Language Subtag Registration Form: variant "signed"

Doug Ewell dewell at
Sun Feb 26 03:13:53 CET 2006

Frank Ellermann <nobody at xyzzy dot claranet dot de> wrote:

> [...]
>>    Prefix: en-GB
>>    Prefix: en-IE
>>    Prefix: en-US
> [...]
> Hi, how about focussing on this part as indicated by your
> evidence ?  It's already interesting to discuss why you
> pick these three region prefixes out of dozens plausible
> combinations, notably en-AU, en-CA, en-NZ, and en.

According to Michael's page at:

these are signed spoken languages identified by SIL in Ethnologue.  Most 
or all of these are also identified on Gallaudet University's FAQ page.

The goal was to provide a tag using "-signed" corresponding to each of 
the tags Michael proposed that use "sgn-" plus an ISO 639-2 language 
code and ISO 3166-1 country code:

Signed Afrikaans (South Africa): sgn-afr-ZA
--> af-signed
Signed Chinese (Taiwan): sgn-chi-TW
--> zh-TW-signed
Signed Danish (Denmark): sgn-dan-DK
--> da-signed
Signed Dutch (Belgium): sgn-dut-BE
--> nl-BE-signed
Signed Dutch (Netherlands): sgn-dut-NL
--> nl-NL-signed
Signed English (United Kingdom): sgn-eng-GB
--> en-GB-signed
Signed English (Ireland): sgn-eng-IE
--> en-IE-signed
Signed English (USA): sgn-eng-US
--> en-US-signed
Signed Finnish (Finland): sgn-fin-FI
--> fi-signed
Signed French (Belgium): sgn-fre-BE
--> fr-BE-signed
Signed French (Canada): sgn-fre-CA
--> fr-CA-signed
Signed French (France): sgn-fre-FR
--> fr-FR-signed
Signed Japanese (Japan): sgn-jpn-JP
--> ja-signed
Signed Norwegian (Norway): sgn-nor-NO
--> no-signed
Signed Portuguese (Portugal): sgn-por-PT
--> pt-signed
Signed Swedish (Sweden): sgn-swe-SE
--> sv-signed

Note the distinction between (for example) Signed Dutch for Belgium and 
Signed Dutch for the Netherlands.  Since a difference between these two 
is asserted, and since Michael assigned two different tags, I assigned 
two different prefixes with built-in region subtags.  I do not 
personally stipulate that such a distinction exists; that is left to 
better scholars.  I merely provided encodings for the distinction that 
others have claimed.

I did not indicate a region for the others, since no indication was 
given that (for example) Signed Japanese for Japan would differ from 
Signed Japanese for somewhere else.  The sole exception was Signed 
Chinese for Taiwan, because the differences between varieties of spoken 
and written Chinese are well known and it seems likely that a putative 
Signed Chinese for China or Hong Kong or anywhere else would also 
differ.  In retrospect, I might have left this out, or alternatively, 
might have applied the concept to Portuguese as well.  This is why we 
have a review period.

> Is that so ?  What's the difference between say en-GB-signed
> and en-IE-signed ?  I'd prefer to skip the regions and just
> say  Prefix: en  if that's good enough.

If we assign "en" as a prefix, we cannot subsequently split it into 
"en-US" and "en-GB" and "en-IE".  (See Section 3.4, item 4.)  Doing so 
would exclude "en-AU" and "en-CA" and others, narrowing the set of 
prefixes rather than broadening it.  We can only start with a narrow set 
of prefix definitions and broaden it, or else leave it alone.

> The 3066bis example says that sl-IT-nedis is okay for the
> registered Prefix: sl

Yes, and if we indicated "en" as a prefix for "signed" it would be 
possible to write "en-AU-signed" without any indication that such a 
signed spoken language exists.  OTOH (devil's advocate), this is the 
same situation as for "en-AQ".  What does the rest of the list think: do 
these prefix distinctions help or hurt?

John Cowan <cowan at ccil dot org> wrote:

> I continue to believe that signed spoken languages should be
> handled with an extension so that both the lexical source
> (a SL) and the grammatical source (a spoken language) can be
> fully specified.

Then please propose an extension as per Section 3.7.  I'll work with you 
on it if you like.  On the LTRU list I got a strong sense from most 
participants (which I did not share) that extensions were a horrible 
hack that should never actually come to pass.

Doug Ewell
Fullerton, California, USA

More information about the Ietf-languages mailing list