Document: ISO/IEC JTC1/SC22/WG20 N651 Source: John Clews, Committee Chair: ISO/TC46/SC2: Conversion of Written Languages Title: Ordering labio-velar characters in Ethiopic Action: For noting when Unicode 3.0 and ISO/IEC 10646 2nd edition characters are added to ISO/IEC 14651. In corresponding with experts from Ethiopian Standards body regarding ISO NP 15921: Generalized conversion methods, being developed in ISO/TC46/SC2, I have been having a long and very cooperative discussion with Daniel Yacob in Addis Ababa about transliteration of Ethiopic. The aim has been to develop a case-insensitive transliteration for Ethiopic script from the widely used case-sensitive SERA transliteration, to be added to other case-insensitive transliterations in ISO NP 15921: Generalized conversion methods, which is being developed in ISO/TC46/SC2/WG8: Transliteration and Computers. Ethiopia is now a P-member of that committee. As part of these discussions, it is clear that a variation from the UCS order is required for Ethiopic script. Probably if any of us knew this beforehand, this might have meant that in UCS the labio-velar characters would have been coded slightly differently, but it's obviously too late now to change UCS. However, it is worth noting in the appropriate stage of ISO/IEC 14651. ---------------------------------------------------------------------------- Order 1st 2nd 3rd 4th 5th 6th 7th 8th 10th 11th 12th 9th +----------------------------------------------------------------------- UCS | -0 -1 -2 -3 -4 -5 -6 -7 -8 -A -B -C -D +----------------------------------------------------------------------- 120- he hu hi ha hie h' ho 1 le lu li la lie l' lo lwa [+8] 2 121- hhe hhu hhi hha hhie hh' hho hhwa 3 me mu mi ma mie m' mo mwa [+8] 4 122- sze szu szi sza szie sz' szo szwa 5 re ru ri ra rie r' ro rwa [+8] 6 123- se su si sa sie s' so swa 7 she shu shi sha shie sh' sho shwa [+8] 8 ---------------------------------------------------------------------------- 124- qe qu qi qa qie q' qo qwe qwi qwa qwie qw' 9 125- qhe qhu qhi qha qhie qh' qho qhwe qhwi qhwa qhwie qhw' 11 ---------------------------------------------------------------------------- Order 1st 2nd 3rd 4th 5th 6th 7th 8th 10th 11th 12th 9th +----------------------------------------------------------------------- UCS | -0 -1 -2 -3 -4 -5 -6 -7 -8 -A -B -C -D +----------------------------------------------------------------------- ---------------------------------------------------------------------------- The chart above shows the differences between the coding (in the line beginning "UCS") and the ordering (shown in the line beginning "Order") [Tip: Use a monospace font like Courier when reading/displaying this]. Only UCS row, UCS column and Order numbers are relevant here: other data is related to work in ISO/TC46/SC2 still at a draft stage. Only the final labio-velar character is affected (marked "9th") for each character, and not all syllable series have labio-velar characters: 1240-124F and 1250-125F are shown as examples above. However, similar departures for the default table will be required for all the appropriate characters in the Unidata tables in the Unicode Ordering Algorithm documentation, and in the ISO/IEC 14651 tables. In the remainder of this email, I reproduce excerpts of corespondence from Daniel Yacob to back this up. Daniel Yacob was one of the main authorities providing input to Joe Becker on the Ethiopic proposals in Unicode and ISO/IEC 10646 and has several years of experience in all aspects of Ethiopic script computing. I had noted that in UCS the ordering of the labio-velar characters almost matches, but not quite matches, the ordering of the non-labio-velar characters. So, rather than the order coded in UCS: ---------------------------------------------------------------------------- 124- qe qu qi qa qie q' qo qwe * qwi qwa qwie qw ---------------------------------------------------------------------------- Ethiopians would prefer this order, e.g. in the UOA and ISO/IEC FCD 14651: ---------------------------------------------------------------------------- 124- qe qu qi qa qie q' qo qwe qw qwi qwa qwie * ---------------------------------------------------------------------------- > Date: Sat, 13 Mar 1999 10:56:49 -0500 (EST) > From: Daniel Yacob > Message-Id: <199903131556.KAA29949@piano.cs.indiana.edu> > To: Ethiopic@sesame.demon.co.uk > Subject: Re: New Alternative transliterations for Ethiopic/Amharic Daniel Yacob replied: > qwe qw qwi qwa qwie > > is traditional. The order shown in Unicode is the main grievance here in > Ethiopia about the standard. The Unicode order for labiovelars was the > work of a foreign linguist (Dr. Gragg of the U of Chicago). It bewilders > people here. Given the nature of labialization "w" and "wu" are nearly > identical which is also why their collation gets confused. In SERA we let > both qWu and qW retransliterate back onto the same syllograph. John Clews, 31 March 1999 END -- John Clews, SESAME Computer Projects, 8 Avenue Rd, Harrogate, HG2 7PG Email: Converse@sesame.demon.co.uk; telephone: +44 (0) 1423 888 432 Committee Chair: ISO/TC46/SC2: Conversion of Written Languages Committee Member: CEN/TC304: European Localization Requirements Committee Member: ISO/IEC/JTC1/SC22/WG20: Internationalization Committee Member: ISO/IEC/JTC1/SC2: Coded Character Sets