ISO 639 name change: Songhai languages

John Cowan cowan at
Sat Jan 6 20:26:01 CET 2007

Don Osborn scripsit:

> Having a more or less systematic methodology for classification is a
> blessing; respecting its limitations is all the more so.


> This may get beyond the range of topics appropriate for the lists I'm
> posting to, so I won't pursue the issue here.

I've trimmed to just ietf-languages.

> Nevertheless, it seems to me worth considering. For
> any who are interested, I raise it in another way at

>From that message:

> Tavultesoft Keyman has a list of available keyboard layouts with
> indications of which languages each is intended for, as based on
> Ethnologue.  In the case of some African languages the Ethnologue
> language classification is not the appropriate level - e.g. a keyboard
> is really not set up for a variety(ies) of Fula/Fulfulde/Pulaar as
> Ethnologue defines it(them), but for Fula in the orthography defined
> on a country level(s).

This is exactly the sort of evidence that would persuade 639-3/RA to add
a new macrolanguage.  They are aware of the issue in principle, and only
need to be persuaded, now that the macrolanguage level has been invented
(or discovered).  It's true that hitherto macrolanguages have only been
used as a shim between 639-2 and 639-3, but new ones independent of
639-2 will be set up as the evidence comes in.  (Another new application
of macrolanguages will be when the RA decides that one of the individual
languages needs to be split: the old code will be changed to macrolanguage

> BTW, <annoying question>will codes used for what are called "language
> groups" reside eventially in 639-5?</annoying question>

Yes.  The final system will look like this:

ISO 639-3 contains 3-letter codes for individual languages and

ISO 639-5 contains 3-letter codes for language collections, mostly
genetic but a few of other types.

ISO 639-2 contains selected 639-3 and 639-5 codes, plus a few oddballs
like mul (multiple languages), und (undetermined languages), and zxx
(non-linguistic material).

ISO 639-1 provides 2-letter codes for a selection of 639-2 entities.

ISO 639-6 contains codes for language collections, macrolanguages,
individual languages, and varieties of individual languages, including
modal (spoken/written/signed), geographical, and sociolinguistic variants.
Entities also in 639-3 or 639-6 will share codes; other entities will
have 4-letter codes.

Evolutionary psychology is the theory           John Cowan
that men are nothing but horn-dogs,   
and that women only want them for their money.  cowan at
        --Susan McCarthy (adapted)

More information about the Ietf-languages mailing list