Requests for new languages/Wikipedia Tunisian/The Discussion with Language Committee and Incubator Administrators about some facts of Tunisian

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

Proposal about Wiktionary Tunisian[edit]

We have managed build a Wiktionary in Tunisian. However, Tunisian is written in LTR Format and the system is made in RTL Format. Can you change this for us? How can we apply?--Csisc (talk) 08:32, 10 April 2015 (UTC)

Csisc: Take a look at Help:Manual on how to start a new project. —Justin (koavf)TCM 14:14, 10 April 2015 (UTC)
Koavf: Tunisian is written in LTR Format. I ask if you can change the format from RTL to LTR for us. --Csisc (talk) 11:32, 16 April 2015 (UTC)
Csisc Tunisian Arabic (aeb) is written in Arabic script which is right to left not left to right. Mjbmr Talk 12:08, 16 April 2015 (UTC)
Mjbmr: Tunisian is written in Latin Script. There are many works using Latin Script in writing Tunisian. For Tunisian, the used scripts are Latin and Arabic. You can see Tunisian Wikipedia discussion page and Tunisian Arabic in Wikipedia to verify this fact. The Lablebi and Karmous Dictionaries for Tunisian provide the words in both Scripts... You can also open the Wt/aeb and Wp/aeb Wikis to verify this.--Csisc (talk) 12:26, 16 April 2015 (UTC)
Csisc Please contact iso639-3@sil.org and tell them this. Mjbmr Talk 12:34, 16 April 2015 (UTC)
I think that Maik Gibson from SIL who spent years studying Tunisian knows about this. However, I will contact them.--Csisc (talk) 12:39, 16 April 2015 (UTC)
Csisc I've submited a patch for you: gerrit:204503, but I don't think people would like that, since this language is called Arabic. Mjbmr Talk 13:41, 16 April 2015 (UTC)
Mjbmr: Thanks very much. Who won't like that? Which people? Tunisian Wikipedia and Wiktionary editors support Latin script! The biggest and most professional dictionaries for Tunisian use Latin script! The word "Arabic" is just a part of name, which can be removed one day, once Tunisian becomes the official language of Tunisia for example! Maltese Arabic is Arabic, but it uses Latin script... And we are dealing with the subject from a Political point of view. From another point of view, Latin is better, because Arabic script can't denote the 8 Tunisian vowel + 2 borrowed vowels from French + Nasal vowels. Impossible! It's possible though to forget about vowels and denote emphatic consonants instead, but Tunisian has 9 emphatic consonants! Only Latin script can't show those emphatic consonants because it's more widely used so you have a bigger choice of letters. Nobody can support the Arabic script, unless he has political views on the subject (Pan-Arabism, which is no more accepted in Tunisia). Thanks a lot. --GeekEmad (talk) 16:12, 17 April 2015 (UTC)
Hi,
As I wrote in the patch, there's a problem. It's not that I just "don't like" it, and the problem is not that the "language is called Arabic".
I don't disagree with the general idea of having Wikipedias in modern variants of Arabic, such as Egyptian or Moroccan. The issue is there is supposed to be some kind of standard and agreement about what language this is and how is it written. Currently there's a mixup:
  • The Wikipedia and Wiktionary incubators are essentially in Latin, but use different variants: the Wikipedia says "Marħba", and the Wiktionary says "Marḩbê", and uses much more diacritics.
  • Ethnologue says "Arabic".
  • Translated messages in translatewiki are both Arabic and Latin (!!!)
Having a project in two alphabets is wrong. We have that with Azerbaijani, and it doesn't really work well and the language committee won't agree to creating another one.
I don't want to force neither the Arabic script nor the Latin, and I don't want to prefer any particular variant. Most importantly, the community that starts the project must agree as early as possible on which standard to use, and that standard is supposed to be based in real life and not a new invention. That's why I repeatedly ask: Are there other publications in this language? Newspapers? Magazines? Textbooks? Websites (professional, edited websites; not social networks)? A Wikipedia is supposed to be written like they are. A Wikipeia is supposed to reflect reality and not create it.
That said, it's even conceivable that a Wikipedia would be the first major publication in a new written standard, but I'd still love it to be described somewhere, for example a grammar book.
