Wikipedia talk:Manual of Style/Arabic/Archive 3

The many problems of Urdu
I think it's a mistake to extend this policy to cover Persian, and certainly to Urdu. There are several problems Urdu introduces. First of all و can represent 4 different phonemes (v, ū, o, au) as can ی (y, ī, e, ai), though this isn't a greater problem than Arabic vowels. The larger problems are with consonants: Urdu can have x different combinations whose base Latin consonant would be T, t, tʰ, ʈ, and ʈʰ from Hindustani and tˤ and θ from Arabic - though the latter are pronounced as allophones to other Urdu phonemes. These combinations arise because Urdu orthography takes into account retroflex and aspirated distinctions from Indic languages and Emphatic ones from Arabic - and aspiration is a particularly important phoneme, usually indicated by an h. However, this would result in confusion between th تھ and th ث in the current proposal (I've seen formal Urdu transliterations that use ṯ for the latter). Conflicts would arise for any consonant that is written with an h-digraph in the current proposal.

Accommodating these complications would clutter up run-of-the-mill Arabic transliterations greatly, and so I think extending this policy to cover Urdu, or even Persian, is a mistake. Moszczynski 17:18, 7 June 2006 (UTC)
 * I agree, as I commented above. There should be separate style manuals for Persian and Urdu. --Cam 19:38, 7 June 2006 (UTC)


 * I disagree. I don't know much about Urdu, but all the complications you mentioned are not unique to Urdu. In Arabic the و can represent ū, u, or au, the ی can represent y, ī, i, ai... and this page already deals with the case of تھ vs ث, see the tables. There are a few consonant differences, like the  و is pronounced as a "V" instead of a "W" when used as a consonant, but those differences are not extensive and they are already noted. Not to mention that a large percentage of words in both Persian and Urdu come directly from Arabic. Cuñado -  Talk  01:52, 8 June 2006 (UTC)


 * I mentioned that I agreed that the vowel problem wasn't greater than for Arabic. It's the consonants. To accommodate Urdu in a single standard we would have to abandon th for ث, so we'd have to use something like ṯ, which is decidedly non-optimal. And what about ڈ? Would we use another diacritic like...ƭ? Our transliteration system would wind up needing a transliteration system of its own. It would be best if Urdu used its own system, with Arabic letters (which our rarer and allophones) as the ones with diacritics Moszczynski 23:36, 8 June 2006 (UTC)

ibn/bin
Above someone mentioned:
 * "pronounced ibn in the Middle East and bin in North Africa."

I'm not incredibly familiar with this, but I've noticed that most wikipedia articles use ibn, but the MOS currently says to use bin. If it's true that it's pronounced ibn in the Middle East then we should change the page. Cuñado - Talk  02:20, 22 June 2006 (UTC)


 * I'm not sure about the dialects, but as I understand it, the rule is that when bin is part of a name (as in Osama bin Laden), bin is the pronunciation, but when it starts a name in its own right, (as in Ibn Battuta) the i is pronounced. This is because the Arabic word for ibn starts with hamza al-wasl, which is assimilated when preceded by another word...although the various western usages (the bin Laden family specifies that bin should always be used, for example) have muddied this up a fair bit. I think it'd be tricky to put down a hard and fast rule, though from what I've noticed modern usages favour bin, as ibn is hard to pronounce for most Westerners Moszczynski 03:05, 22 June 2006 (UTC)

Now, what about bint vs binte for female names? Currently both are used equally frequently. --tyomitch 05:16, 14 August 2006 (UTC)

hamza and ayin
I have been trying different things on many Arabic articles, and the best appearance of the hamza and ayin appear to be the half turned commas, or ...  ‘ ’ 

In small font you can't tell the difference, and the current proposal to use the acute accent  `  looks kind of funny. Would anyone object if I re-worked the standards so the half turned commas are used in the standard transliteration? Cuñado - Talk  07:10, 8 August 2006 (UTC)

what to do with the hamzat-ul-wasl?
How are we supposed to transliterate the assimilated hamza, as in this sections' title? In the article on Alif, it's called hamzatu 'l-wasl, but this doesn't scale well. Should e.g. ‘Abdu’llah (name) be instead moved to ‘Abdu ’llāh (name), and Abd-al-Rahman (name) to Abdu 'l-Rahman (name)? Not to mention that the last case violates this MOS by not assimilating Lam: according to it, the title should be Abd ar-Rahman (name). --tyomitch 05:13, 14 August 2006 (UTC)


 * According to the current page, there are four choices: `Abd ar-Rahman (name), `Abdu'r-Rahman (name), `Abd al-Rahman (name), or `Abdu'l-Rahman (name). Also, which characters to use for the hamza and ayin are sort of being discussed, so theoretically more. This gives you an idea of the problems that people have when transliterating Arabic. I can assure you, however, that hamzatu 'l-wasl is not correct. There should not be a space. And also there are no underdots, underscores, or lines over letters in page titles, so ‘Abdu’llāh (name) is incorrect (because of the ā). Also, Abd-al-Rahman (name) is incorrect because there is an extra dash (not to mention its missing an ayin). This is in conformance with the various forms of Arabic transliteration, and is not made by any wikipedians. Cuñado - Talk  07:04, 14 August 2006 (UTC)


 * Could we please never ever use backticks in transliterations. This symbol doesn't exist typographically speaking. Eiter use a typewriter apostrophe or typographical quotes. —Ruud 10:58, 14 August 2006 (UTC)
 * Well it looks ugly. What do you mean it doesn't exist? Cuñado - Talk  16:48, 14 August 2006 (UTC)
 * Also, here is a table I made of the possible markers...

