"Turkish cannot go back to Arabic script"

Russell McGorman

interesting

“We had a language which was very suitable for science; yet we slept over it one night and the next morning it was gone. Now we have been dragged down to the level of a country which learns and teaches science in foreign languages. Thousands of words and languages are forgotten. The structure of the language, which used to be suitable for deriving new words and expressions, was curbed.”
These sentences belong to Turkish President Tayyip Erdoğan. He was delivering a speech on Dec. 24 in the award ceremony of Turkey’s Scientific and Technological Research Board (TÜBİTAK).
It was not the first time Erdoğan has mentioned this subject, though not as clearly as this time. It is important, since the remark was made in the wake of the debate to make the Ottoman language a compulsory course in the Turkish curriculum by the National Education Ministry.

AzizMostafa AzizAli

Compared to Latin, Arabic script is smarter, easier and more attractive.

Latin letters = 34 standalone shapes = 26 (A-Z) + 8 (a+b+d+e+g+h+q+r) not taking the differneces in the other 18

Arabic letters - Dots = only (19) sweet and dancing shapes.

Unlike Latin, Arabic letters love to go hand in hand and shoulder to shoulder as elaborated in the attached QB-English.pdf

So, why the Arabic Jawi script was replaced with Latin?

How do historians justify that decision?

Hope Turkey as well as Malaysia will go back to Arabic script. http://www.utusanmelayu.com.my

k.l.

(Historians do not make any such decisions. They merely reconstruct what happened and – with luck – why. Justifying anything is not their job.)

Christian Thalmann

Compared to Latin, Arabic script is smarter, easier and more attractive.

OK, I'll take the bait.

The first statement is debatable. Arabic script is extremely well suited to semitic languages with their reliance on TRY-CNSNNTL RWTS; not so much for most other languages where vowels are treated as first-class citizens of the phonology.

The middle one is blatantly wrong. Just think about how much effort it takes to make Arabic look right on screen: The letters twist and mutate and sit on each other. For Latin, all you ever have to learn is the individual letters.

The third statement is entirely subjective. I'd say it is attractive in a different way. In practice, the beauty of Arabic writing is often hidden under technical difficulties (such as everything looking really tiny, with the information-critical dots running together and being reduced to single pixels). I suppose modern fonts are changing that more and more nowadays.

The biggest counter-argument to Arabic script as used for Arabic, of course, is that it's illegible for people who don't know the Arabic language, including Arabic learners. I guess that could be different for Arabic script as used for Turkish language, but then you have tons of vowel signs floating around and exacerbating the already technically challenging vertical spread of the script.

Khaled Hosny

Without debating which script is better than which as I find it pointless; I think the issue of vowel representation is a red herring as one can always write the vowel using full letters, modern Uyghur orthography (which is a Turkic language, BTW), does that. Arabic speakers often do that when transcribing foreign words, though it is frowned upon as it goes against the rules of the orthography.

Digital fonts being ugly (mostly old system fonts done by or for USA corporations) is not any fault of the script itself. Making beautiful Arabic fonts is not as difficult as some make it sound, if me, a complete novice with no training in type design can make it with nothing but FontForge at my disposal, then anyone can do. People are either making excuses for their own incompetence, or trying to make what they do look more important than what it actually is.

Hrant Հրանդ Փափազեան Papazian

Just ask Azerbaijan how many times they've switched, between three scripts...

AzizMostafa AzizAli

By all means Dear Christian Thalmann.

> Arabic script is extremely well suited to semitic languages...; not so much for most other languages where vowels are treated as first-class citizens of the phonology.
@ That's how Arabic words are smartly made shorter in length and that's how the technically challenging vertical spread of the script is tackled.

> Arabic letters twist and mutate and sit on each other. For Latin, all you ever have to learn is the individual letters.
@ Arabic letters change tails not heads making self-explanatory and space-saving ligatures and that's how the technically challenging vertical spread of the script is tackled.