Sorry for making this hard; we are just trying to be professional :) --Amir E. Aharoni (talk) 08:55, 17 April 2015 (UTC)
Amir E. Aharoni: First, I thank you for replying me. First, let me tell you that there are many booklets written with both scripts. However, linguists use mainly the Latin Script as it is the most intelligible for them. Marçais, Cohen, Stumme, Nicolas, Saada and many other linguists used this script. Arabic Script is used as well in other situations... So, we had to make a choice... We chose that the Script that would be used is Latin One because Arabic one would not be efficient in transcribing all words...
However, we faced something strange. The Latin Script used in each of the works is not the same (Arabizi, Cohen, Peace Corps, Marçais...) and the same thing happens for Arabic Script as well. So, we had to standardize Tunisian based on the existing literature... For this, we contacted a senior researcher from the University of Pennsylvania who is Mr. Mohamed Maâmouri who worked about the standardization of Tunisian for years... We got finally a common version of Latin Script for Tunisian this month... We asked some linguists for reviews and we adjusted the output...
The right spelling is the one existing in Wiktionary and we will update the Wikipedia and the TranslateWiki depending on that and eliminate all Arabic Scripts.
As for the standardization booklet, we are writing it now with the collaboration of Nizar Habash and Mohamed Maâmouri and when we complete it by mid-2015, we will be happy to introduce the final output to you so that you can review it...
If you are interested to see works in Tunisian with Latin Script: You can consult Karmous Dictionary when words are written with both scripts, the Facebook groups La langue tunisienne, I speak Tounsi..., Peace Corps Tunisian Books and Social Networks...
Yours Sincerely,
--Csisc (talk) 10:56, 17 April 2015 (UTC)
Amir E. Aharoni: Hello, all of Tunisian Wikipedia editors agree with using Latin script, and none of them support the Arabic script, due to the defectiveness of Arabic script, which is not suitable for Tunisian phonology. Now, the Latin script, at first, we adapted a method called STUNdard which use a lot of digraphs and ignore emphatic consonants. This method was regulated by League of Tunisian Humanists (LTH). Members of LTH, and volunteers suggested many reforms for STUNdard but none of them was supported by a linguist. Later, Csisc succeed in contacting some Tunisian linguists who gave so many suggestions for a better Tunisian writing method (So the actual work is a linguistic work). This beta method (beta, because the pronunciation is not standardized yet) is used on Wiktionary, and after finishing the first standard Tunisian language dictionary, we will update Wikipedia and Translatewiki.
Saying Tunisian Arabic doesn't mean that it's a dialect of Arabic (They are mutually unintelligible) and doesn't mean that we are forced to use Arabic script. Maltese Arabic for example uses Latin script. The used script is related to the phonology of the language.
Arabic script is not used to translate Wikimedia messages on TranslateWiki... The text in Arabic script is actually Arabic language. Early translators thought that copying from Arabic language would make it faster to accept Tunisian Wikipedia. Those messages will be reviewed later. The biggest problem for translating Wikimedia messages is that Arabic language is used as a secondary language (to help the translator understanding the meaning of the message). We actually need French, Moroccan and Algerian to more understand the meaning of the message, and maybe to keep Tunisian, Algerian and Moroccan Wikipedias related to each other.
The reality is that Tunisian is not used in education, so it's rare to find written literature in this language. Most of the written literature forces Arabic letters for Tunisian language, but this method is so defective, because it doesn't show the 9 emphatic letters of Tunisian, nor the difference between [ɑ] and [ɛ]. The two most complete dictionaries of Tunisian language use Latin script, Peace Corps Dictionary (old) and VICAV dictionary (new). The first prefer the use of [ɑ] and [ɛ], and the second prefer the use of emphatic letters. This is an overview of writing systems used for Tunisian language.
Conclusion: We are using Latin script for all Wikimedia projects. Also, Amara approved translating TED Talks' videos (and all other videos on this platform) to Tunisian language using Latin script.