I can't see the half-rings. On a related note, why at all should Ayin be represented with a diacritic mark? It's a full-fledged letter, unlike Hamza. How about using the  symbol from IPA? --tyomitch 04:57, 15 August 2006 (UTC)
 * There needs to be a middle ground between accuracy and readability. Striver is fighting against the very basics of standardized Arabic. Some people have tried to implement the most strict form of transliteration in titles and article text. The accents representing the hamza and ayin can be read by a person unfamiliar with Arabic, and they don't get in the way, while at the same time identifying important information to those familiar with what they represent. I wouldn't say the same for ʕ.
 * What browser are you using? Do the half rings appear as squares? Cuñado - Talk  01:03, 16 August 2006 (UTC)
 * IE6 on WinXP (I expect this to be the most common configuration among all the Wikipedia visitors). Yes they do appear as squares.
 * And to a person unfamiliar with Arabic, ` conveys as little information about Ayin as does . Both are just obscure confusing marks to him. --tyomitch 07:33, 16 August 2006 (UTC)

The problem of Ottoman Turkish
The recent spate of activity over at Abdülmecid/‘Abdu’l-Mijid has prompted me to make a first step towards at least introducing the (hitherto undealt-with on this page) problem of Ottoman Turkish. The main problem is that, despite the statement that the guidelines apply to "any language that uses this script, such as Arabic, Persian, or Ottoman Turkish", the last language of the three mentioned presents something of a special case insofar as it has a (reasonably, with very small exceptions) standard transliteration system already in place in the modern Turkish language's romanization of Ottoman names, and those names are—without any doubt—the most common ones searched for.

So, for example, to put "Abdülmecid" (عبد المجيد) under "‘Abdu’l-Mijīd" so as to make it conform to a standard applicable to Arabic transliteration (and pronunciation) rather than the Turkish-language based system in place since at least the 1930s, is quite simply madness (even if a redirect does end up taking you to the right place).

Any thoughts on the issue? —Saposcat 09:48, 22 August 2006 (UTC)


 * (Without doing too much research) Abdülmecid seems to be the most common name in use in the English speaking world, which is therefore prefered over any transliteration of the name in the original script. —Ruud 13:50, 22 August 2006 (UTC)


 * My first thought is to agree with you. If there is a standardized version of Turkish then I would support using it and separating Ottoman Turkish. I'm really not familiar with Turkish language other than historical facts. But I hope you realize there is a good argument to keep Arabic transliteration, considering they were using Arabic script, and most of the time using Arabic names. When Ataturk tried to purge Turkish of Arabic roots it became impossible because they were so fundamentally integrated. I couldn't find any MOS pages for Turkish. I suggest you make a new page under the title Manual of Style (Turkish) and make clear guidelines on how to deal with Turkish names, places, etc, with definitions as to what cases call for Turkish vs Arabic. Then try to get peer review and approval from editors familiar with Turkish.


 * What I really oppose is the use of non-standardized transliterations. Sometimes people just make up their own way of spelling, and insist on their way in article titles. The problem gets harder with Persians, who insist that their language is not Arabic, and ignore the common script, common words, and the need for standardization. Cuñado - Talk  02:09, 23 August 2006 (UTC)


 * Yes, Turkish is standardized, including names of historical figures. There is no more of a transliteration question for Turkish than there is for German or French - the only question is whether the article names as used in URLs flatten letters with umlaut or cedilla to the unaccented letters. Arabic script should be provided as auxiliary when the subject of the article was pre-script reform and when we actually have a source giving the Arabic-script name as used in Ottoman Turkish, not just a source for the cognate Arabic word.


 * It is true that Ataturk did not eliminate all words of Arabic origin, but that does not mean Arabic script is used at all for the Turkish language today. "Abdülmecid" gets 103,000 Google hits, "‘Abdu’l-Mijid" gets 99 condensed to 4 distinct hits, all of which are Wikipedia.


 * I would change "any language that uses this script, such as Arabic, Persian, or Ottoman Turkish" to "Arabic, or any language using Arabic script that does not have an established standard transliteration system".


