Simplicissimus

Bolorsoft CEO and consultant discuss Unicode standard of Mongolian script – The UB Post
27 avr. 2018
By M.OYUNGEREL
▻https://www.pressreader.com/mongolia/the-ub-post/textview
(note : je reprends tout le texte ici car le format de Pressreader est très pénible, le UB Post n’a plus l’air d’être en ligne en version html et l’adresse ci-dessus n’a pas l’air très spécifique)

Mon­gol News sat down with Founder of Bolor­soft Co. S.Badral and Con­sul­tant of Bolor­soft T.Jamiyansuren to dis­cuss the in­ter­na­tional stan­dard for Mon­go­lian script. Last week, they at­tended the Uni­code tech­ni­cal com­mis­sion meet­ing in San Jose, Cal­i­for­nia, USA to im­prove the ex­ist­ing Mon­go­lian script stan­dard/pho­netic model.

$The public has just re­cently be­come aware of the con­tro­ver­sial sit­u­a­tion con­cern­ing the im­prove­ment of in­ter­na­tional stan­dard for Mon­go­lian script, which is whether to en­code with Mon­go­lian pho­netic model or with graphetic, like that used to en­code Chi­nese char­ac­ters. Since you have par­tic­i­pated in these dis­cus­sions, would you please give our read­ers in­for­ma­tion re­gard­ing this is­sue?

*S.Badral: Uni­code is a com­pany of in­ter­na­tional char­ac­ter en­cod­ing stan­dards. It’s an in­te­grated consortium of cor­po­ra­tions which de­velop the en­cod­ing stan­dards for all the scripts in the world. In other words, it pro­duces one com­pre­hen­sive stan­dard which identifies the com­puter codes for Latin “a”, Cyril­lic “a”, Mon­go­lian “a”, and Chi­nese char­ac­ters. If the script in ques­tion is not en­coded by Uni­code stan­dards, all the global play­ers, such as Face­book, Google, Adobe, Ap­ple, Mi­crosoft, and IBM, would not sup­port it. That means the script will not be sup­ported on any op­er­at­ing sys­tem, com­puter, or phone. Al­though Mon­go­lian script was first en­coded based on a pho­netic model in 2000, small un­solved is­sues have caused it to drag out with­out a so­lu­tion for 17 to 18 years.

So, dur­ing the Uni­code tech­ni­cal com­mis­sion meet­ing in Ho­hhot last Septem­ber, it was al­most de­cided that the cur­rent model is com­pletely wrong and a graphetic model de­vel­oped by Chi­nese ex­perts Liang Hai and Shen Yilei was nearly adopted. At that time, we flatly op­posed, say­ing, “Mon­go­lian script has let­ters, and it’s writ­ten by its pho­net­i­cally.

...In this last meet­ing, we have achieved our ob­jec­tive for the past few years and de­fended our her­itage...

There­fore, we need to im­prove the ex­ist­ing pho­netic model in­stead of adopt­ing graphetic en­cod­ing”. With sup­port from the In­ner Mon­go­lian party, the graphetic model was not ap­proved. In this meet­ing too. Our dis­cus­sion re­volved around aban­don­ing the pho­netic model and chang­ing to the graphetic model.

T.Jamiyansuren: Had we ap­proved the Uni­code stan­dard for Mon­go­lian script as the graphetic model that the Chi­nese de­vel­oped, it would’ve then been dis­cussed and ap­proved at the ISO in­ter­na­tional stan­dard meet­ing, and ev­ery­thing would’ve been over. Be­cause these two meet­ings were sched­uled right af­ter an­other, we tried very hard to not take it to the ISO meet­ing. That Uni­code tech­ni­cal com­mis­sion meet­ing was al­most like war.

Do any state rep­re­sen­ta­tives take part in these im­por­tant meet­ings? What is the par­tic­i­pa­tion of the Mon­go­lian state and govern­ment in this in­ter­na­tional dis­cus­sion con­cern­ing na­tional script and cul­ture?

