Re: [Ietf-languages] [Ltru] Propose adding ak as prefix to variants akuapem and asante

[John Cowan <[e-mail]>]

On Wed, Mar 1, 2023 at 2:34 AM Martin Hosken <[e-mail]> wrote:


> After some further digging (see
> https://www.ethnologue.com/14/iso639/An_analysis_of_ISO_639.pdf section
> 3.9), it seems that Twi [tw] and Fanti [fat] are both dialects of Akan [ak].


As is well known, there is no hard and fast line between dialects and
languages;  "a language is a dialect with an army and a navy."

In ISO 639-3, on which BCP 47 is based, twi [sic] and fat were originally
separate and unrelated language entries, whereas in 639-1 there were
entries for ak and tw.  This was resolved by making ak a macrolanguage
encompassing tw and fat.

Which lends argument for the regular use of ak-asante and ak-akuapem and
> inclusion of ak as a prefix. It might even be argued that tw and fat should
> be deprecated?
>

You mean as prefixes or as entries?

> From https://www.ethnologue.com/language/aka/ we read (in other comments):
>
> The two main subdivisions of Akan are assigned codes in the ISO 639-3
> standard: Fanti (fat) and Twi (twi). Fanti dialects: Agona, Anomabo Fanti,
> Abura Fanti, Gomua. Twi dialects: Ahafo, Akuapem, Akyem, Asante, Asen,
> Dankyira, Kwawu.
>

You'll note that the Akuapem and Asante varieties are listed under Twi,
which is why they have a prefix of tw.

Thus fat and tw are not distinct from ak


They are encompassed by the macrolanguage term.  Chinese is likewise a
macrolanguage encompassing numerous languages, but nobody says that (e.g.)
Cantonese and Mandarin are "not distinct from Chinese"; that would be a
category error like saying that English and Dutch are not distinct from
Germanic.

(I'm not sure why Akuapem and Asante are so privileged to have variants,
> over the others).


Because nobody has requested them.  We register variants based on need.
Having a variant subtag is not a privilege.

> In that case (which I agree with), what is the value of the Prefix: field
> for a variant?
>

It is a non-normative field indicating that this variant is meaningful as
part of a tag containing the specified subtag(s).  For instance, the
basiceng variant subtag is appropriate with the en language subtag.  The
list of prefixes is not necessarily complete, and some variants such as
fonipa (IPA phonetic transcription) do not have a prefix field because they
are widely or universally applicable.