 * The Turkish page on the Ottoman alphabet is tr:Osmanlıca harfler though it looks like English Wikipedia currently has more detail. JWB 03:13, 23 August 2006 (UTC)


 * You're pointing out a good problem with Arabic. Abdul Mejid gets 26,000, Abdul Mijid gets 120, Abdul Majid gets 700,000. And those don't even distinguish between what kind of markings are used for the ayin and hamza. You can get more versions by using non phonetic spelling (Abd al-Mijid instead of Abdul Mijid)
 * Like I already said, there are good arguments for using both languages' standardized forms, so if you're familiar with the subject, start working on the Turkish MOS. Until then I'll be using the current guidelines. Cuñado - Talk  03:51, 23 August 2006 (UTC)


 * You say that "there are good arguments for using both languages' standardized forms", but in this case, I really believe that there are not. Turkish is a language that moved from the Arabic abjad to the Latin alphabet, and has a standardized and regular spelling system in that alphabet, as well as a standardized way of transcribing words that were previously written only in the Arabic abjad. When you point out that "most of the time [they were] using Arabic names" and that thus we should transliterate them according to Arabic transliteration standards, it is at bottom no different from arguing that the Franz Ferdinand song "Michael" should in fact be under the name "Miḫaʾel".


 * I respect your opposition to "the use of non-standardized transliterations", and I am strongly opposed to them myself. However, in this case there is a standardized and non-Arabic-based transliteration to work with, and I think there is absolutely no reason not to use it, especially insofar as that is the policy of ALA-LC—which you, earlier on this page, said that you support—on Ottoman Turkish. Any proposed Manual of Style (Turkish) page would be fruitless, because effectively there is almost no issue. —Saposcat 06:01, 23 August 2006 (UTC)

If there is a published standard, and if it's ALA-LC standard to use modern Turkish standards for Ottoman Turkish, then it should be easy and fruitful to make the Turkish MOS. At the least, there should be a referenced section on the Arabic MOS to Ottoman Turkish. I imagine there are details that need to be worked out. For example, someone just changed Abdülmecid to Abdülmecit. Also, there can be a hazy gray line between what article titles should be in Turkish or Arabic spelling. A good example is Rumi, who had an Arabic given name, born in present day Afghanistan (then Persia), travelled throughout Persian and Arabic lands, then settled in present day Turkey, and his descendants today call themselves Turkish. So what category would he be under? There are constant revert wars on that page as to his identity. Should it be standardized by birth or self identification or something else? <b style="color:#AF7817">Cuñado</b> - Talk  01:26, 24 August 2006 (UTC)


 * Again, I don't think there is a need for an MOS pertaining to Turkish, or even to Ottoman Turkish. Instead, I will work on a separate section in the Arabic MOS detailing the fact that Ottoman Turkish is (or at least, should be) an exception to the guidelines on the Arabic MOS page. I will use ALA-LC as a basis, but also have a look at other romanization systems as well. Any help in this regard would be greatly appreciated.


 * As for the "Abdülmecid" vs. "Abdülmecit" bit, the former, with "d" (ذ) rather than "t" (ت) at the end, is the correct transliteration for pre-spelling reform names, as per Cretanforever's explanation below.


 * Also, the example of Rūmī that you give is not actually a very good one, because, despite the efforts on that page (which I am very familiar with) of Turkish nationalists or pseudo-nationalists to enlist him as a Turk, he was not Turkish. Category-wise, he would probably be under whatever goes for Persian transliteration. As for his descendants, they are the Çelebi family, which uses the standard modern Turkish transliteration of چلبی. —Saposcat 06:42, 26 August 2006 (UTC)

I do some editing on Ottoman-related topics. I have no pretension in getting all details right all the time, but I say to myself, this is what wikipedia is about, we correct each other's mistakes (although for a move like that of from Abdülmecid to ‘Abdu’l-Mijid, it is more civil to have a prior discussion or at least a notice, it may concern a number of people). The case differs from the Chinese context where you have several dialects (all admitted to be part of the same linguistic body - the Chinese language), each of which may have different pronounciations for the same chararacters, but then you have a standard Chinese (Mandarin) on which transliterations are based. Persian, Turkish etc. have always been distinct languages from the language whose script they were using (Arabic), although heavily influenced. Other languages used the Arabic script as well for differing periods, sometimes surprising ones. I would certainly opt for sticking to the modern Turkish standard for the transliteration of Ottoman (and pre-Ottoman) names, even if, at times, I am not comfortable with the fact that an English-speaker will not immediately interpret a "c" to be "dj" and may lose direction with "ğ", "ç" and "ş". My worry is less acute with the wovels, because I guess (I am only guessing) that umlauts are generally recognized for what they change in the pronounciation in our day.

It is true that no one transliterated that Ottoman sultan's name into Latin characters as Abdülmecid in his lifetime (I would also be very surprised if anyone transliterated it as ‘Abdu’l-Mijid). Abdulmejid has been the most frequent use I assume. Therefore, it is a novelty. But a novelty that lasts, as Saposcat remarked, since the thirties.

As for Abdülmecit, it is a horror that stems from a number of Turkish linguists proclaiming that no word in Turkish could end with a -d. The evolution, within the Turkish language, from -d into -t for pronouncing and writing the last letter of a word is pertinent in some cases, but -in my opinion- it is far from being a generality, and an Ottoman sultan (or any other historic personality) certainly do not fit into the category. Cretanforever

A possible error in the guideline?
In the "Alphabetization" section, the first point says "Alphabetize by family name in modern cases where there is one, otherwise by the first component in the commonly used name, e.g. ibn Khaldun under i, Mu'awiya under m." Shouldn't ibn Khaldun be alphabetized under K? Point 4 says to ignore ibn. If so, perhaps we should use a different example, such as Abdul Aziz.

