Portuguese subtags

Luc Pardon lucp at skopos.be
Thu Sep 15 12:50:09 CEST 2011

On 09/14/2011 07:16 PM, Doug Ewell wrote:
> João Miguel Neves <joao at silvaneves dot org> wrote:
>> If the group decides that this is not enough,
>> I'll just drop the matter and use non-official tags.
> If you do end up having to go this route, please, PLEASE use private-use
> tags that are syntactically valid according to BCP 47.  That basically
> means "pt-x-mumble" or "pt-BR-x-mumble" or such, where the "mumble"
> subtag is 2 to 8 letters and/or digits.  Please don't ignore the BCP 47
> validity requirements as some organizations have.

  As far as the "syntactically valid" part is concerned, this is a
reasonable request. However, so is the request for a subtag for the 1990
spelling of Portugese.

  The fact that a government makes a particular spelling official should
be sufficient to demonstrate a valid need for proper tagging. Unless and
until this list is empowered to overrule governments around the globe,
(some of) those government's subjects have a legal requirement to write
in that spelling, so they will and do write in it. Thus there is a need
for tagging the writings they produce. Period.

  Whether people like the spelling or not is totally irrelevant. Whether
the dictionary for it is printed in the proper typeface or on the proper
paper size to be recognized by Michael as a valid source is totally
irrelevant. In fact, whether or not there is a dictionary at all is
irrelevant. The requester must simply provide a "Reference to published
description of the language (book or article)". Note that it says
"article", which seems to imply that the description need not be
exhaustive to be acceptable. In fact, BCP 47 3.5 seems pretty clear to me:

>  In most cases, reference to an authoritative grammar
>    or dictionary of that language will be useful; in cases where no such
>    work exists, other well-known works describing that language or in
>    that language MAY be appropriate.

If a one-paragraph publication in some government gazette were supposed
to be sufficient for that govenment to impose a spelling on its
citizens, it should be sufficient for this list to allow it to be
tagged. The only alternative is to declare documents in this spelling as
"not intended primarily for human communication" (BCP 47 pt 2, 2nd
sentence) and that seems a pretty strong statement to me.

  Michael wrote:

> OK, so where's the list? Not the decree. The specification. The thing you can turn to to make sure that you're spelling a given word correctly. Is there such a source? If not, then there is nothing reliable to point an orthography subtag to, and that is problematic.

  That may be so. But in that case Michael (or others) should probably
explore the possibilities of having this particular international
agreement overturned in the proper court(s) for being "problematic", and
even a succesful outcome does not eliminate the need for a subtag for
the documents that were produced in the meantime. As a foreigner,
Michael may not be bound by the decree, but the citizens in those
countries are, problematic or not. He should not deny them tagging just
because their politicians did a lousy job (politicians usually do), or
because their opponents say they did (they always do). Correct me if I'm
wrong, but the burden Michael is imposing here on the requestor has no
basis in the procedures.

  Minutes after the original requester submitted his proposal, Michael

> Well, I have an extremely negative opinion about this "accord" which does nothing whatsoever good for European Portuguese. And I know it is controversial.

  Correct me if I'm wrong, but I can't see any requirement for a
particular spelling "to do whatsoever good" as a prerequisite for it
being allowed to be tagged. It's not because people wish something
didn't exist that it doesn't.

  I know I'm beating dead horses here, but that is the point, kind of.
People come here because they feel they have a need for a specific
(sub)tag, and because they want to play by the rules. It is perfectly
acceptable that they be required to demonstrate that their request make
sense, in every way required by the relevant RFC/BCP documents. But no
more than that.

  It is not acceptable that, if people request tagging for one
particular spelling of a language, it is imposed on them that they
demonstrate linguistic and political and whatever consensus on all
possible spellings of that particular language since Lucy came out of
Africa. There seems to be agreement on this principle among list
members, yet it is being done all the time, and by the time some list
members protest, the original requester is halfway to being worn out and
p***ed off. Then comes the dictionary requirement argument to finish him

  Actually, sidetracking requests like this is a most efficient
mechanism to block any and all new registrations.

  If this is what this list wants, i.e. if the list considers itself the
Keeper of the Registry's Virginity, it is on the right track.

  If, however, this list intends to help well-intending people tagging
things in a proper way (as I believe it is mandated to), it should ask
itself if it is a good idea to keep inventing new roadblocks all the
time, thereby turning away people with valid and properly established
tagging needs. If the answer is yes, it shouldn't complain when people
start ignoring it, and by extension, also the BCP.

  In this particular case, as I see it, the requester has fulfilled all
the requirements for approval (almost from day 1, I'd say), so the
request should be approved. If it's not, I would encourage him to go
ahead and tag with pt-1990aolp anyway (or even pt-acor1990, or anything
else that is syntactically valid). BCP47-style private-use subtags are
"intended for private interchange" only, they are "not recommended for
general interchange", and they are "useless for information exchange
without prior arrangement", so tagging with "pt-x-mumble" in this case
would be at least as much an abuse of the tagging mechanism as using a
tag that is valid on all counts but was rejected for invalid/irrelevant
reasons only. And nobody will go to jail for using either.

  This list cannot demand that people follow the BCP 47 rules if it
keeps bending the rules itself (e.g. by inventing arbitrary requirements
for approval), and encourages people to do the same (e.g. by suggesting
private tags for public communication).

  So much for the views of an outsider.

  Luc Pardon

More information about the Ietf-languages mailing list