S.Badral: Pre­vi­ously, rep­re­sen­ta­tives from the Agency for Stan­dard­iza­tion and Metrol­ogy and the In­sti­tute of Lan­guage and Lit­er­a­ture at the Mon­go­lian Acad­emy of Sci­ences reg­u­larly at­tended these meet­ings. But we don’t know why the is­sue has got­ten to this point. Be­fore we went to Ho­hhot in Septem­ber 2017, we viewed that the fu­ture of Mon­go­lian script re­lates to the na­tional in­ter­ests and in­tel­lec­tual in­de­pen­dence of Mon­go­lians and con­tacted the pres­i­dent. Pres­i­dent Kh.Bat­tulga then met us within 10 days of our re­turn, called the rep­re­sen­ta­tives of rel­e­vant or­ga­ni­za­tions, and or­dered them to ur­gently take nec­es­sary mea­sures. We de­ter­mined that an ac­tion plan to im­prove the Mon­go­lian scripts pho­netic model will be de­vel­oped by the Com­mu­ni­ca­tions and In­for­ma­tion Tech­nol­ogy Author­ity (CITA) and ap­proved by the Agency for Stan­dard­iza­tion and Metrol­ogy, and formed a work­ing group.

How­ever, the work­ing group did noth­ing be­cause they didn’t have a bud­get. The Na­tional Se­cu­rity Coun­cil obliged them to send a re­port ev­ery week, but we have no idea what re­port was given or what work was done. The agency tried to dis­cuss the fund­ing is­sue in a govern­ment meet­ing, but was post­poned. By then, the bud­get dis­cus­sion had al­ready been con­ducted, hence, no so­lu­tion. Ba­si­cally, they took this is­sue very idly.

Govern­ment Memo No. 54 was passed. In there, they as­signed six agen­cies to take care of the ex­e­cu­tion and fund­ing of this is­sue, three for each. While the Mon­go­lian script en­cod­ing im­prove­ment is­sue was bounced be­tween state or­ga­ni­za­tions like a ten­nis ball, it was time for the sched­uled meet­ing.

T.Jamiyansuren: Ap­prox­i­mately 20 days be­fore we left, an­other work­ing group was es­tab­lished by the Min­istry of Ed­u­ca­tion, Cul­ture and Sci­ence, and they held a meet­ing. Dur­ing that meet­ing, there was some kind of talk, “What do we do . Ei­ther we give S.Badral and T.Jamiyansuren a state as­sign­ment, or no­tify them that they do not rep­re­sent the state”. A let­ter was sent to the Uni­code tech­ni­cal com­mis­sion stat­ing, “These two men do not have the right to rep­re­sent the state,” and that a per­son named Enkhdalai will be com­ing. Those two men have just re­turned from par­tic­i­pat­ing in that meet­ing. But po­lit­i­cal sug­ges­tions and con­clu­sions do not af­fect the Uni­code tech­ni­cal com­mis­sion meet­ing. They sent a per­son called Enkhdalai with a po­si­tion of con­sul­tant at CITA. We were in­tro­duced at the meet­ing.

Did the Mon­go­lian rep­re­sen­ta­tives first met each other once they were at the meet­ing?

S.Badral: Yes. We had ar­rived two days be­fore, met our trans­la­tor, and care­fully pre­pared the is­sues of con­cern and presentations. Uni­code tech­ni­cal com­mis­sion chose Gan­ba­yar Gan­sukh (G.Gan­ba­yar), a man who lives in Ok­la­homa, as our trans­la­tor. On our end, it was un­clear who was go­ing to voice the state po­si­tion even a week be­fore the meet­ing.

T.Jamiyansuren: Uni­code Consortium be­lieved that CITA is of im­por­tance and had them at­tend as a li­ai­son mem­ber. The re­spon­si­bil­i­ties should be even higher in re­gard to this.

Did you in­tro­duce a Mon­go­lian pho­netic model that you de­vel­oped your­selves, dur­ing the meet­ing?