Speaking of which, what's the standard and strict translation of Abdul Aziz? Abd al-Aziz? – Quadell (talk) (bounties) 19:13, 24 August 2006 (UTC)
 * Most likely, Abd ul-Aziz. And technically, the term is transliteration, since it is alphabets we are switching between, and not languages, really. --Fsotrain09 19:25, 24 August 2006 (UTC)
 * Transliteration, yes, that's the word I mean. I'm not convinced "Abd ul-Aziz" is correct. The guideline says that the definite article should be written "al-" and not "ul-". – Quadell (talk) (bounties) 20:30, 24 August 2006 (UTC)
 * The correct transliteration is either `Abd al-`Aziz or `Abdu'l-`Aziz (ignoring long vowels). In strict transliteration it would be ‘Abd al-‘Azīz or ‘Abdu’l-‘Azīz. Nobody has decided that one form or the other is better, one being phoneticly correct and the other ignoring combination of words. <b style="color:#AF7817">Cuñado</b> - Talk  06:09, 25 August 2006 (UTC)
 * Why not ‘Abd ul-‘Azīz, which both respects word boundaries and is phonetically correct? --tyomitch 08:53, 25 August 2006 (UTC)
 * Well I may be wrong, but no standards out there transliterate that way. ALA-LC uses ‘Abd al-‘Azīz, and many many others use ‘Abdu’l-‘Azīz. This is supposed to be based on what academic and international standards exist, and as far as I know none of them would use ‘Abd ul-‘Azīz. <b style="color:#AF7817">Cuñado</b> - Talk  14:39, 25 August 2006 (UTC)

poll for standard transliteration
There have been a number of debates lately about Arabic transliteration, and the MOS seems to be failing because nobody wants to keep the same standard. I think part of it a matter of education, but also preference. I've tried to summarize the main issues by using the following example, which demonstrates how quickly the standard can turn into a mess. These all come from the various forms of Arabic transliteration used by various international organizations and academic standards.

A figure everyone should be familiar with: Gamal Abdel Nasser (جمال عبد الناصر Jamāl ‘Abdu’n-Nāṣir)

Keep in mind these are for the standard transliteration, and not the strict one. Please comment on the separate issues, and please be familiar with the MOS before commenting. <b style="color:#AF7817">Cuñado</b> - Talk

hamza and ayin
They cannot be ommitted, and they must be different. They both strongly affect pronunciation, and in different ways. We could use the grave accent: ` and the apostrophe  ' , both located on the keyboard, but people think the grave accent is ugly, especially when it's in the middle of a word (compare Shi`a vs Shi‘a). Another option is the half rings ʿ and ʾ, but they're not supported by Explorer browsers and turn into squares. I suggest using the typed turned comma ‘ and typed apostrophe ’, both on the clickable insert tab. They are the choice for ALA-LC, UNGEGN and other prominent standards. <b style="color:#AF7817">Cuñado</b> - Talk  01:30, 29 August 2006 (UTC)

phonetic and grammatic
I just made these words up, I'm not sure what the technical term is. I strongly suggest using the phonetically correct version, which is easier to pronounce and carries the same information. <b style="color:#AF7817">Cuñado</b> - Talk  01:30, 29 August 2006 (UTC)

solar letters
For the same reasons, I suggest incorporating the changes brought by solar letters. For non-native speakers, they will have no idea that al-Nasir is actually pronounced an-Nasir, and it carries the same information. <b style="color:#AF7817">Cuñado</b> - Talk  01:30, 29 August 2006 (UTC)

Ottoman Turkish
As an aside, I suggest we vote on whether to use Turkish or Arabic standards for Turkish people under the Ottoman Empire. I suggest Arabic, unless someone makes clear guidelines on Turkish, which currently don't exist. <b style="color:#AF7817">Cuñado</b> - Talk  01:30, 29 August 2006 (UTC)

comments

 * Guidelines in which Ottoman Turkish would have certain specifities (as treated on the basis of a number of exceptions to the standards existent for Arabic, as suggested by Saposcat with, at the same time, an eye for modern Turkish in order not to alienate the descendants, as well as the international reader) would be a good point of departure. That makes three different angles to watch. The same rules could also apply to the pre-Ottoman era. Cretanforever


 * Hamiş - Addendum: For transcription guidelines for Ottoman Turkish, the following document could be useful - Ottoman historians Cretanforever


 * That's a good find, Cretanforever, and overall a good standard to use, I think. However, I would make a couple of alterations in its strictness to accord better with actual pronunciation (which is, of course, the whole purpose of transliteration in the first place).