--GeekEmad (talk) 12:15, 17 April 2015 (UTC)
Amir E. Aharoni: Nowadays, the Latin Script is used for typing Tunisian in Social Networks and linguistic works. So, there is no need to write Tunisian with Arabic Script. All works with the Arabic Script like the books and romans can be translated to Latin Script as there is a mutual interconvertibility between Arabic and Latin Script--Csisc (talk) 12:22, 17 April 2015 (UTC)
Amire80 These descriptions was not enough for you? The community says and willing to contribute in Latin script. Csisc, GeekEmad Please also vote on w:Talk:Tunisian Arabic#Requested move 18 April 2015 Mjbmr Talk 15:32, 18 April 2015 (UTC)
Mjbmr: Thanks for your efforts, but it seems that it's impossible to change the title of that page. Let's just keep it, changing it is not a big issue. --GeekEmad (talk) 22:16, 18 April 2015 (UTC)
Mjbmr: I have sent all references to SIL. I was a bit surprised by the discussion about Tunisian Arabic. The Arabic Script does not provide excellent coverage for vowels and that is why we adopted Latin Script.--Csisc (talk) 12:52, 19 April 2015 (UTC)
Mjbmr: As for the situation of Tunisian, Tunisian is a divergent dialect... This means that Tunisian is not intelligible to Arabic Speakers who are not familiarized with Tunisian... However, it contains some structures that are extracted from Arabic and having currently the same format as in Arabic. That is why it is not a language... This subject caused controversies... So, we should probably keep it as Tunisian Arabic. However, recognizing Latin Script is required as Tunisian is transcribed in Latin Script since 1893 with Hans Stumme... --Csisc (talk) 12:58, 19 April 2015 (UTC)
Csisc OK, start translating mediawiki messages in translatewiki with Latin script and rewrite all previous messages with Latin script, then we will ask Amire80 to merge that patch. We have like Tajik Wikipedia written in Cyrillic script and a convert code for Latin script. There is no such a resource for each translations we make in translatewiki, the actual resource for translations are users, so if we have all contributions in Latin, then Latin will be the main script for Tunisian projects. Mjbmr Talk 16:57, 19 April 2015 (UTC)
The difference from Tajik is that the Cyrillic script has been widely used for many decades, and the Latin script was used for some time before that in some easy to find publications, so both are documented.
Tunisian in the Latin script, however, is hardly documented. It's not quite enough that the community writing the Wikipedia agrees upon it - what's really needed is to know that the wide community of people who speak this language will use this script, will be able to read in this script, will use this script to search the web in this script, and so on. As I said, it's not a total blocker, but it will be much easier if such information would be known.
Also, what do you mean when you say "The right spelling is the one existing in Wiktionary"? According to which standard is it right? --Amir E. Aharoni (talk) 18:08, 22 April 2015 (UTC)
We had analyzed all the Latin Scripts used for Tunisian and created a standard one. All the references are explained in [1]. --Csisc (talk) 12:21, 24 April 2015 (UTC)
Csisc Can you respond to this? Mjbmr Talk 17:36, 24 April 2015 (UTC)
Mjbmr: There are six main Latin Scripts used for Tunisian:
  • DMG Transcription (Based on interconvertibility to Arabic Script and the transliteration of voices): Created in 1969, this transcription had been used for Tunisian since 1988 with the works of Maik Gibson who is now a senior consultant of SIL International. The work has been used by Ines Dallaji in 2010 to transcribe six dialogues about Wedding in Nabil Region.
  • Peace Corps Tunisia Transcription (Based on the transliteration of voices): Created in 1970 by the team of Inglefield et al., this transcription is used only for Tunisian and was a result of the researches done with the collaboration of Tunisian Assistants conducted in the University of Virginia. This transcription had been used until 1990 in transcribing Tunisian in Peace Corps Tunisia Works
  • French Script (Based on the transliteration of voices): Created in 1910, this transcription used French Pronounced Latin Letters and Digraphs to transcribe Tunisian. Introduced by Alfred Nicolas, it was used later by William Marçais, David Cohen, Lucienne Saada and others... It worked for many linguistic works until 1989
  • Tunisian Arabizi (Based on the transliteration of voices): Appeared in 1995, this transcription is based on changing the non-Latin Voices into Arabic Numbers... This is used by users till now... However, it is widely seen as informal.
  • Buckwalter Transliteration (Based on interconvertibility to Arabic Script): Appeared in 1991, this transcription developed by Tim Buckwalter from the University of Pennsylvania used ASCII characters to transcribe Arabic Letters. It is used since 2010 for Tunisian Arabic. However, many problems had occured
  • Standard Transliteration (Based on interconvertibility to Arabic Script and the transliteration of voices): Till now under review and also known as STUNdard Transliteration, the work made a synthesis of the five experiences and made a new method that benefit from all the characteristics of the other transliteration methods. We consulted the leaders of each transcription method and made a new one.
