Aside,

At least some of transliterations (particularly "names") can be treated using existing BCP-47 tokens.

For example, in OCN:100011210: 

700 [0] 1_   [33$6] 880-04   [1$a] Vukić, Ljiljana.   
880 [0] 1_   [33$6] 700-04/(N   [1$a] Вукић, Љиљана.   

Mapping these fields (and using other clues in the record) can reasonably produce:
    rdfs:label "Ljiljana Vukić"@en ;
    rdfs:label "Љиљана Вукић"@sr ;
Romanization rules may have been used to generate the 700 form, but capturing that fact doesn’t seem very important.

This mechanism doesn’t work well for non-names, which end up being more about capturing phonetics as opposed to “language". In those cases, the literals would have to be translated instead of transliterated in order to attach more useful “languagy" language-tags.

I would also note that web browsers are configured to use BCP-47 tokens which servers can leverage for display. It’s unlikely that someone would choose extension language tags to control displays of the data.

Jeff

From: Bibliographic Framework Transition Initiative Forum <[log in to unmask]> on behalf of Andrew Cunningham <[log in to unmask]>
Reply-To: Bibliographic Framework Transition Initiative Forum <[log in to unmask]>
Date: Friday, July 1, 2016 at 12:48 PM
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: [BIBFRAME] Language tags

But if record is being consumed by system that requires bcp47, or being output to web, you still have the issue of passing along or generating an appropriate language tag.

On 2 Jul 2016 1:58 am, "Young,Jeff (OR)" <[log in to unmask]> wrote:
Another approach might be to use SKOS-XL instead of language tags like so:

:A1 a skosxl:Label;
rdfs:label “Ελληνική Δημοκρατία”@gr
skosxl:literalForm “Hellēnikē Dēmokratia”;
.

It’s a little heavy, but a construct like this would tie the key pieces together. A custom language tag on the transliteration can’t tie in the native term.

Jeff

From: Bibliographic Framework Transition Initiative Forum <[log in to unmask]> on behalf of Andrew Cunningham <[log in to unmask]>
Reply-To: Bibliographic Framework Transition Initiative Forum <[log in to unmask]>
Date: Friday, July 1, 2016 at 11:39 AM
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: [BIBFRAME] Language tags


On 1 Jul 2016 2:19 am, "Joseph Kiegel" <[log in to unmask]> wrote:
>
> The approach of creating variant subtags for specific editions of the ALA-LC romanization tables is outmoded, since the tables are now on the Web. 
>
>  

It is less than ideal, yes, but it was the only one that was proposed as a variant subtag. And no one has proposed a way to tag anything else.

It doesn't help that romanisation tables lacked good versioning information, etc.