> In practice, the beauty of Arabic writing is often hidden under technical difficulties...
@ Beauty of Arabic can not be hidden under technical difficulties?!

> ... for Turkish language, but then you have tons of vowel signs floating around and exacerbating the already technically challenging vertical spread of the script.
@ Without exacerbating the technically challenging vertical spread of the script, Arabic letters are multiplied to cover Turkish and other languages simply by adding one or more dots instead of placing tons of vowel signs.

Happy exploring.

Andrew Wood

Christian Thalmann said:

 Arabic script is extremely well suited to semitic languages with their reliance on TRY-CNSNNTL RWTS; not so much for most other languages where vowels are treated as first-class citizens of the phonology.

And, of course, Turkish is not a semitic language. It's closer to Korean and Finnish than Arabic. Perhaps hangul would be a better choice.

AzizMostafa AzizAli

Why hangul?!  The majority of Turks are Muslims. They do recite the Glorious Quran in Arabic!
https://www.quora.com/Do-Turks-recite-Quran-in-Arabic

Christian Thalmann

@ That's how Arabic words are smartly made shorter in length and that's how the technically challenging vertical spread of the script is tackled.

For most non-semitic languages, removing vowels is not a «smart» move.

@ Arabic letters change tails not heads making self-explanatory and space-saving ligatures and that's how the technically challenging vertical spread of the script is tackled.

Many Arabic letters look quite different in their different versions. It's a bit like saying that Latin capitals are just large versions of the lowercase letters.

@ Beauty of Arabic can not be hidden under technical difficulties?!

Really? Do you think it's not possible to make an ugly Arabic font, or to display an Arabic font at such a small size that it becomes annoying to decypher?

Take even just this small sample from the URL of the Arabic Wikipedia:

The Latin letters are sized to be small but comfortably legible, whereas the bodies of the Arabic letters are compressed into about half the size and the dots melt into each other.

This is not the fault of the Arabic script in general, but of the Latin-biased technical implementation of the two scripts and the Latin-biased design of the browser.

I suppose both are getting better nowaways with more quality fonts becoming available, but it's still an existing problem.

> ... for Turkish language, but then you have tons of vowel signs floating around and exacerbating the already technically challenging vertical spread of the script.

@ Without exacerbating the technically challenging vertical spread of the script, Arabic letters are multiplied to cover Turkish and other languages simply by adding one or more dots instead of placing tons of vowel signs.

The Uyghur use of the Arabic script that Khaled linked to is a good examle of why the benefits of Arabic script for writing Arabic do not translate well to other languages. You end up with a small number of base glyphs that are turned into a large number of individual letters by a dense cloud of tiny diacritics. That moves the information content of the text into the most vulnerable and least legible part of the glyph (namely, the tiny wiggles floating around), sort of like Vietnamese (IMHO a very poor use of the Latin script). Also, all these individual wiggles defy the purpose of a fluid connected script in the first place.

Latin scripts that use diacritics usually only use one or two types of accents, so readers only need to see whether there is an accent present rather than what its exact shape is. That leaves the main volume of information encoded in the letters themselves, which take up most of the available vertical space.

Christian Thalmann

AzizMostafa said:

Why hangul?!  The majority of Turks are Muslims. They do recite the Glorious Quran in Arabic!
https://www.quora.com/Do-Turks-recite-Quran-in-Arabic

It's not a surprise that they use Arabic script for Arabic language — that's what it's made for! The question is what script is most fitting for Turkish, which they presumably use much more often in daily life than Arabic. 

(Also, Turkey is officially still a secular country with separation of religion and state, despite Erdoğan's efforts.)

Andrew Wood

AzizMostafa said:

Why hangul?!  The majority of Turks are Muslims. They do recite the Glorious Quran in Arabic!
https://www.quora.com/Do-Turks-recite-Quran-in-Arabic

I've studied both Turkish and Korean. Their grammars are similar. Why not use hangul: it's very easy to learn, and it's designed for a language in the same family as Turkish.

