Re: [Ltru] Proposal: include new "Language-Type:" field

I think this is out of scope for the registry; our goal should be to have
sufficient description to identify the language (that is, to distinguish it
from others), and no more -- not to be the kitchen sink of information about
languages.

In addition, I don't see much of a need to duplicate information that is in
639, and I doubt that this additional information is "very useful to help
shrink the large number of languages to a more manageable size". Most people
will be picking among living languages, and this classification only
separates out 8% of those. Making a list of 7K languages 8% smaller is not
the key to an effective UI ;-). Those who need this information can go to
639 for it.

  living 6989 92.0%  extinct 417 5.5%  ancient 114 1.5%  historic 53 0.7%
constructed 24 0.3%  oddball 3 0.0%  total 7600

Mark

On 10/4/06, John Cowan <cowan@ccil.org> wrote:
>
> A proposal:
>
> The 4646bis registry should capture the language type information from
> ISO 639-3.  Each language, including macrolanguages, is labeled in -3
> as either living, extinct, ancient, historic, or constructed.  These
> terms are defined precisely at http://www.sil.org/iso639-3/types.asp .
> Though informative rather than normative, this is very useful to help
> shrink the large number of languages to a more manageable size.
>
> The current draft of 639-3 contains 6989 living languages, 417
> extinct languages, 114 ancient languages, 53 historic languages,
> and 24 constructed languages.  The codes 'mul', 'und', and 'zxx' are
> special cases.
>
> The registry should also capture the individual language vs.  language
> collection information from ISO 639-2.  If a code element appears in
> -2 but not in -3, it is a language collection; there are 68 such code
> elements.  Now that we have 639-3 code elements for essentially every
> language on the planet, language-collection subtags are extremely vague
> and provide little guidance to the recipient.
>
> (I'm not really happy with the vagueness of "Language-Type", and would
> prefer "Language-Status", but it's the term used in the FDIS.)
>
>
> I propose the following language for 4646bis section 3.1.2:
>
> o Language-Type
>         o Language-Type's field-body contains one of the values
>           'collection', 'extinct', 'ancient', 'historic', 'constructed',
>           or 'special'.  This field MUST NOT appear except in records
>           of type 'language'.
>
>
> And here's a draft of the new section 3.1.3.8:
>
> 3.1.3.8.  Language-Type field
>
>         The field 'Language-Type' MUST only appear in records whose
>         'Type' field-body is 'language'. This field MUST NOT appear
>         more than once in a record.  Most of the language records in
>         the registry represent individual living languages.  This field
>         indicates those which are not.
>
>         The value 'collection' indicates a language collection appearing
>         in ISO 639-2 but not ISO 639-3.  The values 'extinct', 'ancient',
>         'historic', and 'constructed' indicate languages which are so
>         designated in ISO 639-3; precise definitions of these terms can
>         be found in that standard.  The value 'special' is used for the
>         three subtags 'mul', 'und', and 'zxx', which do not actually
>         designate languages at all.
>
>
> Finally, here's a rule for section 4.1:
>
>         8.  Language subtags with a 'Language-Type' field of 'collection'
>         do not represent specific languages, and SHOULD NOT be used
>         unless more specific information is unavailable.
>
> Appropriate adjustments would be needed to 3.3, 3.4, and 3.5 as well.
> We should be able to set this field if and when we ever register
> a language subtag directly, and change it when 639-3 changes.
>
> --
> John Cowan
>         cowan@ccil.org
>                 I am a member of a civilization. --David Brin
>
> _______________________________________________
> Ltru mailing list
> Ltru@ietf.org
> https://www1.ietf.org/mailman/listinfo/ltru
>