For references, please consult this website and not the one you had made. I will publish the final output of this important standard transliteration booklet when I finish it.--Csisc (talk) 08:16, 25 April 2015 (UTC)
Csisc I DID! he was the one who found mine! Mjbmr Talk 10:20, 25 April 2015 (UTC)
Mjbmr: I finished the booklet that standardizes the transliteration of Tunisian. I will send it for review and next week, I will publish it. --Csisc (talk) 12:03, 25 April 2015 (UTC)
Mjbmr: You can consult the booklet here. --Csisc (talk) 11:24, 26 April 2015 (UTC)
Mjbmr: As shown in this important booklet, Tunisian is written as Tûnsî and not Tûnsi. So, please review this soon.--Csisc (talk) 10:19, 27 April 2015 (UTC)
Csisc As I told, I've set a patch but someone else must approve it. I was found that from your edit. BTW according to CLDR, there is already a Arabic to Latin transliteration rules which will be used to convert Arabic articles into Latin, there is no need to have two variants. Mjbmr Talk 15:02, 27 April 2015 (UTC)
Csisc Update from langcom, patch to set aeb fallback to aeb-latn. Mjbmr Talk 15:57, 27 April 2015 (UTC)
Csisc Based on Amir E. Aharoni's comment and Gerard's respond I have no more word to say. Mjbmr Talk 16:27, 27 April 2015 (UTC)
Mjbmr: According to the paper, the interconvertibility of the two scripts is not complete particularly for the vowels. The interconvertibility of Scripts is not a drawback... It is an advantage... and this is why we have standardized Tunisian... You have adopted many scripts that are totally convertible for languages like Kazakh... However, we need only using Latin Script and you can not to use Arabic Script. Furthermore, we are not against Arabic Script... We have reported it. However, we recognize that Tunisian is a variant of Arabic that is not intelligible to Arabic. It involves many vowels that may be confusing in Arabic Script... So, adopting Latin Script was in order to transliterate Tunisian as it should be and not because of a social reason. Furthermore, the Google Site entitled Stundard Tunsi is very old and it was written before three of the four authors joined STUNdard. We are not interested in updating this website now because we are standardizing Tunisian. However, we will do that. —The preceding unsigned comment was added by 197.0.249.61 (talk) 14:16, 28 April 2015 (UTC)
As for the comments, there are many people who spoke about STUNdard and that is why we could get help from linguists... We even discussed STUNdard in WikiArabia last month... Try to change the name of Tunisian from Tûnsi to Tûnsî--Csisc (talk) 08:25, 28 April 2015 (UTC)
Mjbmr: With AEB-LATN, we had translated 48 pc. of MediaWiki most used messages, 49 pc. of TranslateWiki and 19 pc. of the TranslateWiki main page. As for AEB, it contains many deficiencies. I propose that you will adopt Latin Script and transfer AEB-LATN as AEB and drop all messages in AEB.--Csisc (talk) 09:29, 28 April 2015 (UTC)
Amir E. Aharoni, Gerard Meijssen, Oliver Stegen, SPQRobin: I have seen that I had been misunderstood. In fact, it is more than true that STUNdard had the attitude cited by Meijssen. However, at that moment, I was not affiliated in LTH Team… So, all what is included in the references found in Internet does not seem to be relevant according to all the new members including me and User:GeekEmad.
The first question raised is that STUNdard is against the adoption of Arabic Thoughts and Civilization. That was true for the versions 1.0, 2.0 and even 3.0 of STUNdard and when it was introduced to me for the first time, it clearly noticed this. However, what is clear to us now is that Tunisian is a variant of Arabic because removing all unmodified aspects of Standard Arabic from Tunisian will let it dysfunctional. So, Tunisian is dependent of Arabic and that is why it is a dialect and not a language. But, Tunisian is not intelligible to other native Arabic Speakers because it involves structures and sounds derived from other languages like Berber, Maltese, German, Italian, French, Punic… So, Tunisian is a divergent dialect of Arabic.