 * For long vowels, I think "â", "î", and "û" would be better than "ā", "ī", and "ū", insofar as the circumflex is still fairly standard in modern Turkish, as well as most Ottoman transliteration systems (to the best of my knowledge); certainly it's more common than the macron.
 * I don't think there is a need, in Ottoman Turkish transliteration, to write " ق " as "q", insofar as the letter was (and is) phonetically (and phonemically) pronounced not as /q/ but as /k/. —Saposcat 09:02, 30 August 2006 (UTC)


 * I haven't commented on these pages before, but I have been following much of the discussion over the past couple months. I really think it enhances Wikipedia when a standard is decided upon, and is followed.  (That is why there is a general MoS, and other MoSs). I'm also in favour of following a standard that is academic, and contains all the information.  In general I'm in favour of almost all of Cunado19's suggesstions.  In particular, I would use the grave accent: ` and the apostrophe  '  in the title and the lead, and we can use the apostrophe in the rest of the article. Having the characters on the keyboard are really important. In regards to the solar letters, I would use how it's pronounced, since non-native speakers of Arabic, really don't understand the al- etc.  I took me a long time.  Anyways, that's my input. -- Jeff3000 02:49, 29 August 2006 (UTC)


 * Ditto, for my part. I definitely support rendering sun letters via pronunciation rather than orthography.  As for the grave - while it can admittedly be aesthetically displeasing in some situations, speaking as one who has gone through many an article correcting or raising the level of transcription, the simple utility of having the characters on the keyboard is indispensable.  Article re-transcription is a slow, dull, head-wracking process, and those little niceties here and there are greatly appreciated. Keldan 05:53, 29 August 2006 (UTC)


 * i support using the MESA/IJMES/EI3 system, with the modification of representing the sun letters. i also think it is in the general benefit to represent 'the sun' as 'ash-shams' instead of 'al-shams'. as far as ottomans, generally the accepted standard is to use modern turkish orthography and i agree that it makes more sense to spell `abd al-aziz as 'Abdülaziz'. thank you for being on top of all this! dgl 03:46, 29 August 2006 (UTC)


 * Thanks to Cunaldo for putting this poll together. I support the phonetic transliteration over the grammatic one. I support incorporating the solar letters into al-. As for hamza and ayin, I think the grave accent and apostrophe should be used, as the turned comma and typed apostrophe are very difficult to distinguish in many smaller fonts. I have no opinion on Turkish or other non-Arabic languages. – Quadell (talk) (bounties) 03:52, 29 August 2006 (UTC)


 * I have no opinion on Arabic, Persian, Urdu, or other languages, but as for Ottoman Turkish, I think that it would be best to conform (roughly) to modern Turkish transliteration methods. I say "roughly" because, in fact, I would prefer to conform to ALA-LC's system of transliteration, which largely respects those standards, and can be found here in downloadable PDF format. —Saposcat 05:57, 29 August 2006 (UTC)


 * Just an extra note about the reasoning behind my arguments for using modern Turkish transliteration methods for the Ottoman version of the Arabic script. Insofar as—to the best of my knowledge—transliteration fundamentally exists as an aid to pronunciation, there is a basic problem with applying Arabic-based transliteration methods to Ottoman Turkish. For example: the name عبدالعزيز would be rendered as `Abd al-`Azīz in (one variety of strict) Arabic-based transliteration, indicating an (Arabic) IPA pronunciation of (I may be a bit wrong here, so bear with me) /ˌʕabdalʕaˈziːz/. On the other hand, an Ottoman (or modern) Turkish speaker would pronounce عبدالعزيز as /abˌdylaˈziːz/. This is, in point of fact, a large difference from the Arabic pronunciation, and one that would be better rendered as it is, in fact, virtually always rendered: "Abdülaziz" (or, to be perfectly strict, "Abdülazîz"). —Saposcat 21:02, 29 August 2006 (UTC)


 * I'm not going to argue for a long time about this a I won't be using the standard transliteration (primary and strict will do just fine). I'd try to keep the standard tranliteration close to the strict transliteration (ALA-LC). I would transliterate the ayin with a typewriter apostrophe and drop the hamza (the standard transliteration is all about simplicity, we're dropping the underdots as well). At least don't commit the typographical crime of using a "grave accent". —Ruud 07:11, 29 August 2006 (UTC)
 * for simplicity, I strongly suggest that whenever close transliteration is needed, people should use DIN 31635, employing the ArabDIN template. Thus, ǧamāl ʿabdu-n-nāṣir. For Hamza, Ayin use ʾ, ʿ or ᾿, ῾, but not ’, ‘ (which are punctuation symbols)! Make sure to segment the i`rab properly (ʿabdu-n-nāṣir, not ʿabd-un-nāṣir!) I see no harm in using j for ǧ (don't regulate things to death), so that Jamāl ʿAbdu-n-Nāṣir is fine with me, too. Really close transliteration (like ǧamālu ʿabdu-n-nāṣiru) is probably only needed in philological/qur'anic topics. (ᛎ) qɐp 09:06, 29 August 2006 (UTC)
 * I have an Explorer browser on XP professional 2002. Almost all the characters you just wrote are squares to me (ie, my browser doesn't support them). I see this: ǧamāl ʿabdu-n-nāṣir. For Hamza, Ayin use ʾ, ʿ or ᾿, ῾, but not ’, ‘ (which are punctuation symbols)! Make sure to segment the i`rab properly (ʿabdu-n-nāṣir, not ʿabd-un-nāṣir!)
 * Since we're talking about article titles, these would be unacceptable. ~Cunado
 * Please! we are not talking about article titles at all. Try to avoid further confusion. Obviously, nobody is suggesting to move Gamal Abdel Nasser to ǧamāl ʿabdu n-nāṣir. We are discussing recommendations for technical transliteration of Arabic, not article titles (in the case of names; I see no problem with grammatical terms like iʿrāb residing at their DIN transliteration, but that's a wholly different issue!). And there is no way that we let these recommendations be guided by broken browsers. I've been over these exact points so many times for so many languages on so many talkpages that I am extremely bored with them now. Why cannot we just have a clean poll with options documented on Arabic transliteration (1. SATTS, 2. UNGEGN, 3. ALA-LC, 4. DIN-31635, 5. ISO 233)? All details would follow from such a decision. At present we should just recommend using some official standard and stating which standard you are using. Concerning idafa, I recognize the hyphen is unnecessary, thus ʿabdu n-nāṣir would be more correct than ʿabdu-n-nāṣir. But ʿabduʾn-nāṣir (which I corrected in your table, assuming a simple oversight) is just wrong. dab (&#5839;) 11:14, 9 September 2006 (UTC)


