Re: [Ietf-languages] ISO TR 21636 Dimensions

[Doug Ewell <[e-mail]>]

Richard Wordingham wrote:
 
>> I know Richard knows this, but just as a reminder: "Region subtags
>> are used to indicate linguistic variations *associated with or
>> appropriate to* a specific country, territory, or region." They have
>> nothing to do with where the speaker, writer, signer, drummer, or
>> whistler is actually located. A Siamese accent could be used in Peru.
>
> So if the form of the Tham script and its encoding appropriate to Pali
> in the province of Udon Thani in Thailand is the same as the form and
> encoding appropriate to Laos, and not the commonest form in Thailand,
> is it indeed appropriate to tag it as pi-Lana-LA if one considers only
> the written aspects?
 
Yes, I would think so, given the conditions you stated.
 
Tagging language usage by region at the granularity of entire countries certainly isn't perfect, but it's well established and very often good enough.
 
The 'u' extension does provide a way to tag at the level of national subdivisions, which I always wanted; you could write "pi-Lana-u-sd-th41" to pinpoint usage in Udon Thani. But this runs into the usual problem that not all tag-aware software understands extensions, especially the relatively new 'sd' key, and especially if Peter's perception that "the 'u' extension is primarily (solely?) used in connection with CLDR" is widely shared. Furthermore, and perhaps more importantly, "pi-Lana-u-sd-th41" loses the connection with common usage in Laos. This might be especially important for matching. So, probably not the best choice here.
 
--
Doug Ewell, CC, ALB | Thornton, CO, US | ewellic.org