[Ietf-languages] Adding prefixes with dialect variants to Occitan orthographic variants

[Doug Ewell <[e-mail]>]

I've been asked about the following case, broadly paraphrased:

Suppose one wants to create a tag for Occitan that specifically calls out both (a) the Gascon dialect and (b) the Mistralian orthography. This is not the most commonly encountered combination of dialect and orthography, but seems to exist in the wild, which is all the more reason why one might want to tag it.

The logical choice of tag would seem to be "oc-gascon-grmistr". However, all 13 of the variant subtags for Occitan -- those for dialects as well as those for orthographies -- have only one Prefix value:

Prefix: oc

That means, while it is technically permissible to create the tag "oc-gascon-grmistr" (just as any combination of discrete variant subtags is technically permissible), BCP 47 does not recommend it, and tools that create or validate tags will usually flag such a tag with a warning.

The approach recommended in BCP 47, and the one we actually use in practice, is to add additional Prefix values:

Prefix: oc
Prefix: oc-gascon
Prefix: oc-nicard
Prefix: oc-provenc
etc.

This has the added advantage of indicating the recommended order of variants, so that one will be encouraged to write, for example, "oc-gascon-grmistr" and not "oc -grmistr-gascon".

I'd like to get a feel for the Registrar's and other list members' inclination toward augmenting the Occitan orthographic variant subtags by adding Prefix values that include some or all of the dialect variants, as illustrated above. I've also cc'd stakeholders on this message, the requester, David Mediavilla, as well as Lo Congrès Permanent de la Langue Occitane and its representative who originally proposed the 13 variants in 2018.

If it is not expected that all combinations of 10 dialect variants and 3 orthographic variants might reasonably be used together, then we would need a list of which combinations do need to be added.

If we do not decide to add these Prefix values, then we should probably provide some sort of advice to the stakeholders how they should tag such a combination, in light of what BCP 47 says.

--
Doug Ewell, CC, ALB | Lakewood, CO, US | ewellic.org