André G. Isaak

Korean and Turkish are not generally considered to be in the same family. At one point it had been proposed that the Altaic family, consisting of Turkic, Mongolic, and Tungusic, should be extended to include Korean and Japanese, but that was highly controversial and never gained widespread acceptance. Today, the Altaic family itself is rejected by the vast majority of linguists — the grammatical similarities between these various language groups (even Turkic and Mongolic which formed the core of Altaic) don't appear to be the result of common ancestry.

John Savard

It is certainly true that one can't claim Latin script is so much better than Arabic script in an absolute sense that the Arabic world should switch. Even if it seems more legible to people used to the Latin script, that doesn't prove it's a problem for the people who are used to it.

The thing is, the Turkish people are used to the Latin script now, so switching over would be just as hard for them as it would be for speakers of English. So I think I can say that it would be a really bad idea.

In any case, Erdogan's comment makes no sense. Changing the script did not make Turkish as a language less suitable to doing science. Yes, using the Latin script has made English and other Western languages more accessible to Turks, and so they are borrowing scientific terms from those languages. How is this a bad thing? Scientific terms in English don't come from Latin roots, they're borrowed from Latin and Greek, and this hasn't reduced the English language to an ignoble position. Although not borrowing from other languages hadn't hurt science in Germany or China, it must be admitted.

Erdogan appears simply to be preaching insane xenophobic nationalism. However beautiful and easy to write the Arabic script may be, I cannot support or encourage such thinking.

John Savard

Christian Thalmann said:

(Also, Turkey is officially still a secular country with separation of religion and state, despite Erdoğan's efforts.)

Yes, but its people are overwhelmingly Muslim. There is nothing wrong with Turkey being a Muslim nation in the same sense as the United States is a Christian nation, it would only be wrong if religious minorities were denied full equality.

Before Erdogan, Turkey had anti-clerical laws (like France and Italy do among Christian nations), and laws forbidding women from wearing the veil (think of Peter the Great forbidding beards). While there is much about Erdogan's regime and policies, changing these things were steps towards, not away from, compliance with the First Amendment, since Muslims are entitled to religious freedom too. The United States, unlike Italy, would not consider attempting to tell Catholic priests they can't preach sermons against abortion, Negro slavery, or American policy on immigration from Mexico.

If nearly all Turks are fully familiar with the Arabic script because they use it to read the Quran, that is a rational reason for writing Turkish in that script as well: why go to the bother of learning two writing systems, instead of just using one?

Of course, reading the Quran in Arabic, but in Latin transliteration, would perhaps be even easier, but this may be considered religiously disrespectful or something.

So, while my preceding post opposes the idea, I do have to admit there are two sides to the question.

AzizMostafa AzizAli

ChristianThalmann said:

> Many Arabic letters look quite different in their different versions. 

@ I say:

Basically, Arabic has only 17 letter forms + dots.

Medial forms = initial forms + short tails at the their right.

Final Forms = initial or Medial forms + longer tails at their left.

Out of any Latin font, you can easily make ALL the Arabic letterforms as follows:

pick @ (as is) for هـ ـهـ ـه ه and drop its petal for ـه ه

pick iïî (as is) respectively for نـ ـنـ ـن ن + تـ ـتـ ـت ت + ثـ ـثـ ـث ث  and v-flip them for بـ ـبـ ـب ب + يـ ـيـ ـي ي + پـ ـپـ ـپ پ

pick o (as is) for مـ ـمـ ـم م

pick ö (as is) for قـ ـقـ ـق ق. and drop one dot for فـ ـفـ ـف ف

pick q (as is) for و

pick S for كـ ـكـ  ـكـ 

pick j (as is) for ز ـز and  drop the dot for ـرر

pick c (as is) for عـ ـعـ ـع ع and center one dot above for غـ ـغـ ـغ غ

pick c + h-flip it for د ـد and center one dot above for ذ ـذ

pick b (as is) for ط and center one dot above the round part of it for ظ ـظ

pick V + rotate it 75 CCW forـح ح حـ ـحـ and center one dot above for خـ ـخـ ـخ خ and one dot below it for جـ ـجـ ـج ج

pick D + rotate it 90 CCW for صـ ـصـ ـص ص and center one dot above it for ضـ ـضـ ـض ض

pick m + flip for سـ ـسـ ـس س  and center 3 dots above it for شـ ـشـ ـش ش

ChristianThalmann said:

> Do you think it's not possible to make an ugly Arabic font, or to display an Arabic font at such a small size that it becomes annoying to decypher?

