Re: [Ietf-languages] adjectival usage of variant subtags

That's a strawman; one doesn't need to enumerate them.

One can have a simple policy for when it comes time to add a variant that
could have general applicability, such as "vestland" (for 'Western
Norwegian' = vestlandsk). In such a case, use the corresponding general
word "western" instead, and add to the description the example 'no-western
= Norwegian Vestlandsk'. Then there is no need to have a special variant
for Western Slovak dialects, and innumerable other cases.

Of course, where such a variant is not applicable (eg, there is no dialect
of Chickasaw that it would make sense to call 'Western'), there is no
purpose to using that variant. But that is no different than
other productive uses of language subtags, say, ja-AQ.

Mark

On Fri, Dec 4, 2020 at 7:32 PM Michael Everson <everson@evertype.com> wrote:

> This is why we have preferred specific subtags referring to specific
> entities, rather than
>
> northern, southern. eastern, western, northeastern, southeastern,
> northwestern, southwestern, central, peripheral, secret, male, female,
> hill, valley, plains, mountain, coastal, insular, archaic, old, middle,
> catholic, protestant… all of this is too vague and there would be scores of
> additional geographical or social generic tags. It is not wise to try to
> enumerate these.
>
> Michael
>
> > On 30 Nov 2020, at 00:16, Mark Davis ☕️ <mark@macchiato.com> wrote:
> >
> > Well, if the registry had {eastern, western, northern, southern,
> central} variants, then someone could have sv-northern. The meaning would
> be the customary meaning of 'northern' — a variety spoken in the northern
> part of the area of the Earth in which that language is typically spoken.
> It is quite common to distinguish varieties based on general geographical
> location: look at Kurdish.
> >
> > For many languages that 'adjective' would not be particularly useful,
> and like any subtag, should be avoided where it does not mark a useful
> distinction. For example, for English it would not be particularly useful.
> >
>
> _______________________________________________
> Ietf-languages mailing list
> Ietf-languages@ietf.org
> https://www.ietf.org/mailman/listinfo/ietf-languages
>