New extension for transformed languages

Doug Ewell doug at
Mon Mar 5 16:23:50 CET 2012

Philippe Verdy <verdy underscore p at wanadoo dot fr> wrote:

>> What problems with language coding do you see?
> The first problem is to have the language recognized as such and not
> just as an informal, undesired dialect of another official form of the
> same language. There have been lots of battles (still continuing
> today) about the variants of Serbo-Croatian, still continuing today
> between Serbian and Montenegrin, or Romanian vs. Moldovan.

This is a problem of language classification, not encoding as such, and
it will probably be around forever. The level of precision needed to say
without question that two "languages" are the same, or dialects of one
another, or truly different language, simply doesn't exist.

Unfortunately, but probably unavoidably, ISO 639 contributes to this by
defining common or distinct "ethnolinguistic identities" as criteria for
calling languages the same or different, which ends up being interpreted
as "nationality" or "religion" and which leaves the door open for the
battles Philippe mentioned.

Fortunately, ISO 639 has no concept of "undesired" dialects; the words
"dialect" and "pidgin" and "creole" are often used prescriptively or
pejoratively in popular usage, but only descriptively in the standard.

> The question of codification comes next after the resolution of those
> linguistic battles of classification between what are considered plain
> languages or dialects of the same language, even if they have several
> standardized forms made official but contradicting the language
> continuums between them.

Language coding has to be useful to ordinary people, not just linguists.
There is always going to be some level of distinction, if a trained
professional looks hard enough, between the way any two people speak and
write, but in order to classify and encode languages in a useful way,
some lines have to be drawn, and some are going to seem arbitrary. The
annual changes in ISO 639-3 reflect the ongoing attempt to get this
right; note that many of the "new" and "deleted" languages are actually
splits and merges of existing languages.

Doug Ewell | Thornton, Colorado, USA | @DougEwell ­

More information about the Ietf-languages mailing list