 * I'm for the grammatic/solar variant, unless there's a clear policy declaring when consecutive words are to be joined with a hyphen. Otherwise there are just too many variants to choose from (e.g. Zaynab bint Khuzayma, Zaynab bint-Khuzayma, Zaynab-bint-Khuzayma, etc.) --tyomitch 17:27, 29 August 2006 (UTC)
 * A hyphen only goes after a definite article... How is that confusing? Zaynab bint Khuzayma, Harun ar-Rashid, Abu Bakr ibn Abi Qahafa ... ~Cunado
 * In the revision of the proposal that I was commenting on, the phonetic variant was given as Jamal `Abdu-n-Nasir, with two hyphens. It's kind of not nice to change the proposal in the middle of a poll.
 * Anyway, is now `abd the only word which gets merged with the following definite article? I really don't see why the phonetic variant is Jamal `Abdu'n-Nasir and not, say, Jamalu`Abdu'n-Nasir. Also, to let you know, there's no hamza there, so no need to put an ' --tyomitch 03:50, 30 August 2006 (UTC)


 * Sorry I didn't realize the table was changed. That demonstrates why I'm so frustrated and asking for consensus. `Abdu-n-Nasir is a completely improper use of the hyphen. As to Jamalu`Abdu'n-Nasir, that's a good point. I just looked through tons of examples and didn't see any words combined unless it was followed by a definite article, and even then sometimes it wasn't. <b style="color:#AF7817">Cuñado</b> - Talk  04:50, 30 August 2006 (UTC)


 * Another downside of the phonetic transliteration is that it's irreversible: it's impossible to tell from it whether the joined words were written together in the original Arabic script. IIRC, the reversibility was your major point in promoting the ticks for ayin and hamza; by promoting the phonetic transliteration now, you're contradicting your own old arguments. --tyomitch 15:23, 30 August 2006 (UTC)


 * of course the transliteration should be reversible. In fact, transliteration schemes add information not contained in the original script (such as hyphens). Would you care to elaborate why you think that " `Abdu-n-Nasir is a completely improper use of the hyphen."? Note that there is no hamza in the article, and I have no idea why your table puts a hamza where there is none. We can discuss how to transliterate hamza, but we can hardly debate about transliterating letters that aren't there in the first place. dab (&#5839;) 11:02, 9 September 2006 (UTC)


 * dab, I'm not sure what to tell you, but there is a hamza there. The alif is just a seat that the hamza sits on (or under). That's pretty standard Arabic knowledge, so if you don't know that (ال) is a hamza plus a lam, then you need to review Arabic. <b style="color:#AF7817">Cuñado</b> - Talk  03:38, 10 September 2006 (UTC)


 * Nothing of the kind. See the article on Alif. --tyomitch 17:47, 10 September 2006 (UTC)


 * OK I saw. It supports what I said. <b style="color:#AF7817">Cuñado</b> - Talk  05:23, 11 September 2006 (UTC)


 * I am opposed to a phonetic transliteration as defined above, for a couple of reasons. First, there is no standard system for it; Wikipedians basically have to cobble one together. We're making something up that doesn't exist out in the real world. (For example "Abdu-n-Nasir" produces one Google hit, from a blog.) Second, it opens the door for users exposed to diverse Arabic pronunciations to quibble with the transliteration ("this is supposed to be phonetic, but no one says it that way," "that vowel is e, not i" etc.). I would really prefer the adoption of UNGEGN without underdots (the grammatic-with-solar system). --Cam 14:51, 30 August 2006 (UTC)


 * I have no idea what you are talking about. What "phonetic transliteration"? We are looking at the simple matter of looking at which existing transliteration standard we should endorse. There can be no question of 'cobbling together' one of  our own. We have a handful of standards to choose from. I recommend DIN. That's the long and the short of it, it really isn't complicated. dab (&#5839;) 11:02, 9 September 2006 (UTC)

Standard transliteration
"The standard transliteration uses a systematic convention of rendering Arabic script into English which is used and standardized by academics and linguists. The current proposal for the standard transliteration from Arabic to Roman letters is found below."


