Re: Registration requests: akuapem, asante

There is a unified orthography for Akan. What does that mean when, say, Twi Asante, Twi Akuapem, and Fante (which from all descriptions are mutually intelligible) are written in it? One might refer simply to writing Akan (ak)? Or would new subtags for the unified orthography in question reference it under tw or fat?

It seemed that an alternative would be to tag under ak for the unified orthography and those for Asante, Akuapem, and Fante (4 subtags). I'm not arguing for that, and recognize the potential BCP 47 issue you mention (and also, as a not-speaker wouldn't have the knowledge of the languages to make the case), but it was a thought.

This is why, in addition to clearer documentation on usage (for instance on Wikipedia), involvement of experts in the language(s) in question would be helpful. Not intending that as a criticism, but an observation.

Don

PS- A mailbox issue may have booted me off the subscriber's list - will sort it out, thanks

-----Original Message-----
From: Doug Ewell [mailto:[email protected]] 
Sent: Sunday, May 28, 2017 2:29 PM
To: [email protected]
Cc: Don Osborn <[email protected]>; Michael Everson <[email protected]>; [email protected]
Subject: Re: [Ietf-languages] Registration requests: akuapem, asante

Don Osborn wrote:

> It would be nice to maintain a unified approach to Akan, but probably 
> not necessary to have two ISO 639 codes in this tag.

I'm not sure what you mean by "a unified approach." ISO 639 has assigned separate code elements for Twi and Fanti, and therefore individual subtags have been assigned in the LSR as well. Do you suggest that users ignore these and tag all Akan content with the macrolanguage subtag 'ak'?

I never got a response (from anyone) to my question about current tagging practice, quoted yet again below:

| The purpose of BCP 47 is to provide identifiers for language content, 
| based when possible on ISO standard code elements.
|
| ISO 639 has coded Twi [tw, twi] separately from Akan [ak, aka].
|
| Irrespective of the fact that "Twi" is actually a name for two of the 
| three major dialects of Akan (the two for which Sascha has proposed 
| variant subtags), I believe it would be destabilizing for BCP 47 to 
| encourage users to tag Twi content as "ak" or "ak-anything" instead of 
| using "tw".


> Sent this to the list, but apparently a snafu with my subscription 
> (?).

Try re-subscribing. Remember it's @ietf.org now, not @alvestrand.no.

--
Doug Ewell | Thornton, CO, US | ewellic.org