S.Badral: How could we have the Mon­go­lian script en­coded graphet­i­cally, like that of Chi­nese char­ac­ters on our watch? So, I and T.Jamiyansuren dis­cussed and co­op­er­ated with Mon­go­lian script ex­pert Lkhag­va­suren, and pre­pared a pre­sen­ta­tion on ways to im­prove the pho­netic model. We dis­cussed the dis­ad­van­tages of the ex­ist­ing model and ways to fix it, and proved it with a re­al­is­tic ex­am­ple. Af­ter three days of the meet­ing, the Uni­code tech­ni­cal com­mis­sion had a pos­i­tive at­ti­tude and said, “If we re­ally fix it like this, the model will be eas­ier and bet­ter”. That’s be­cause the graphetic model con­tains only char­ac­ters and not let­ters, which makes it com­pli­cated to sort, cat­e­go­rize, and de­velop etc. There would be many prob­lems such as iden­ti­fy­ing the text and spellcheck­ing it. The pro­posed graphetic model was not even for the clear Mon­go­lian script drawn, so it was hard for the user to write with the word in mind or even the root of the word . For in­stance, the “a” and “e” at the be­gin­ning, mid­dle, and end of a word, and “n”, were to be writ­ten by press­ing one “teeth” or aleph, and the “crown”, “tooth”, and “tail” (el­e­ments of Mon­go­lian script writ­ing) were to be au­to­mat­i­cally man­aged. This might break the Mon­go­lian think­ing and one but­ton will eas­ily break from too much pres­sure. I think the Uni­code tech­ni­cal com­mis­sion peo­ple started to un­der­stand it. The In­ner Mon­go­lians on the other hand, sug­gested to de­velop both the graphetic and pho­netic mod­els, maybe be­cause they were in a rush to de­cide on some so­lu­tion to pre­vent fur­ther drag out, or they lost faith in the pho­netic model.

Any­how, main­tain­ing the pho­netic model which was to be aban­doned, fix­ing its bugs, and hav­ing a de­ci­sion made for it to be de­vel­oped with pref­er­ence is a big achieve­ment.

So the rep­re­sen­ta­tives lis­tened to you and ac­knowl­edged your pre­sen­ta­tion. Isn’t the mes­sage “They don’t rep­re­sent the state” a way of sav­ing their skin in case some­thing went wrong?

S.Badral: It just looks like that. Since last Septem­ber, that’s the stance our govern­ment held. In the first work­ing group meet­ing of the Min­istry of Ed­u­ca­tion, Cul­ture and Sci­ence, CITA rep­re­sen­ta­tives kept say­ing, “It’s not suit­able for pri­vate com­pany rep­re­sen­ta­tives be in­cluded in a state work­ing group, that’s pri­vate party in­ter­est,” so I even re­moved my­self from the work­ing group (laughs).

T.Jamiyansuren: Be­cause the is­sue couldn’t have been qui­eted down, they had to send some­body as a rep­re­sen­ta­tive, which hap­pened to be Enkhdalai, as a con­sul­tant at CITA. I had the im­pres­sion that that per­son read and re­searched quite a lot too. But it’s not ef­fec­tive to have some­one who’s in­ter­ested in the Uni­code stan­dard of Mon­go­lian script read ready pre­pared ma­te­ri­als and retell them. You have to be metic­u­lous with your words, and be able to prove your point. His few un­sure words dur­ing his pre­sen­ta­tion had Chi­nese ex­perts stand up and say, “This is ex­actly why there should be a graphetic model”. Of course, it was not easy to give this much work in tight sched­ule to a per­son who lacked ex­pe­ri­ence.

Peo­ple are say­ing that the Mon­go­lian govern­ment will pay at­ten­tion to our na­tional script be­fore the next Au­gust meet­ing to pre­vent the graphetic model from be­ing en­coded. Is there re­ally such dan­ger, or is ev­ery­thing be­hind us now?

S.Badral: If we hadn’t given a pre­sen­ta­tion at this last meet­ing and changed the com­mis­sion’s un­der­stand­ing, our Mon­go­lian script re­ally would have been en­coded graphet­i­cally. But now, the Uni­code tech­ni­cal com­mis­sion has de­cided to de­velop both mod­els to ra­tio­nally solve the is­sue. That means we have to fix and im­prove our pho­netic model and in­tro­duce it to use. Even a sin­gle sym­bol can­not be changed once it’s reg­is­tered in the Uni­code stan­dard. There­fore, there’s a strict rule that we have to fix with­out chang­ing the pre­vi­ous one. In the meet­ing, we in­tro­duced a pos­si­ble so­lu­tion that we can im­prove it like that. So our govern­ment has to take care of this is­sue for this to con­tinue on a big­ger scale. We wouldn’t beg them if it was only tech­ni­cal work, but it’s re­lated to so­ci­ety, cul­ture, and pol­i­tics.