@ I say:

1. Do you think it's not possible to make an ugly Latin font if made Arabic-looking?

https://www.urbanfonts.com/fonts/arabic-fonts.htm

@ 2. Is not still readable that small sample from the URL of the Arabic Wikipedia?

@ 3. The Arabic letters are comfortably legible though compressed into about half that size is still readable and even if dots are removed as in the attached JawiMissingDots.pdf and contested here: http://www.emenang.com/community/index.php?/topic/54920-online-free-fill-jawi-missing-dot-contest/

ChristianThalmann said:

> The Uyghur use of the Arabic script that Khaled linked to is a good examle of why the benefits of Arabic script for writing Arabic do not translate well to other languages. https://en.wikipedia.org/wiki/Uyghur_Arabic_alphabe

@ I say: Stupid modernization for the old alphabet?!

John Savard said:

> If nearly all Turks are fully familiar with the Arabic script because they use it to read the Quran, that is a rational reason for writing Turkish in that script as well: why go to the bother of learning two writing systems, instead of just using one?

@ I say: Absolutely True.

Andrew Wood

John Savard said:

> If nearly all Turks are fully familiar with the Arabic script because they use it to read the Quran, that is a rational reason for writing Turkish in that script as well: why go to the bother of learning two writing systems, instead of just using one?

Koreans, Japanese and Chinese learn roman scripts. 

Russell McGorman

Here is the article that prompted me to start this thread: http://www.hurriyetdailynews.com/opinion/murat-yetkin/turkish-cannot-go-back-to-arabic-script-76063

I'm posting the link again... And more prominently this time, because when I revisited the thread, it took me a while to figure out where I got the quote from. 

Vasil Stanev

Going to a previous script would render a century of media content unreadable, and not just any century, but the 20th! It would push Turkey nearer to it's Arab neighbours and further from Europe, which is contrary to the country's politics. It's like if England returned to Insular scripts or if Japan replaced the kanas with the hanzi they are derived from.
Then again, this leads me to a compromise: I would welcome an app that transliterates Turkish text into the Arabic script, the same as when the Chinese type pinyin and the autocorrect suggests the hanzi. I think for Turkish children to be taught basic Arabic orthography would not be valueless, but then again this is not my call to make by a long shot.
And what version of Arabic?? The turkish alphabet is just that - localized in a country. But there are differences between different Arabic scripts (ASFAIK), and the language has moved past Ottoman Turkish. Who is to decide what the orthography should be?

Vasil Stanev

For the record, I have created beautiful latinizations of Arabic letters. The trick is to not use the Arabic at all, because it goes RTL, but to take a kalem (or your trusty tablet with a digital kalem brush) and write out the Latin alphabet in the style of arabic (mine was Nastaliq). That is what the creators of the fonts Mr. Mostafa linked to did not get, so they were bound to fail. They drove the cart butt-front. Of course it's going to crash! But a little calligraphy and respect for foreign cultures goes a long way. And I hope I do need to remind everybody of Palatino Arabic, do I?

Here you can see this principle applied by a much better hand to a classical Bulgarian children's book (The Tales of Sheherazad) that I grew up on. 




Fun fact: a big part of the Ottoman Imperial Library (some train wagons full) was sold as old paper by Ataturk... it now resides in our National Library and waits to share it's wonders.

John Savard

Vasil Stanev said:

It would push Turkey nearer to it's Arab neighbours and further from Europe, which is contrary to the country's politics.

