On Thu, 16 Jan 1997, Stu Weibel wrote:
> The following is a note I received about the Dublin Core
>
> I'm particularly interested in what people think of his comments
> vis a vis language codes
>
> stu
>
> ----- Begin Included Message -----
>
> >From [log in to unmask] Thu Jan 16 12:34:59 1997
> From: [log in to unmask]
> X400-Originator: [log in to unmask]
> X400-Recipients: [log in to unmask]
> X400-Mts-Identifier: [/PRMD=SANDOZ/ADMD=400NET/C=CH/;0034700003269798000002]
> X400-Content-Type: P2-1988 (22)
> To: <[log in to unmask]>
> Subject: Dublin Core Metadata Set
> Date: Thu, 16 Jan 1997 18:35:36 +0100
>
>
> Hi Stuart,
> may I shortly introduce myself:
>
> I have been working in the context of dictionary data to support our
> large databases, inclusive creation of metadata databases about
> documentation archived at our company.
> The company I am working for is NOVARTIS, the outcome of the merger of
> SANDOZ and CIBA. Novartis is a 100 000 employees worldwide chemical
> company focusing on healthcare, agribusiness and nutrition with a total
> sales of around 35 000 Million Swiss Franks. The headquarters is based in
> Basel, Switzerland, Europe.
>
> I very much appreciate your attempt to standardize on a very small set of
> fields to index any data. We have come several times across this problem
> in the past and have found internal solutions to some of the problems.
> Therefore I have two recommendations:
>
> First: The labels you have chosen conflict in some programming languages
> with reserved words, e.g. Date. I would therefore recommend to prefix all
> your labels with something like "DC_" for Dublin Core, which would make
> them definetly free. The label for Date would then read DC_Date.
>
> Second: Your recommendation to code languages according to Z39.53 three
> character codes for written languages, I do not like. This codification
> is a national one and does not code dialects, like American English or
> Swiss German. There is already an international agreed Codification for
> Languages and dialects inclusive a recommendation for their use
> especially in librarian contexts: It is the "ISO-639 (1988-04-01) Code
> for the representation of names of languages" issued by the International
> Organization for Standardization.
>
>
> Regards
> Armin Weltin
I think I already explained this once before concerning the language
codes.
The current ISO standard for language codes, ISO 639 is for a 2-character
code, generally established by the terminology community. It was found to
be inadequate for many uses (particularly the library community) because
there were only so many combinations possible if using only 2 characters.
Work began around 1988 on establishing a 3-character code by an ISO
committee (TC37/SC2 TC46/SC4 Joint Working Group on Language Codes). This
has been a very long process and a Draft International Standard was issued
about a year ago; ballotting was completed last month (and we don't yet
have the results). This new standard will be called ISO 639-2. The
maintenance agency for the 2-character code list (ISO 639-1) is INFOTERM
in Austria; the maintenance agency for the 3-character code list (if it's
approved) will be the Library of Congress. (LC issued the Draft
International Standard ISO 639-2.)
The 3-character list was generally based on Z39.53 (the American National
Standard on language codes); hence the citation to that standard in the
Dublin Core, since ISO 639-2 has not yet been approved. If it is
approved, DC should cite ISO 639-2. Since the 3-character code list was
based on Z39.53, which is the national language code standard that MARC is
based on, the intention is to have the MARC and ISO language codes the
same. There are some lesser-known languages in the draft standard that
now have different codes in MARC; these will be changed once the
international standard is approved. There are about 20 languages that use
different codes depending upon whether it is the bibliographic or
terminology community using them.
Incidentally, there are NOT separate codes for dialects in either ISO
639-1 or 639-2, e.g. both German German and Swiss German use the same
code. I'd be glad to answer any further questions about these code lists.
Rebecca
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^^ Rebecca S. Guenther ^^
^^ Senior MARC Standards Specialist ^^
^^ Network Development and MARC Standards Office ^^
^^ Library of Congress ^^
^^ Washington, DC 20540-4020 ^^
^^ (202) 707-5092 (voice) (202) 707-0115 (FAX) ^^
^^ [log in to unmask] ^^
^^ ^^
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|