Appeal to ISO 639 RA in support of Elfdalian

[Luc Pardon]

On 23-04-16 06:07, Peter Constable wrote:
> Based on information I've heard about linguistic distinctness — and I have no reason not to trust the combined sources of that information — I have been persuaded that Elfdalian should be coded as a separate language.
> 
> However, I am concerned about process.
>

   The proper process would be for us to go it alone, i.e. assign a
BCP47-only language tag and be done with it.

   Let's not loose perspective of what is going on here.

   Mats came here because he has texts that need tagging. We sent him
over to ISO, as required by BCP47 2.2.1.5 and 3.6.

   Mats went and asked, they replied "no", and now he's back here.

   Now we could take that "no" as final and do what must be done. After
all, the RA knows his business better that we do, by definition.

   However, we happen to think that, in this particular case, the RA
made the wrong decision. At least for that reason by itself, going it
alone is likely to cause problems in the future. We could decide to not
care about that. We could say: "their decision, their problem".

   Instead, we are basically asking the RA : "Are you really, really sure?"

   That is all there is to it.

   Yes, we are explaining that we are asking because we think the
decision is wrong, and we also explain why we think so. In that sense
our letter can be read as a kind of appeal, but strictly speaking we're
only asking him to have another look.


> The only justification I see given is that people don't want to wait
through 8 months with uncertainty about the outcome.

   The justification, as I see it, is simply that we don't want to
initiate the proper process at our end (i.e. assign a BCP47-only
language tag) without making sure that the RA really stands behind his "no".

   If he does, we can go ahead with a clear conscience, knowing that at
least we have tried.

> I completely understand that waiting 8 months may be inconvenient for
some.

   It's not so much about 8 more months as about how many times we
should send an applicant back and forth. If Mats initiates an 639-3
appeal (which he is not even required to do) and comes back here next
year, with yet another "no", what are we going to do then? How would the
situation be different from now?

   In fact, if Mats really wanted to push things, he could simply file a
formal BCP47 application form (something he is now entitled to do).
Given the overwhelming (or should I say "unprecedented") consensus over
here that Elfdalian should indeed be tagged as a separate language,
"approve" is the only possible outcome.

   So if he wanted, Mats could have his tag in two weeks and one second
from now (two weeks for the mandatory review period to pass and one
second for the LTR to push the "approve" button).

   Instead Mats has been waiting patiently while we drafted and
redrafted and re-redrafted our letter, and he agrees to await the answer
- without any guarantee that there will be any answer at all, let alone
when. Nobody can reproach him a lack of patience.

> But what people are not considering is that, for the RA, to engage in
a non-public process may be _far, far more damaging_ to them.


   I have no idea how and why that would be a "non-public process", and
if so, what the extent of the damage would be.

   However, that should not prevent us from sending our letter.

   The RA is much better placed to assess the potential damage than we
are, and he can act accordingly.


> It is _the openness of the RA’s process_ that has enabled us to be discussing this at all! Otherwise, you would have received a rejection without any idea as to why or as to who else provided what input. 

   First off, unless I missed something, nobody here has complained
about lack of openness by itself. The debate over here is about how to
deal with Elfdalian.

   But since you brought it up:  Yes, the rejection letter is public,
and yes, that enables us to discuss it, and yes, we do have an "idea as
to who else provided what input", but no, we have no idea as to why the
linguistic arguments of the loosing party were rejected. As far as I
understand, they were not even rejected, they were simply ignored, and
nobody knows why.

   If that is indeed the way 639-3 is designed to work, I can understand
that people complain about lack of openness.


   More important is the fact that, if it takes years to fix such
obvious "errors in iudicando", the process is in fact unworkable. There
should be a way for the RA to have another look and say "oops, you're
right, I forgot about your argument nr. 4711", and then there should be
a way for him to fix that without necessarily requiring the "victim" to
go through a new yearly cycle.

   Since I have seen nobody say that we're wasting our time in
requesting the RA to fix this "out of cycle", I assume there _is_ such a
way.

   So let's give it a try and see what comes out of it.

   The sooner the better.

   Luc