The second question is the doubts about creating Wikimedia Projects in Tunisian using Latin Script. To address the issue, I will give some examples… In Tunisian, we have borrowed from Slavic Languages the principle of the stress given to nouns because we have been occupied by Vandals prior to the Arabic Era. So, unlike other dialects, we have dâr to mean house (a noun) and dêr to mean turn (a verb) for example. The situation worsens when the two words are from the same grammatical form like for the situation of Kêri and Kâri that are both nouns. This is not existent in other dialects that pronounce both of them as dâr and that is why using Arabic Script within them does not cause problems. However, this problem commonly exists in Tunisian. So, the transliteration of vowels using Arabic Script would be defective… Droping this aspect of Tunisian would mean neglecting an important aspect in Tunisian. In Tunisian, we have more vowels than in Arabic and even different structures than in other Arabic Dialects and that is why all linguistic works about Tunisian used Latin Script since 1893 and not the Arabic Script unlike in the situation of other Arabic dialects. Many had proposed to me to drop vowels from the Script and use Arabic one. The problem is that Tunisian People do not use Arabic Script as it should be in writing Tunisian. The audience of the Wikimedia Projects is mainly constituted of young people who use Arabizi to transliterate Tunisian. Some of them are living abroad. So, they are unfamiliar to Arabic Script when used to transliterate Tunisian. When standardizing a language, the main authors have to choose the script that allows them to guess the most of the aspects of the dialect… That is why we created this script… and that is why we have decided to use this script to write Wikimedia Projects in Tunisian…
The third question is the doubts about the reputation of the new form of STUNdard. For this particular fact, I cannot predict anything… However, I think that this script has been made by taking into consideration the common experiences of the transliteration of Tunisian in both Scripts. The new script is intuitive as the adopted characters are derived from the configuration of Arabic Script… If you allow us to do a Wiktionary project using STUNdard, we can try this and see. Moreover, the use of Latin Script is excellent in TranslateWiki for example. The translators of Algerian Wikimedia Projects had spent a lot of time in translating the messages into Algerian using Arabic Script although this script is not defective for Algerian. As for our situation, it is quite impossible. However, when we used Latin Script, we have exceptional results in TranslateWiki. So, the use of this script would be efficient and its success would depend only on how we will introduce the new script to audience. That is why we will try after finishing the standardization booklet to see how we will advertise it in Social Networks and Wikimedia Projects... However, I think that the new script is very efficient and can influence Tunisian people easily…
The fourth question is why we did not post the new method and spread it as it should be. We have done that because we liked to publish all the practical standardization of Tunisian in one booklet and not in a series of booklets. The booklet is working and is verified regularly by linguists and you have seen the part about syntax we had finished recently. We will introduce the full booklet when we will begin the discussion. In that moment, we have to finish translating TranslateWiki, get recognition from SIL and begin working in Incubator and this is not possible if you could not admit the Latin Script for us now. So, please accept this script and adjust the name of Tunisian as “Tûnsî”, recognize it as a linguistic entity in TranslateWiki and adopt it and we will do our best to publish the book as soon as possible and finish translating the TranslateWiki Main Page, TranslateWiki and core MediaWiki important messages. If you have further details to provide me, please put them here because I cannot verify Phabricator, Gerrit… I will answer you as soon as possible and provide you with more explanations about the aspects of Tunisian and what we are doing now.