Is it contrary to Erdogan's politics? It may be contrary to Turkey's position as a member of NATO, but Erdogan is seeking to enhance the role of Islam in Turkey, so that is the direction in which he is seeking to move the country.

However, this seems like such a far-reaching step that I am willing to consider the possibility that while he is talking about it as a possibility, as it is in line with the other things he is doing, he will not go to the point of implementation. Turkey, after all, does not have oil, so modernization and industrialization are priorities for that country no matter what the political orientation of its government may be.

Having to republish the technical literature of the 20th Century for the new generation would not help that.

John Savard

Christian Thalmann said:

The first statement is debatable. Arabic script is extremely well suited to semitic languages with their reliance on TRY-CNSNNTL RWTS; not so much for most other languages where vowels are treated as first-class citizens of the phonology.

While I do not think that Erdogan is interested in such a modified version of Arabic script for his country, that issue could be overcome the same way it was for the writing system of Yiddish.

One could switch from indicating vowels by subsidiary marks to indicating vowels by the same kind of script element as is used for consonants. Then one still has the beauty and ease of writing of the Arabic script without a mismatch to a Turkic language.

In Turkey, scientists trying to point out to the public errors and misstatments of fact in some elaborately printed books by a Creationist there have been successfully sued by him in the new climate created by Erdogan. So it would not surprise me if the idea of an alphabet change is to insulate Turkish science from borrowed ideas as well as borrowed words. For public consumption, he says that he wants the Turkish language to regain its dignity as an instrument of science; but the real agenda may be that he views science as a threat.

So I think one can admit that the Arabic script is graceful and beautiful, and that it was successfully used for Turkish before Ataturk came along, while being concerned because the intent may be to keep a new generation of Turks isolated and ignorant.

Vasil Stanev

It's simply a rambling of a madman elected into the highest office. Seems to be the rule nowadays.

RobPratley

Just a small off topic comment, Christian's comment on Vietnamese being a poor use of the latin script imho goes against his comments on Arabic use in Turkey. How does one qualify their judgement of a languages' use of a script without substantial linguistic and culturally relative knowledge? If the phonology is actually very well represented by that script with those 'wiggles' does it make it bad because one is used to that script without those 'wiggles'? I'd argue Vietnamese does rather well with the Latin script, if we excuse some circumstances by which that norm became norm, which are arguably outside of such discussions of 'poor use' of a script. Phonology and the symbols used to express that are two separate entities 90% of the time and phonology forms the basis for the language most of the time. We'd do well to remember that when getting into discussions about 'efficiency' or 'difficulty' of a script for specific languages. 

John Savard

RobPratley said:

Just a small off topic comment, Christian's comment on Vietnamese being a poor use of the latin script imho goes against his comments on Arabic use in Turkey.

I would consider your post on topic. Unfortunately, this discussion can't entirely be separated from politics, which is off-topic and controversial, since politics appears to be part of the motivation for a script change.

Since Christian Thalmann comments that the Arabic script is a bad fit to Uzbek or Turkish, it seems to me that saying the Latin script is a bad fit for Vietnamese is consistent with it.

I've noted that it's easy enough to use the Arabic script while modifying the writing system to be the same as that of Latin - the same thing was done for the Hebrew script to write Yiddish. So I do agree that the Arabic script is not seriously incompatible with the Turkish language, even if the way in which the Ottomans used it was a bad fit.

What about Vietnamese?

Well, Vietnamese is a tonal language, closely related to Chinese. Before the Latin script was imposed by France as a colonial power, the existing script was "chu nom", derived from Chinese characters. (It involves newly drawn characters to match Vietnamese words, unlike the direct use of Chinese characters as Kanji in Japanese. I lack the space and the expertise to further discuss this fascinating writing system.)

While the Chinese writing system has its defenders, the number of symbols does mean it takes longer to learn, and thus it is not well suited to mass literacy in a poor country. A wealthy one, like Japan, can manage.

