What Doug says is definitive and correct. At 19:55 -0600 2008-06-26, Doug Ewell wrote:
Kent Karlsson <kent dot karlsson14 at comhem dot se> wrote:The subtag 'ase' will replace "sgn-US" as the (sub)tag meaning "American Sign Language". You can then use region subtags with it as needed. E.g. ase-US or sgn-ase-US (based on the current draft).I still say that it is not even a possible interpretation of RFC 3066 to remove the region code in these cases, no matter how the sgn-* tags were discussed at registration time.You are wrong. Sorry to be so discourteously direct, but let facts be submitted to a candid world.The "US" part in "sgn-US" is and remains a region code, and the appropriate target code (for Preferred-Value of the to-be-deprecated tag sgn-US) should be "asl-US".The "US" part is there to modify and narrow down "sgn", to show that it means not merely "sign languages" but specifically "American Sign Language." It is NOT functioning as a true region code. It is actually functioning just like an extlang, although that concept had not yet been formalized in 2001. Forget about the syntactic point that "US" is two letters long, and focus on the parallel semantic relationships:"zh" means "Chinese." "zh-yue" means "not just any Chinese, but specifically Cantonese." "sgn" means "sign languages.""sgn-US" (as registered) means "not just any sign language, but specifically ASL."These tags, registered by ietf-languages in 2001, used two-letter ISO 3166 codes in an extremely loose sense of the term "region identification." They don't refer to regional variations of a single language, as is the case for normal generative tags; rather, they refer to *completely different languages*. This is the change in semantics that Peter Constable was referring to.By arguing that the Preferred-Value for "sgn-US" should be "asl-US", you are arguing that the "US" in "sgn-US" actually modifies "sgn" in TWO ways:1. to indicate "American Sign Language" 2. to indicate the region "United States" which is a position I find completely untenable.Consider the special rule added to RFC 3066, Section 3, that attempted to lend legitimacy to these tags:"This procedure MAY also be used to register information with the IANA about a tag defined by this document, for instance if one wishes to make publicly available a reference to the definition for a language such as sgn-US (American Sign Language)."Notice that this is phrased in terms of making a reference publicly available, not changing the definition of "a tag defined by this document" from its generative meaning to something else, but in fact it does change the meaning. I know you understand this, because you do not claim that "sgn-US" as registered means "any old sign language as used in the United States," which would be its generative meaning.Notice, too, that this rule has never been invoked for ANY other language -- no matter how well-known and exhaustively documented, no matter how obscure and in need of documentation -- except for the batch of sign languages registered in 2001, shortly after RFC 3066 was published. And notice that this rule no longer exists in RFC 4646, or in draft-4646bis. It was only in RFC 3066 BECAUSE of this batch of tags that were coming down the pike, that DID use two-letter ISO 3166-based second-position subtags to mean something OTHER than "as used in such-and-so country."You are right in thinking that "xxx-US" for some language 'xxx' SHOULD mean "language xxx as used in the United States." In all *generative* tags going back to RFC 1766, that is exactly what they mean, no exceptions. That is exactly why Peter and I and I'm sure others dislike these nonstandard sign-language tags so much, and can't wait to replace them with generative tags that mean just what they say.You can choose not to believe the Description field in the Registry, nor the original registration form from 2001 which is still archived by IANA (http://www.iana.org/assignments/lang-tags/sgn-US), and you can choose not to take into account the intent of the requester, but the fact remains that the "US" in "sgn-US" is NOT serving double duty as both a de-facto extlang AND a region qualifier. No subtag has ever been asked to serve double duty like that.I am cc'ing Michael, the original requester, again in hopes that he will have something definitive to say, because evidently what Peter and Addison and I are saying is not definitive enough."asl-CA" (e.g.) is simply not covered with tags in the current LSR, along with 100+ other sign languages (plus regional variants of those sign languages) that will be taggable using 639-3 codes.This part, at least, is true. -- Doug Ewell * Arvada, Colorado, USA * RFC 4645 * UTN #14 http://www.ewellic.org http://www1.ietf.org/html.charters/ltru-charter.html http://www.alvestrand.no/mailman/listinfo/ietf-languages ?
-- Michael Everson * http://www.evertype.com _______________________________________________ Ltru mailing list Ltru at ietf.org https://www.ietf.org/mailman/listinfo/ltru
Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.