2012/2/21 Michael Everson everson@evertype.com

On 20 Feb 2012, at 18:23, Osmar Valdebenito wrote:

...

But there are several problems, especially with Mapudungun Wikipedia.

Currently, ISO code for Mapudungun is arn from Araucanian, an offensive word used by the Spanish and Chilean conquerors till the past century.

What is offensive about it? To an English-speaker it simply looks like some sort of place-name.

I'm not a Mapuche person so I'm not the one to say why are they offended. One of the etymologies for Araucanian is "savage", "barbarian" and "enemy", used first by the Incas and then by Spanish and Chilean conquerors. It is not longer used anymore and today is totally outdated. Not in English, not in Spanish. For me, it is not a big deal... but, as I said, I am not a Mapuche person... but I know it is important for them and we must deal with their sensitivities. Using "araucano" or any derivation would be like using the N-word, trust me. But these sentivities are important for all our movement, especially if we are dealing with different cultures and different backgrounds. We have seen those cases before with, for example, the portraits of Muhammed or other issues when our culture clashes with other.

...

That name is also used by Unicode for their translation (and so, it is

used in our {{#language}} magic word).

Unicode does not make "translations", so I do not know what you mean. I also not know what "it is used in our {{#language}} magic word)" means.

Sorry, I meant localization. {{#language}} is a magic word for MediaWiki that shows the localized name of a language. {{#language:es}} would be <Spanish> at English Wikipedia, <Español> in Spanish Wikipedia and <Wingkazugun> in Mapudungun Wikipedia, just to name few. Currently {{#language:arn}} says <Araucanian> and <Araucano> for English and Spanish and, as I said before, it is not only pejorative and offensive but also outdated and wrong. In fact, as you said earlier, the name used by the ISO code for arn in English is Mapudungun and not Araucanian. To be honest, this is a minor issue and I just wanted to correct something it is wrong. The real problem is with the ISO code.

...

The use of the code for the community is a serious issue... nobody would

like to work in a project with an offensive word attached to it.

Standardization bodies have to deal with stability. If you retire a code it can't be used by any one else. And if "arn" were retired and something else were established for Mapudungun, that would result in one less three-letter code available for *other* languages which do not yet have codes. Those minorities also deserve a chance, do they not?

Although it is true that standarizations have to deal with stability, it also tries to reflect the truth. There have been several changes for different ISO standards; not sure about ISO 639 about languages (it is difficult that a language can change) but for example, there have been a lot of changes on ISO 3166 about contries: HV changed to BF when Alto Volta was renamed Burkina Faso, for example.

Considering 26 letters on the alphabet and three letters per code, there are 17.576 different combinations. Currently, less than 500 are being used. I'm pretty sure a change on the code won't affect other minorities.

...

We have been trying to contact the ISO authority and also the US Library

of the Congress (which is in charge of ISO 639-1) but we didn't get an answer.

LOC is in charge of 639-2, which is relevant. And if you look at http://www.loc.gov/standards/iso639-2/php/code_list.php you will see that by the code "arn" the name "Araucanian" does *not* appear.

No, Araucanian is not longer used, but the code is. And that is the code every Mapudungun speaker will use on their own Wikipedia. And "arn" doesn't seem like random letters to asign that language. Of course, they will (and are) upset.

...

I know that the Language Committee has an strict rule about the use of

ISO code but I think this is an exceptional case where the use of that code is a really sensitive issue that could damage any development.

Why is this "really sensitive"? And what specific remedy do you want?

Michael Everson * http://www.evertype.com/

I think I said before why it is sensitive and shouldn't be used. My proposal would be to look for an alternative language code for use on Wikimedia projects instead of arn. There are available codes at ISO (for example, mpg or mpn) although they can be used on the future. Other codes (for example, map) is used for a group of languages, not for a single language, so it probably won't be used on Wikipedia. And other option could be using a diferent code, for example mapu, that won't clash with any other project.