[osis-core] Schema: type on language

[Chris Little]

Patrick,

On Sat, 11 Oct 2003, Patrick Durusau wrote:

> Chris,
> 
> Just so I understand the issues:
> 
> 1. Language prose should conform to the schema

Yes.  I think that's always a good idea, TEI traditions notwithstanding.

> 2. Language element's type attribute should equal (values as listed)
> 
> ISO-639 list as: ISO-639-1, ISO-639-2

yes

> SIL list as: Ethonologue

I have to take back my previous comment.  We should probably go with "SIL" 
as the value.  There's a document from SIL, by Gary Simons and/or Peter 
Constable I'm guessing, that identifies, for example, x-SIL-ENG as their 
prefered form for making SIL/Ethnologue codes RFC 3066/xml:lang-compliant.  
I take that to mean they prefer "SIL"--it might just be a length issue, 
however.

> Linguist

It's identified as "LINGUIST" (all caps), actually.  Or "LINGUIST List".  
It's just one of those LISTSERV things that has been carried further than 
it really needed to be.

> other
> 
> (will just enumerate a list)
> 
> 3. lanuage element should have a role attribute like date
> 
> Not sure I understand, not that I disagree, just don't understand its 
> function.
> 
> Reading your suggested value:
> 
>  > Some possible values for language's role include: original, translation,
>  > interlinear, quotation (for excerpts), didactic (for grammars), source &
>  > target (for dictionaries).
> 
> I am still puzzled by what I would use the role function to indicate.
> 
> Can you say a little bit more about what it is that you want to do?

If I have a work that is a translation, e.g. the Divine Comedy, I might 
mark something like:
<language type="ISO-639-1" role="original">it</language>
<language type="ISO-639-1" role="translation">en</language>

Similarly, for BDB, I would identify Hebrew as role="source", English as
role="target", and perhaps Arabic, Syriac, Greek, etc. as
role="quotation".

An English interlinear NT would have Greek as role="original" and English 
as role="interlinear".

--Chris