
   division into main tag and subtags is an administrative mechanism,
   not a navigation aid.

   The relationship between the tag and the information it relates to is
   defined by the standard describing the context in which it appears.
   So, this section can only give possible examples of its usage.

    -    For a single information object, it should be taken as the
         set of languages that is required for a complete
         comprehension of the complete object. Example: Simple text.

    -    For an aggregation of information objects, it should be taken
         as the set of languages used inside components of that
         aggregation.  Examples: Document stores and libraries.

    -    For information objects whose purpose in life is providing
         alternatives, it should be regarded as a hint that the
         material inside is provided in several languages, and that
         one has to inspect each of the alternatives in order to find
         its language or languages.  In this case, multiple languages
         need not mean that one needs to be multilingual to get
         complete understanding of the document. Example: MIME
         multipart/alternative.

    -    It would be possible to define (for instance) an SGML DTD
         that defines a <LANG xx> tag for indicating that following or
         contained text is written in this language, such that one
         could write "<LANG FR>C'est la vie</LANG>"; the Norwegian-
         speaking user could then access a French-Norwegian dictionary
         to find out what the quote meant.






Alvestrand                                                      [Page 4]

RFC 1766                      Language Tag                    March 1995


3.  The Content-language header



   The Language header is intended for use in the case where one desires
   to indicate the language(s) of something that has RFC-822-like
   headers, like MIME body parts or Web documents.

   The RFC-822 EBNF of the Language header is:

    Language-Header = "Content-Language" ":" 1#Language-tag

   Note that the Language-Header is allowed to list several languages in
   a comma-separated list.

   Whitespace is allowed, which means also that one can place
   parenthesized comments anywhere in the language sequence.

3.1.  Examples of Content-language values



   NOTE: NONE of the subtags shown in this document have actually been
   assigned; they are used for illustration purposes only.

   Norwegian official document, with parallel text in both official
   versions of Norwegian. (Both versions are readable by all
   Norwegians).

      Content-Type: multipart/alternative;
             differences=content-language
      Content-Language: no-nynorsk, no-bokmaal

   Voice recording from the London docks

      Content-type: audio/basic
      Content-Language: en-cockney

   Document in Sami, which does not have an ISO 639 code, and is spoken
   in several countries, but with about half the speakers in Norway,
   with six different, mutually incomprehensible dialects:

      Content-type: text/plain; charset=iso-8859-10
      Content-Language: i-sami-no (North Sami)

   An English-French dictionary

      Content-type: application/dictionary
      Content-Language: en, fr (This is a dictionary)

   An official EC document (in a few of its official languages)




Alvestrand                                                      [Page 5]

RFC 1766                      Language Tag                    March 1995


      Content-type: multipart/alternative
      Content-Language: en, fr, de, da, el, it

   An excerpt from Star Trek

      Content-type: video/mpeg
      Content-Language: x-klingon

4.  Use of Content-Language with Multipart/Alternative



   When using the Multipart/Alternative body part of MIME, it is
   possible to have the body parts giving the same information content
   in different languages. In this case, one should put a Content-
   Language header on each of the body parts, and a summary Content-
   Language header onto the Multipart/Alternative itself.

4.1.  The differences parameter to multipart/alternative



   As defined in RFC 1541, Multipart/Alternative only has one parameter:
   boundary.

   The common usage of Multipart/Alternative is to have more than one
   format of the same message (f.ex. PostScript and ASCII).

   The use of language tags to differentiate between different
   alternatives will certainly not lead all MIME UAs to present the most
   sensible body part as default.

   Therefore, a new parameter is defined, to allow the configuration of
   MIME readers to handle language differences in a sensible manner.

    Name: Differences
    Value: One or more of
         Content-Type
         Content-Language

   Further values can be registered with IANA; it must be the name of a
   header for which a definition exists in a published RFC.  If not
   present, Differences=Content-Type is assumed.

   The intent is that the MIME reader can look at these headers of the
   message component to do an intelligent choice of what to present to
   the user, based on knowledge about the user preferences and
   capabilities.

   (The intent of having registration with IANA of the fields used in
   this context is to maintain a list of usages that a mail UA may
   expect to see, not to reject usages.)



Alvestrand                                                      [Page 6]

RFC 1766                      Language Tag                    March 1995


   (NOTE: The MIME specification [RFC 1521], section 7.2, states that
   headers not beginning with "Content-" are generally to be ignored in
   body parts. People defining a header for use with "differences="
   should take note of this.)