 * This is blatantly false/misleading. Academic and scholary works always use a strict tranliteration (including the overlines, underdots and half rings), see the list of reference at al-Khwarizmi. The standard tranliteration is only used in low quality print works and on the internet (probably both due to technical limitations). —Ruud 10:05, 29 August 2006 (UTC)


 * The word "always" is rarely true. In fact, some scholarly works use a less strict translation that does not distinguish between every set of various characters. See, for example, the ISO 233-2 standard. But I see your point. Perhaps the paragraph should be changed? – Quadell (talk) (bounties) 12:44, 29 August 2006 (UTC)


 * We need something in between strict and chaos. It would be impractical and undesirable to use the dots and underscores in the article text. Having no standard results in reverts and chaos, as there are literally hundreds of possible ways of spelling things. See Muammar al-Qaddafi. <b style="color:#AF7817">Cuñado</b> - Talk  14:35, 29 August 2006 (UTC)


 * The standard transliteration is useful and should stay. The question is, should the introductory paragraph describing it be rewritten? – Quadell (talk) (bounties) 17:11, 29 August 2006 (UTC)


 * "Always" might have been an overstatement, but "usually" would have been an understatement :) Could you give a few examples where a standard would be preferable over strict translitartion AND a primary transliteration is not available? I agree it would be unnecessary to consistently use Shīʿah instead of Shi'a. But I would consider the latter to be a primary transliteration. —Ruud 22:19, 29 August 2006 (UTC)


 * A good example is the one I gave above: Muammar al-Qaddafi. The strict transliteration would read Mu‘ammar al-Qadhāfī and no primary transliteration exists (there are more than 32 ways of spelling it). Keeping it with diacritical marks throughout the article would be cumbersome and unless it is universally applied, it would remain forever inconsistent. So an in between is Mu`ammar al-Qadhafi, which can be typed on the keyboard and read by any browser. I would support using a stricter standard, but as you can see by Striver's comment below, there are two extremes that have never agreed on this issue. Striver is advocating getting rid of any kind of standard at all. ~Cunado

Guys, i hate to see Shīʿah instead of Shi'a. Every time i see Shīʿah or Nikāḥu’l-Mut‘ah, i start to have high blood preasure. I want to see "Shi'a" and "Nikah Mut'ah". I do not want to see 'Umar or 'Ali, rather Umar and Ali. I dont mind seeing the hated version in a etymology section, but not everywere in the article. And there is a loot of people supporting my view, just see Wikipedia talk:Manual of Style (Islam-related articles)/names. We are divided in two sections. I do understand why you people prefer 'Umar, its more accurate to the Arabic version. But thing is that we do not care for that. Why cant we have an etymology section were we go though the pronounciation, the strict Arabic and whadever more, and then use the relaxed version in the rest of the article? --Striver 23:32, 29 August 2006 (UTC)

al- and solar letters
If I understand correctly, many articles are currently at the wrong location and should be moved. For instance, Al-Razi should be Ar-Razi, and Mohamed Atta al-Sayed should be at Mohamed Atta as-Sayed. Perhaps these article titles also have other errors as well. Would it be an appropriate time to move these, or should we wait for more discussion? – Quadell (talk) (bounties) 14:51, 6 September 2006 (UTC)


 * I moved to Muhammad `Ata as-Sayyid and requested the move for the other. I have been fixing articles for months, I wouldn't be shy about it if I were you. <b style="color:#AF7817">Cuñado</b> - Talk  03:49, 10 September 2006 (UTC)


 * Please don't move those! Al-Razi is commonly known as al-Razi, not ar-Razi. I've only seen solar letters used in some German texts (DIN uses solar letters), but never in English ones (ALA-LC doesn't use English letters either). —Ruud 06:45, 10 September 2006 (UTC)


 * And why are we using a backtick? Your "standard" transliteration, I assure your, is very non-standard. Can we please wait moving/updating articles before we've agreed on what the standard transliterions is? —Ruud 06:49, 10 September 2006 (UTC)


 * Some arguments for not using solar letters and backticks:
 * It deviates from common "standard transliterations", making it useless for picking a prefered "primary transliteration" among others.
 * It doesn't help people write names, as solar letters aren't transliterated this way in English and the backtick is problematic to write down.
 * It doens't help pronouncing the name. Most speakers of English can't correctly pronounce an Arabic name from the original Arabic or a strict tranliteration; the standard tranliteration loses some information making it impossible to pronounce correctly anyway. Why bother differentiating between ayin and hamza?
 * It would be strange to tranliterate solar letters in the standard tranliteration, while not using them in the strict tranliteration.
 * It would be strange to use a backtick in the standard transliteration, while primary tranliterations use an apotrope (for example shi'a ).
 * —Ruud


