Adding variant subtags 'aluku' and 'nduyka' and 'pamaka' fordialects
pascal.vaillant at guyane.univ-ag.fr
Fri Aug 21 11:35:49 CEST 2009
> One thing that bothers me: the Description of 'djk' is "Aukan". Ethnologue, at least,
> doesn't mention "Busi Nenge Tongo", at least by that spelling (it does mention Aluku
> and Boni, though). I don't pretend to know what any of these language are, but it would
> probably make sense to use the same description here and in 'djk' or consider adding
> more Description fields to 'djk'
> Thus, let me suggest that the record could be:
> LANGUAGE SUBTAG MODIFICATION
> File-Date: 2009-08-27
> Type: variant
> Subtag: aluku
> Description: Aluku dialect
> Description: Boni dialect
> Added: 2009-08-27
> Prefix: djk
> Comments: Aluku (which is sometimes called Boni) is a specific dialect of
> Aukan (djk), also called "Busi Nenge Tongo", an English-based Creole
> continuum in eastern Suriname and western French Guiana.
I very strongly oppose to this last suggestion.
As I said in the initial request form:
"the [Ethnologue - aka ISO 639-3] language code -somewhat inappropriately-
refers to the whole language by the name of its most important variant",
which means that (to put it again very explicitly), Aluku and Pamaka are
not *dialects of Ndyuka*, but that Aluku, Ndyuka and Pamaka are closely
related dialects of the same language continuum.
(Recall that this is not a statement from me as an individual person, but
that I send the request on behalf of a research group in linguistics
specialized in the area).
I understand perfectly Mr Phillips' point, which is a quite logical one:
as a matter of fact, most people, when deciding which tag to use, refer
(1) to ISO-639-3 to check the presence and meaning of a language tag, and
(2) to the IANA language subtag registry to check the availability and
meaning description of a proper subtag; so the issue of the cross
consistency of both registers is an important one.
However, I feel that this consistency issue should not in any case be managed
in disregard to accuracy of description. In the present case, stating that
"Aluku [resp. Pamaka] is a dialect of Ndyuka" would be a blatant distortion
of the facts. To be more precise: no linguist knowledgeable in this area,
and no native speaker of any of the three variants, would agree with the
description, such formulated. It would be felt like something comparable
(in respectable proportions) to saying that Croatian is a dialect of Serbian,
instead of saying that they both belong to a common language continuum.
A point remains: it is true that Ethnologue has an entry for Ndyuka and
not for the two others (am I the only one in this list aware that Ethnologue
has some inaccuracies, inconsistencies, or incomplete descriptions? I could
name a few others in my area ;-)). It is, very simply probably, due to the
fact that Ndyuka was the first one to be described, which is in its turn
probably connected to the fact that it is the dialect with the most speakers.
I mentioned that in my original post as a problem to try to solve later, by
writing to SIL/ISO-639-3 R.A. to suggest a change in the tag meaning description.
This is a not-so-urgent and long process, and we are aware that even if the
tag meaning description finally gets to change in ISO-639, introducing a new
3-letter mnemonic is an ordeal, as well as a very symbolic issue on which we
do not wish to cling (many languages have mnemonics which refer to an old or
partial denomination anyway, and people don't do a fuss about it).
At present, we need the three tags for digital corpora, and we feel (after
due discussion with colleagues who are specialists of the said dialects),
that the situation is best solved by adding three variant subtags in the
IANA registry, rather than by trying to register Aluku and Pamaka as separate
language entries in ISO-639-3. As a matter of fact, the degree of
similarity/divergence is best described at the language/dialect level
than at the metalanguage/language level (judging by comparisons with
other attested examples). So we can put up with the slight, temporary,
inaccuracy of saying that "Aluku [resp. Ndyuka] is a dialect of Busi Nenge
Tongo [djk]", when a lookup in ISO-639-3 at the present time yields
"djk = Aukan/Nduka", but it is not a good solution to the problem to
say that "Aluku is a dialect of Ndyuka". By the way, this would lead us
to state, symmetrically, that "Ndyuka is a dialect of Ndyuka".
We should try to make the tags fit the reality, not the other way around.
More information about the Ietf-languages