ISO 639-5 reconfirmation ballot (long)

Caoimhin O Donnaile caoimhin at
Thu Jul 21 17:11:12 CEST 2016

Thanks for the explanations, Debbie.  I understand.

I got interested in all this because of my own needs for sorting and 
ordering languages by family:

and possibly this cognates project in the future:

So far, the best framework I have found for my own needs is the set of 
Composite Trees at Linguistlist Multitree (thanks Anthony!).  I took a 
copy of that several years ago, tidied it up a bit, and laboriously add 
ordering among sib nodes on an ad-hoc basis as I need them.

> the 3 letter code space is not large enough for both language and 
> language variation and thus would cause considerable confusion.

I agree.  That is very much the impression I have too.  One of the big 
advantages (at least for some people, some purposes, as Doug pointed 
out) of going to a hierarchic tree of atomic codes is that it escapes 
the difficulty and political arguments and potential instability of 
deciding what is a language, or dialect, or language group.  So you 
might as well include major dialects in the system, and these are needed 
for many applications anyway.  With all the codes that will be needed 
for major dialects, and languages, and language groups, and deprecated 
language groups after theories are revised, you would be lucky to have 
enough three-letter codes.  And it becomes totally impossible if you 
want the codes to be even semi-mnemonic.


More information about the Ietf-languages mailing list