[tei-council] representing transliteration in @xml:lang (was Re: biblscope and imprint)

[Kevin Hawkins]

On 11/14/12 8:51 AM, Martin Holmes wrote:
> The subtag for Russian looks like this:
>
> %%
> Type: language
> Subtag: ru
> Description: Russian
> Added: 2005-10-16
> Suppress-Script: Cyrl
> %%
>
> which I interpret to mean that the default script is Cyrl; that means
> that you don't need to say ru-Cyrl, and in fact ru = ru-Cyrl. Therefore
> if we retain this example, we must supply a script subtag.

BCP 47 says that you should not use a script subtag in combination with 
a language subtag for which a suppress script is given.  That is, 
"ru-Cyrl" is incorrect.  Nothing in BCP 47 says that you must or even 
should use a script subtag if there is a different script.  That is, 
"ru" for transliterated is not a lie, as Lou says.

>> Kevin is right that "ru" is
>> underspecification, but it's very wide, so wide as to imply the default
>> that isn't the case. "ru-Latn" is still underspecification, but much
>> narrower, and with some consequences for machine processing.
>
> Agreed. It's much better than nothing, and I don't think that there's a
> requirement that it be more precise. More precise would be better, and
> more processable, but since "ru-Latn" is actually used as an example in
> the W3C explanatory document, I maintain that it's at least adequate,
> and since the actual subtag required is not (yet) available, we should
> stick with it.

We all seem happy with "ru-Latn".  I was about to make that change, but 
I see that Lou did that already on 4 November.

>> Commenting on this further in the Guidelines seems a matter of taste and
>> balance, possibly a footnote on the existence of many alternative
>> romanized transliterations might be quite useful.

It's tempting to add a footnote, but I we can't put a footnote siglum 
inside of an <egXML> without it being confusing, and I don't see another 
sensible point of attachment.  I think it's fine how it is.

--Kevin