 * Ruud, your argument doesn't hold water. Every standard of transliteration (UN, ALA-LC, DIN, EI) all differentiate between the hamza and ayin, and they never omit either of them. I can tell by your comments above that you don't actually understand the issue here. See this chart and you will see that they either use the typed apostrophe and turned comma, or the right and left half rings. The use of the grave accent ` (what you call the backtick) and the apostrophe ' are a shorthand that can effectively replace the more complex symbols and make typing easier for editors. The consensus so far has been to use the solar letters and the shorthand diacritics. By the way Shi`ah should not have an apostrophe. It is out of complete laziness that people put a hamza mark, when it should be an ayin mark. The pronunciation of Shi'ah would be markedly different from Shi`ah. <b style="color:#AF7817">Cuñado</b> - Talk  05:20, 11 September 2006 (UTC)
 * My "standard trasliteration" referred to the standard transliteration as defined in this MOS, not to e.g. ALA-LC (which this MOS calls strict transliterations). —Ruud 22:28, 11 September 2006 (UTC)

Re: Proposal for Ottoman Turkish standards
I propose that, for transliteration of the Ottoman Turkish language, we go with the ALA-LC standards now outlined in the table found on the Ottoman Turkish language page. It is not only the ALA-LC's standard, but is also used in many standard textbooks of Ottoman Turkish in Turkey, and can to an extent be considered to have gained general acceptance, as can be seen—to give just one example—in its being fundamentally the same as the transliteration system used on the Ottoman Text Archive Project. It is also more accurate a reflection of Ottoman Turkish pronunciation (as per, for example, Omniglot) than the Arabic- (or even Persian-) based standards that have been proposed.

Any opinions on the issue? —Saposcat 19:42, 6 September 2006 (UTC)


 * I think this is the third or fourth time I've said this: nobody has opposed this idea, and everyone seems to support it. But nobody has actually contributed to the policy pages. I suggest someone passionate about this should edit the current page and add a section about Ottoman Turkish, in the same manner that Persian and Urdu are handled. At the very least, write out the details of the policy, like what standards to use, where to find them, and some details on what articles would require Turkish instead of Arabic or Persian, or any other language. <b style="color:#AF7817">Cuñado</b> - Talk  06:09, 7 September 2006 (UTC)


 * I'm working on it, don't worry. صبر صبر ... —Saposcat 14:48, 7 September 2006 (UTC)


 * تمامدر ... it's done (with a slight glitch because the solar letter table is at the very bottom—any idea how to fix up positioning & appearance on that?). —Saposcat 09:48, 8 September 2006 (UTC)

Sorry! I have been away from the discussion for a few days. But as Talleyrand once remarked on an employee he found sleeping at the office, "At least, he is here!". There are tons of knowledgeable English-speaking people in the Turkish wikipedia who could comment or correct or even slightly develop here. I left a note, but no reaction to date.

Question: Do we differentiate between transcriptions for article headings and those within the text of the article? Do we somehow simplify the headings? Omit the circumflexes for example? Or do I change Sefaretname heading into Sefâretnâme for example, according to the the standardized transcription and we do redirects for simplified versions, that people will seek more readily? Cretanforever

Mohamed Atta
The article at Mohamed Atta al-Sayed was recently moved to Muhammad `Ata as-Sayyid. There's discussion here on whether this was the right thing to do or not. Any input would be appreciated. – Quadell (talk) (bounties) 13:12, 13 September 2006 (UTC)

Poll results
I'll try to summarize the poll above regarding the standard transliteration. Forgive me if I made a mistake. On some of the issues, I didn't side with my original proposal. <b style="color:#AF7817">Cuñado</b> - Talk  03:42, 15 September 2006 (UTC)

Hamza and `Ayin
Use the grave accent ` and apostrophe '. support:Jeff3000, Cunado19, Keldan, Quadell. Oppose: Ruud.

Solar letters
Incorporate them. Support: Jeff3000, Cunado19, Quadell, Dgl, dab, Tyomitch, Cam

Definite article
Use the "phonetic" (`Abdu'n-Nasir) instead of grammatic (`Abd an-Nasir). Support: Jeff3000, Dgl, dab. Oppose: Tyomitch, Cam, Cunado19


 * I did not support the transliteration of hamzat al-wasl (the  '  in  'n)!! "`Abd an-Nasir" is not "grammatic" in any sense (Abdu n-Nasir is grammatical, while I suppose "`Abd an-Nasir can be called "orthographical"). I really think the poll should be redone by someone who is aware of the subtleties of Arabic grammar and orthography. dab (&#5839;) 14:54, 16 September 2006 (UTC)


 * I'm plenty familiar. I've been following the standards already mentioned several times. Most of them use the orthographical (I just made "grammatical", I did't know what the correct word was), and Encyclopedia of Islam uses the phonetic version and always transliterates the hamza when it's not the initial letter of the word. After we had this discussion I went and read through a book on Arabic grammar and the definite article. There is definitely a hamza there, but in pronunciation and combined with the previous word, it becomes a hamzatu'l-wasl and is not pronounced, unless the previous sound is a vowel, then it becomes a different kind of hamza and is pronounced. I've never seen anyone use the form of "`Abdu n-Nasir", and you haven't shown any evidence that it's used by anybody. <b style="color:#AF7817">Cuñado</b> - Talk  19:50, 16 September 2006 (UTC)