Re: ISO 639, 'Code for the representation of names of languages'

From: Misha Wolf (
Date: Wed Nov 29 1995 - 22:40:38 EST

As language tagging and RFC 1766 have been attracting a lot of interest
recently, members of this list may be interested in the attached mail I have
just sent to ISO Central Secretariat.



A problem has arisen recently in regard to the accuracy of ISO 639, 'Code
for the representation of names of languages'.

ISO 639 was last published in 1988. Infoterm is the designated Registration
Authority (RA). In 1989 the RA issued a newsletter, detailing some
additions and some modifications. Extracts from the newsletter can be
found below. As will be seen, the newsletter stated quite clearly that the
changes "have been made".

Reuters, in common with many other companies and organisations, took these
changes seriously and implemented them. By now there are many databases
containing numerous items tagged with the changed symbols.

As evidence of the wide acceptance of these changes I include:

1. The following extract from Internet RFC 1766, 'Tags for the
    Identification of Languages'. This RFC has been heavily referenced
    and forms the basis of language-tagging for MIME body parts and Web

    "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)."

2. The appended extracts from a file generated by Keld Simonsen. This file
    has been widely disseminated on the Internet and it too is heavily
    referenced elsewhere.

When I spoke with Infoterm by telephone on 14 November, I was astonished to
hear that the 1989 newsletter should be seen merely as a proposal, with no
formal status.

It is simply not acceptable for a Registration Authority to issue a formal
written notice of changes to an ISO standard and then, six years later, to
seek to annul these changes, especially without any formal written notice.

Misha Wolf
Globalization & Standards Manager, Reuters


Extracts from Keld Simonsen's file follow. In reformatting them, I have
lost some of the accents present in the original.


ISO 639 -- Code for the representation of names of languages

Technical contents of ISO 639:1988 (E/F)
"Code for the representation of names of languages".
Typed by 1990-11-30
Two-letter lower-case symbols are used.
The Registration Authority for ISO 639 is Infoterm, Osterreiches
Normungsinstitut (ON), Postfach 130, A-1021 Vienna, Austria.

<<stuff deleted>>


I S O Registration Authority I n f o t e r m
        for ISO 639 "Code for the Affiliated to ON (Austrian Standards
        representation of names of Institute)
        languages" (ISO 639/RA) Heinerstra<ss>e 38 Wien 2 Austria

                                      International Information Centre for

Postal address: <O">sterreichisches Normungsinstitut (<O">N)
Infoterm Postfach 130 A-1021 Wien (Austria)


             ISO 639/RA N E W S L E T T E R No 1/1989

Upon the recommendation of the Advisory Committee (ISO 639/RA-AC) and after
consulting the national standards organizations or other appropriate
institutions according to the "Guidelines for the Registration of Languages
and their Symbols", changes (additions, deletions, revisions) have been made
as below. Please note that a vacated symbol will not be reinstated for a
five-year period.

(English (original name (instead of) (symbol) (instead of)
 language name) of language)

New symbols included in ISO 639

Uigur Uygurqe ug
   Inuktitut CA Inuktitut iu
   (superseded: Chuang) Saw Cueng za

Symbols changed in ISO 639

Hebrew ‘Ibrit (Iwrith) he (iw)
Yiddish Yidis (Jiddisch) yi (ji)
Indonesian (Bahasa) Indonesia id (in)

<<stuff deleted>>

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:30 EDT