ON LANGUAGE NAMES /// RE: Results of Duplicate Busters Survey #2

Doug Ewell doug at ewellic.org
Tue Sep 9 03:57:56 CEST 2008

Lang Gérard <gerard dot lang at insee dot fr> wrote:

> Autonyms ( written directly with a variant of the Latin alphabet in 
> their native script, or taken as romanized form of an autonym whose 
> native script is not a variant of the Latin alphabet, or taken as an 
> IPA phonetisation of the language name when the considering language 
> has no script and is only a spoken language) are very certainly much 
> more specific and identifying of the considered language name (because 
> they are, in particular, the definitive proof for the existence and 
> for the autonomy [also, of the self-denomination, "auto-nomination"] 
> of the considered language) that possible french or english language 
> names for the same languages.
> So, the consideration of autonyms gives a considerable separation and 
> identification power for language names and the underlying language, 
> that cannot be acheived by the approximation deriving from a french or 
> english designation of the same entities.

Devil's advocate: A given culture, especially a relatively isolated one 
such as those whose native language was not reflected in ISO 639 until 
part 3, might well be unaware of other cultures, hundreds or thousands 
of kilometers away, where languages that happen to share the same 
autonym are spoken.

On the other hand, the English and French names (to the extent they 
differ from the native name, which I still contend is overstated for the 
minority languages) may be more likely to have been bestowed with an eye 
toward disambiguation, since there is bound to be a linguist among the 
millions of English and French speakers who is aware of the shared 

Do I have hard evidence to back my assertion?  Nope.  Do you?

> The case that two autonyms could be identifiying the same language 
> name, and the same underlying language, is certainly very infrequent 
> and should be considered with many precautions.

I thought we were debating the opposite scenario: two autonyms that 
happen to be the same but identify two different languages.

> Are you definitively certain that the native phonetisations of both 
> Ainu variants you consider as two distinct language names are 
> different ? If then, maybe the romanization of the chinese version 
> could be changed as not to be identical with the japanese autonym.

In other words, you would propose a slight modification of the ISO 639-3 
name to eliminate the ambiguity.  That's exactly what I proposed by 
doing away with the ambiguous "Ainu" name in the Registry and retaining 
the unambiguous "Ainu (Japan)" and "Ainu (China)" which were already in 
639-3.  I think this way is simpler and easier to comprehend.

Doug Ewell  *  Thornton, Colorado, USA  *  RFC 4645  *  UTN #14
http://www.alvestrand.no/mailman/listinfo/ietf-languages  ˆ 

More information about the Ietf-languages mailing list