T.Jamiyansuren: Some who un­der­stands the sig­nif­i­cance of this meet­ing are right when they say, “This was like the mod­ern Khi­agt agree­ment”. This is a mat­ter of whether Mon­go­lian script will ex­ist for the next five years, 500, or 5,000. When the rep­re­sen­ta­tives were asked for their opin­ion on the lo­ca­tion and time of the next meet­ing, In­ner Mon­go­lians sug­gested to ur­gently hold it in Ho­hhot, af­ter two months. But the Uni­code tech­ni­cal com­mis­sion head said, “The next meet­ing will be held at least six months later. There’s a sug­ges­tion to or­ga­nize it in Ulaan­baatar,” while our state rep­re­sen­ta­tive stayed quiet. We couldn’t, so we voiced our opin­ion. In other words, there will be a Uni­code meet­ing re­gard­ing this is­sue in Ulaan­baatar, at the end of Septem­ber. If it’s or­ga­nized well, it’s not ours but Mon­go­lia’s name that will bear the good name.

It seems like we are greedy, hear­ing that the Chi­nese have de­vel­oped the graphetic model and en­code the Mon­go­lian script, when we don’t even use the Mon­go­lian script our­selves. There are some who say to stop co­op­er­a­tion with the Chi­nese and de­velop the script alone. What do you say about this?

S.Badral: Lan­guages ex­ist through the use of its script. In the mod­ern times, peo­ple’s writ­ing has trans­ferred from hand­writ­ten notes to typ­ing on a com­puter or a phone. As for Mon­go­lian writ­ing, it has slipped into the list of en­dan­gered lan­guages be­cause there is no dig­i­tal us­age and no op­por­tu­nity to cre­ate con­tent. How can dig­i­tal con­tent be cre­ated when the Uni­code stan­dard of 18 years has a big prob­lem. There­fore, this is­sue will be im­me­di­ately fixed and in­tro­duced into use like Cyril­lic and Latin al­pha­bets. In terms of pos­ses­sion, it shall be­long to those who use it. In other words, Mon­go­lian script is not the prop­erty of Outer or In­ner Mon­go­lians. There could be Amer­i­cans, Ger­mans, British, Bel­gians, Ira­ni­ans, or peo­ple of any other coun­try who have stud­ied and uses Mon­go­lian script. They have the right to learn and use what­ever lan­guage and script they please. It’s im­por­tant we pro­vide that op­por­tu­nity. That’s why these in­ter­na­tional rep­re­sen­ta­tives are putting this much ef­fort into in­tro­duc­ing the Mon­go­lian script in dig­i­tal use. This script is a very im­por­tant world cul­tural her­itage.

Why are you putting this much ef­fort and heart for the Mon­go­lian script?

S.Badral: As for me, I’m a mem­ber of the Uni­code tech­ni­cal com­mis­sion Work­ing Group (WG)-2 and vol­un­teer of the WG3. I re­ceive in­for­ma­tion about this be­fore oth­ers. I’ve seen this as my civic du­ties and re­ported it to the state and govern­ment. Sec­ondly, Bolor­soft is a dig­i­tal lin­guis­tics com­pany. Mon­go­lians know that we have re­leased many prod­ucts re­lated to Mon­go­lian lan­guage and script. Al­though Cyril­lic writ­ing pro­grams are in the mar­ket, most users don’t know that it is based on Mon­go­lian script. That’s why we can’t aban­don it.

Bolor­soft Co. is con­sid­ered a ma­jor provider in the de­vel­op­ment of Mon­go­lian script at Uni­code Consortium. That’s be­cause we were the first to cre­ate the Uni­code font for Mon­go­lian script and have it li­censed. This field was stag­nant since 2013, un­til we solved the Uni­code stan­dard for Mon­go­lian script is­sue. But those fonts be­came the be­gin­ning of big cor­po­ra­tions such as Google and Mi­crosoft. So, Uni­code Consortium al­ways invites us to their in­ter­na­tional meet­ings. We try to at­tend these meet­ings con­stantly to voice the in­ter­ests of Mon­go­lia, but we can’t al­ways due to the ex­pen­di­ture. But I see that there are peo­ple who are jeal­ous and spread ru­mors that we are try­ing to make money us­ing Mon­go­lian script. That’s the only thing they talked about in the last six months, politi­ciz­ing it. On the other hand, we are work­ing for Mon­go­lia’s in­ter­ests.

If we were seek­ing profit from this, we wouldn’t be us­ing ar­ti­fi­cial in­tel­li­gence to de­velop Mon­go­lian lan­guage and writ­ing, but fi­nan­cial de­vel­op­ment. We are one of the first Mon­go­lian com­pa­nies de­vel­op­ing and us­ing ar­ti­fi­cial in­tel­li­gence. In this last meet­ing, we have achieved our ob­jec­tive for the past few years and de­fended our her­itage.