Yours Sincerely,
--Csisc (talk) 07:39, 29 April 2015 (UTC)
Amir E. Aharoni, Gerard Meijssen, Oliver Stegen, SPQRobin, Mjbmr, Koavf: I think that there are some doubts about the Latin Script… I understand that the fact of having many Latin Scripts in Tunisian and that its standardization is new will cause some debates in SIL and Language Committee. However, knowing that Arabic Script is the main script for Tunisian and that Latin Script is the most common one, we can use Latin Script in making the Wikipedia, the Wiktionary, the WikiVoyage, the WikiNews (as contributors use daily the Latin Script and as most of the dictionaries about Tunisian are written in Latin Script) and we can use Arabic Script for Wikisource and Wikibooks (as main literary books are written in Arabic Scripts) and this will give an excellent result. I will talk again with SIL in order to let them recognize the different types of scripts for Tunisian… Arabic Script (main and literary), Latin Script (common and daily)… and get the two ISO codes AEB-ARAB (keep AEB as AEB-ARAB if you like this) and AEB-LATN (as specified clearly by BCP 47). Then, I think that you should correct the spelling of Tûnsi as Tûnsî… After, I will work on messages in both scripts and on test wikis. Furthermore, I will work on commenting Ethnologue Overview about Tunisian. --Csisc (talk) 08:48, 3 May 2015 (UTC)
@Csisc: Have you seen the Serbian Wikipedia? Serbian is written with the Cyrillic and Latin scripts and there's a tab that allows you to change between the two. Could this be worked into Wiktionary somehow? Do we need a separate project? —Justin (koavf)TCM 14:09, 3 May 2015 (UTC)
@Koavf:: This can partially work (Latin to Arabic Script) but not the reverse. However, Amire80 did not support this fact. He had said that a wiki should be written with one formal script (as the experiences of the wikis using two scripts had been not very successful) and that is why I had chosen to do a Wiktionary, a Wikipedia, a Wikinews and a Wikivoyage in Latin Script and a Wikisource and a Wikibooks in Arabic Script. However, we can use this important idea of the tab. We can allow editing in Latin Script and viewing in both Arabic (alternative) and Latin (main) Scripts. But, I think that some adjustments should be done as we use Arabic Numbers like 1, 2 and 3 and not indian ones for Arabic and Latin Scripts and as Latin Script is LTR and not RTL. Furthermore, the transcription is vowel-free (if we do not consider any Haraka) excepting when the vowel comes in the first or the last position and ' (Used also for font) is considered as a letter in Tunisian and its transcription can vary according to the vowel preceding it. Moreover, the hyphen is considered sometimes as an additional letter when not preceded by a space and we can face problems when citing a quote from another language... --Csisc (talk) 08:05, 4 May 2015 (UTC)
It depends on how the tab works. The transcription of the letter in Tunisian depends on the character coming before it and the one coming after it. A full table of conversion could be found here. If the tab could evaluate the characters near the letter and convert it by taking into consideration the characters near it, this could be done. However, if the tab just converts letters, it will be difficult and this is what I see... So, we have to adopt one of the scripts and work on it. I will begin today a Wikisource in AEB-ARAB. So, try to recognize AEB-TRY for Wikipedia and other projects--Csisc (talk) 16:28, 4 May 2015 (UTC)
@Koavf:: As I have seen, the system supports phonemic conversion of scripts. However, this is not the situation for Tunisian when the conversion is phonemic and syntactic at the same time. --Csisc (talk) 09:39, 6 May 2015 (UTC)
@Koavf, Amire80, Mjbmr, and GerardM: As I had seen, you had added the AEB-ARAB and AEB-LATN extensions. However, the name is still not accurate for AEB-LATN as Tûnsi is written as Tûnsî as explained in the writing booklet. Try to adjust it as soon as possible and give me access in order to edit in Wp/AEB-LATN and Wt/AEB-LATN. As for SIL International, they are discussing recognizing the script and would inform us of their decision as soon as possible. --Csisc (talk) 12:13, 6 May 2015 (UTC)
Csisc Fix for the name was submitted by me in gerrit:206634 and merged by someone else, no one has the power to change anything they like, those variants were add me and was merged by someone else, you don't listen to what I told you in phab:T97934, you have mentioned other people also here who has nothing do to with this, I was just trying to help your community out but sadly I was mistaking, if you follow my instruction in phab:T97934, you can get though this, and don't continue this discussion, it's not worth it. Mjbmr Talk 12:37, 6 May 2015 (UTC)
Mjbmr: I thank you for replying me and your exceptional help to me. I took your advice into consideration as I currenly use div to let the script appear in LTR Format. I thank you very much for this. However, I am only asking to adjust the name of the language as "Tûnsî". But, as you have said, I should probably stop it now. You have advised me to work in Wp/aeb/latn and Wp/aeb/arab. However, I have seen today that AEB-ARAB and AEB-LATN had been recognized. So, that is why I have updated the comment in Phabricator. --Csisc (talk) 12:47, 6 May 2015 (UTC)
I see that the fix had been accepted now. I thank you for this and I am really very sorry for the disturbance I caused those days. I will translate the most used MetaWiki messages now. I will wait for SIL answer. --Csisc (talk) 12:52, 6 May 2015 (UTC)