So if chu nom is the alternative, the existing Vietnamese script, even if a bad one, could scarcely be a worse one.

Vietnamese has an extra layer of accent marks above the letters because it is a tonal language. Perhaps it could keep the tone marks, but get rid of the other diacritical marks if it had an alphabet with more letters in it, such as that of Georgian or Armenian. Or one could use Cyrillic and Greek to augment the Latin alphabet - or go with something like the Pitman ITA. That way, with only tone marks, Vietnamese writing would be comparable to pinyin.

But I see nothing to indicate that Vietnamese would be better off with an abugida or a syllabary, and I'm not sure there are any other choices. So I think that the possibilities for improvement in the Vietnamese writing system are marginal at best.

Christian Thalmann

RobPratley said:

How does one qualify their judgement of a languages' use of a script without substantial linguistic and culturally relative knowledge? If the phonology is actually very well represented by that script with those 'wiggles' does it make it bad because one is used to that script without those 'wiggles'?

I don't doubt the Vietnamese alphabet does a good job at mapping the phonology of Vietnamese; that wasn't my point. My point is that cramming lots of tiny wiggles into a small ascender space is a bad design choice. I suspect they would have done better with modifier letters, digraphs etc. Just a hunch though, I haven't thought it through.

John Savard

Christian Thalmann said:

My point is that cramming lots of tiny wiggles into a small ascender space is a bad design choice. I suspect they would have done better with modifier letters, digraphs etc. Just a hunch though, I haven't thought it through.

Actually, that is a better idea than the ones I considered, and it may well be workable. Had the English instead of the French colonized Vietnam, that could have been how it turned out.

Of course, there is one thing that Vietnamese should not do with digraphs or modifier letters: use them to indicate tone. One of the methods used for transliterating Chinese, the Yale system, does that. Unlike Wade-Giles or Pinyin, the Yale system has a reputation for being very confusing and hard to learn for native speakers of Chinese.

Andrew Wood

"One of the methods used for transliterating Chinese, the Yale system, does that. Unlike Wade-Giles or Pinyin, the Yale system has a reputation for being very confusing and hard to learn for native speakers of Chinese."

To be fair to the Yale system: it was designed to teach American soldiers how to speak Chinese quickly during the second world war, and not for native speakers of Chinese.

Also, what Chinese are you talking about :-) Putonghua (mandarin), Cantonese, Hokkien, Hakka, etc. There are many languages that Chinese people speak.

John Savard

Andrew Wood said:

Also, what Chinese are you talking about :-) Putonghua (mandarin), Cantonese, Hokkien, Hakka, etc. There are many languages that Chinese people speak.

Wade-Giles, Pinyin, and Yale are all systems for writing Mandarin, and that only. There are transliteration systems, sometimes related to ones for Mandarin, for the other Sinitic languages, also known as dialects of Chinese, such as Cantonese, Taiwanese and other Southern Min dialects, Shanghai dialect and other Wu dialects, and so on and so forth.

Adam Jagosz

Christian Thalmann said:

sort of like Vietnamese (IMHO a very poor use of the Latin script)

Oh, I was meaning to comment on that utterance, but I couldn't find what thread it was in! So I'll go off-topic just like you.

They could very well base the vowels on IPA and use:

• Əə instead of â
• Ɛɛ instead of e and Ee instead of ê
• Ɔɔ instead of o and Oo instead of ô
• something not based on phonology and IPA (because it has the sound /aː˧˥/ according to Wiki) instead of ă. Maybe Ʌʌ?

Just 4 changes and we have only single-story accents.

I didn't touch on the horns, but I don't think they aren't a lucky idea either. They are cute and maybe they don't clash into the following letters, but are generally difficult (though satisfying) to design and do take up some ascender space. Maybe Ɵɵ and Ʉʉ would be better?

AzizMostafa AzizAli

https://www.ircica.org/event/11th-international-calligraphy-competition-registration-will-be-closing-on-31-december-2018