Registration request for new subtags for Portuguese orthographies

Luc Pardon lucp at
Thu Mar 26 12:03:11 CET 2015

On 25-03-15 16:40, Michael Everson wrote:
> So, again, if a user in Portugal wants his variants of ao1990 and a user in Brazil wants his variants of ao1990, how would you propose to respond to those user requirements? 

  How to respond to these user requirements? Well, how about telling the
user to simply type his preferred variant on the keyboard, and not
typing the other? What's wrong with that? We have done precisely that
for many years when writing our Dutch, and as far as I know nobody has
died of it.

  Next, are you sure this is an actual user requirement, or is it simply
something that you suppose might be (or become) a user requirement?

  Also, supposing that it is a real requirement, what makes you think
that _all_ users in Portugal will want one variant (say with "facto" and
"génio"), and _all_ users in Brazil will want one other variant (say
with "fato" and "gênio")? To me that looks like an oversimplification.

  If that is what they want, then pt-PT-ao1990, pt-BR-ao1990 etc. will
do fine. But I think that won't wash, and then you have a huge problem.

  Suppose I am a writer of Portuguese, and I am forced to retrain my
brain to adhere to ao1990, and I pause to think about how I'm going to
handle that.

  I can imagine that, if I were from Portugal, I would prefer to adopt
the (ex-BR) "fato", simply for consistency with "atualmente", given that
I have to unlearn my familiar (ex-PT) "actualmente" anyway (it is no
longer valid in ao1990). A more conservative user in PT might stick to
"facto", accepting that he has to remember that it's an exception.

  I can imagine that, if I were from Brazil, I would prefer to switch
over to the (ex-PT) "génio", simply because the "é" takes one less
keystroke than the "ê" (dead key acute + e versus shift key + dead key
circumflex + e). A less lazy user in BR might stay with the (ex-BR)
"gênio", accepting more wear on his fingers.

  And if I were living in Belgium or France, I would most certainly
prefer "génio", because there I'd have an AZERTY keyboard in front of me
where the "é" has its own key, whereas the "ê" takes two strokes.

  Under ao1990, I have the freedom to mix and match, so why shouldn't I?

  Now, with only two words, you already have four possible "user
variants". I don't know how many of these words there are, but with just
ten of them, there are 1024 possible combinations, and with 20 you're at
1,048,576 in no time. Of course these words come in sets, so one million
is probably and exaggeration, but still...

  Are you saying that all of these "user variants" must be registered
instead of ao1990?

  Yes, the above is speculation, since I am not a writer of Portuguese,
but at least it's based on my experience with Dutch. There we had the
advantage of having pre-defined subsets. For each word, one of the
spellings was marked as "preferred", which means that it was mandatory
for use in government etc. Otherwise, you were free to pick and choose,
and many users did exactly that, for reasons similar to the above.

  For example, the "preferred spelling" tended to be with "c-", the
other one with "k-" ("cultuur" vs. "kultuur") but there were
inconsistencies. Therefore many users selected all the "c-" forms, or
all the "k-" forms, thereby creating their own spelling subsets. Others
would write "kultuur" with k but "copie" with c, for reasons that I
don't need go into.

   Nor are (pre-2005) Dutch and (post-oa1990) Portuguese the only such
languages. In Greek, both "αδερφός" and "αδελφός" are equally valid for
"brother", and there are many others. Upon reflecting on my own writing,
I find that I happily write "νεκρός αδελφός" (with -λ-) when referring
to the "Ballad of the Dead Brother" by Mikis Theodorakis, and "μεγάλος
αδερφός" (with -ρ-) when referring to "Big Brother". Don't ask me why.
Given that both are legitimate, why would you want me to stop doing
that, and if you don't, how would you propose that I tag it?

  Back to Dutch. The current spelling standard says to write "kerk"
(church) with upper-case "K" when referring to the institution, and with
lower-case when referring to the building. Would you refuse a subtag for
the post-2005 spelling because it is "an umbrella" that allows both, and
if so, how do you propose that this "problem" can be solved by more
"precise" subtags?

   Back to Portuguese. If you want facts, not speculation, these could
be found rather easily. I read that seven of the major newspapers in
Portugal have switched to ao1990. Newspaper editors tend to care about
such things, and they would have had meetings where considerations such
as the above were discussed. Just find out what they came up with. If
these seven "user variants" turn out to be identical (and if their
editorial staff is really independent), then you'd have a case for
pt-PT-ao1990. But I wouldn't be surprised if you came up with seven
different "user variants" instead.

   Next, turn to the Portuguese government gazette. I read that it
switched to ao1990 in 2012. I wouldn't be surprised if you found all of
the (allowed) variants in there, depending on the particular lawmaker
that drafted the law that eventually got published. So you would have no
choice but to tag the government gazette with "ao1990", and it would be
just as well to approve its registration.

   Luc Pardon

More information about the Ietf-languages mailing list