   The mechanism for deciding which body part to present is outside the
   scope of this document.

    MIME EXAMPLE:

    Content-Type: multipart/alternative; differences=Content-Language;
              boundary="limit"
    Content-Language: en, fr, de

    --limit
    Content-Language: fr

    Le renard brun et agile saute par dessus le chien paresseux
    --limit
    Content-Language: de
    Content-Type: text/plain; charset=iso-8859-1
    Content-Transfer-encoding: quoted-printable

    Der schnelle braune Fuchs h=FCpft =FCber den faulen Hund
    --limit
    Content-Language: en

    The quick brown fox jumps over the lazy dog
    --limit--

   When composing a message, the choice of sequence may be somewhat
   arbitrary. However, non-MIME mail readers will show the first body
   part first, meaning that this should most likely be the language
   understood by most of the recipients.

5.  IANA registration procedure for language tags



   Any language tag must start with an existing tag, and extend it.

   This registration form should be used by anyone who wants to use a
   language tag not defined by ISO or IANA.










Alvestrand                                                      [Page 7]

RFC 1766                      Language Tag                    March 1995


----------------------------------------------------------------------
LANGUAGE TAG REGISTRATION FORM

Name of requester          :
E-mail address of requester:
Tag to be registered       :

English name of language   :

Native name of language (transcribed into ASCII):

Reference to published description of the language (book or article):
----------------------------------------------------------------------

   The language form must be sent to <ietf-types@uninett.no> for a 2-
   week review period before submitting it to IANA.  (This is an open
   list. Requests to be added should be sent to <ietf-types-
   request@uninett.no>.)

   When the two week period has passed, the language tag reviewer, who
   is appointed by the IETF Applications Area Director, either forwards
   the request to IANA@ISI.EDU, or rejects it because of significant
   objections raised on the list.

   Decisions made by the reviewer may be appealed to the IESG.

   All registered forms are available online in the directory
   ftp://ftp.isi.edu/in-notes/iana/assignments/languages/

6.  Security Considerations



   Security issues are not discussed in this memo.

7.  Character set considerations



   Codes may always be expressed using the US-ASCII character repertoire
   (a-z), which is present in most character sets.

   The issue of deciding upon the rendering of a character set based on
   the language tag is not addressed in this memo; however, it is
   thought impossible to make such a decision correctly for all cases
   unless means of switching language in the middle of a text are
   defined (for example, a rendering engine that decides font based on
   Japanese or Chinese language will fail to work when a mixed
   Japanese-Chinese text is encountered)






Alvestrand                                                      [Page 8]

RFC 1766                      Language Tag                    March 1995


8.  Acknowledgements



   This document has benefited from innumberable rounds of review and
   comments in various fora of the IETF and the Internet working groups.
   As so, any list of contributors is bound to be incomplete; please
   regard the following as only a selection from the group of people who
   have contributed to make this document what it is today.

   In alphabetical order:

   Tim Berners-Lee, Nathaniel Borenstein, Jim Conklin, Dave Crocker,
   Ned Freed, Tim Goodwin, Olle Jarnefors, John Klensin, Keith Moore,
   Masataka Ohta, Keld Jorn Simonsen, Rhys Weatherley, and many, many
   others.

9.  Author's Address



   Harald Tveit Alvestrand
   UNINETT
   Pb. 6883 Elgeseter
   N-7002 TRONDHEIM
   NORWAY

   EMail: Harald.T.Alvestrand@uninett.no
   Phone: +47 73 59 70 94

10.  References



    [ISO 639]
         ISO 639:1988 (E/F) - Code for the representation of names of
         languages - The International Organization for
         Standardization, 1st edition, 1988 17 pages Prepared by
         ISO/TC 37 - Terminology (principles and coordination).

    [ISO 3166]
         ISO 3166:1988 (E/F) - Codes for the representation of names
         of countries - The International Organization for
         Standardization, 3rd edition, 1988-08-15.

    [RFC 1521]
         Borenstein, N., and N. Freed, "MIME Part One: Mechanisms for
         Specifying and Describing the Format of Internet Message
         Bodies", RFC 1521, Bellcore, Innosoft, September 1993.

    [RFC 1327]
         Kille, S., "Mapping between X.400(1988) / ISO 10021 and RFC
         822", RFC 1327, University College London, May 1992.




Alvestrand                                                      [Page 9]