Wikipedia talk:Manual of Style/Arabic/Archive 4

Solar letters again

I'm going to post this agian as I have not received a single counter-argument above: We should not use solar letters in the standard transliteration because

1. The strict transliteration we chose (ALA-LC) does not use solar letters. Using solar letters in the standard transliteration would be confusing to the reader.
2. Most English-language literature does not incorporate solar letters (see for example the references mentioned here, Britannica, Encarta, Columbia and the Encyclopedia of Islam. Therefore, if we incorporate solar letters in the standard transliteration, the following situations can occur in a single article:
   • Word A has a primary transliteration (which very likely does not incorporate solar letters)
   • Word B has two very common transliterations (which both very likely do not incorporate solar letters); as a compromise the standard transliteration should be used which does incorporate solar letters. —Ruud 10:02, 16 September 2006 (UTC)

sigh, again, I will continue to use DIN (with {{ArabDIN}}), which is a perfectly well-defined international standard, regradless of the frustrating discussions here. From this follows that I will "use solar letters". People who want to use (US centric) ALA-LC are free to do that, of course. If you want to establish consensus to favour one or the other in MoS, make the vote as simple as: (a) DIN-31635; (b) ISO 233; (c) ALA-LC (pick a, b or c, end of story), without all the confusing options and sub-options which will not be understood by half of the voters anyway. dab (ᛏ) 15:02, 16 September 2006 (UTC)

I think you're missing something. The strict form is close to ALA-LC, but the solar letters are an exception. We have actually taken pieces of several standards. So in other words, both standard and strict transliterations should use solar letters, according to this MOS. Cuñado - Talk 19:53, 16 September 2006 (UTC)

I thought, the strict for was exactly ALA-LC? Fully agree with dab that we should use an existing and established standard; Wikipedia should be descripitive, not prescriptive. I was willing to have some good discussions about the "standard transliteration", but if we can't even settle on the "strict transliteration" this MoS is going nowhere. —Ruud 22:13, 17 September 2006 (UTC)

I understand Ruud's argument, (and in specific instances it might make sense to make an exception to this proposed MoS, and not use solar letters), but I think the advantages of using solar letters, by being closer to the pronounciation is more important. -- Jeff3000 21:59, 16 September 2006 (UTC)

Ok, fine. I'll be using ALA-LC (without solar letters) as it is consistent with the sources I base my articles on. —Ruud 21:52, 17 September 2006 (UTC)

It's ALA-LC with solar letters. If you want to change the policy then gain a consensus. Don't pout like a child on the playground and say that you're going to do your own thing. Cuñado - Talk 01:42, 19 September 2006 (UTC)

<sarcasm>I could probably claim to be using a "primary translation" 90% of the time...</sarcasm> (constructive comment will follow later) —Ruud 22:53, 19 September 2006 (UTC)

"Solar letters again" again

Moved to Wikipedia talk:Manual of Style (Arabic)#Alphabetization. - Anas ^talk? 20:35, 22 March 2007 (UTC)

Self-identification

There was a discussion on Talk:Mohamed Atta about his name. There is really no primary transliteration of his name, but he actually used the spelling of Mohamed Atta for himself, and had it on his driver's license. I think we should add a clause and say that if an individual self identified with a particular spelling or form, then that should be used as the primary. This wouldn't apply to anyone who lived in more ancient times when Arabs/Persians didn't use Roman letters. Any supporters? Cuñado - Talk 19:57, 16 September 2006 (UTC)

I think it should depend on whether the majority of the person's notability was involved in spelling his name with Roman letters or Arabic letters. This is a little confusing, but let me explain. Mohamed Atta is only famous because of what he did in America, while using Roman letters to spell his name, so his own spelling should take precedence. But I think Muammar al-Gaddafi is a counter-example. He is only known to have spelled his name with Roman letters once, in a letter to an American girl, and he used the spelling "Moammar El-Gadhafi". But he overwhelmingly used Arabic script, and all of his fame is involved in his living in the Arab World, speaking in Arabic. So I don't think we should necessarily use the "Moammar El-Gadhafi" name, simply because he used it once. We should balance how much the person used Roman letters, somehow, but I'm not really sure how to balance this. – Quadell ^{(talk) (random)} 22:40, 16 September 2006 (UTC)

Self-identification can still be problematic (Gadaffi comes to mind, self identiefies with several different names, which also differ from the most common transliterations). However the spelling used on a driver's license and especially passport would carry significant weight. —Ruud 22:03, 17 September 2006 (UTC)

I agree with both of you. I'll try to update the page with something. Cuñado - Talk 01:48, 19 September 2006 (UTC)

And just where are you going to get this driver's license and passpoort information, without violating Wikipedia:No original research? Gene Nygaard 14:55, 21 January 2007 (UTC)

If the passport/license spelling is published somewhere, would it really be original research to report it with a citation? --Cam 15:03, 21 January 2007 (UTC)

No. Just pointing out that basing the rules on something that won't generally be known and verifiable doesn't make all that much sense. Gene Nygaard 19:25, 21 January 2007 (UTC)

I think there is an obvious assumption with this: that the identification is published somewhere, or some that there is a source in which the person published their own name. Cuñado - Talk 22:02, 21 January 2007 (UTC)

I liked the self-identification idea immediately, but then remembered when I had my name transliterated into another language for business cards, the translation was done not by me but by a translator, and simply copied and pasted by me onto the business card without any understanding of the characters I was copying. Because of this, I would personally not put 100% weight behind the translation that happens to be on those business cards - maybe they're erroneous; I don't really know. Gadhafi may have done the same thing. Atta, having lived in the US, used that particular spelling personally many times, including on the ID, so that surely carries more self-identificatory authority than the Gadhafi example. Tempshill (talk) 19:01, 12 December 2007 (UTC)

Persian again

My expertise is Persian rather than Arabic. I first became aware of the problems of transliteration when looking over the article on Abbas Kiarostami. My understanding is that in the lead sentence the Persian script should be used, and a strict "Arabic" transliteration given, in this case `Abbās Kiyārustamī, which is what I recommended they use. But they didn't much like the u, because to almost all Iranians u represents the long vav/waw. Similarly, i usually represents the long ye/ya.

I very much doubt whether there are half a dozen Persian/Iranian articles that follow the MOS strict transliteration in their opening sentence. They can't even get Mohammed Mossadegh right, spelling him Moḥammad Moṣadeq in the first sentence, not the presumably correct Muḥammad Muṣaddiq. The problem is that people using the dominant Tehrani—or should I say Tihrānī?—pronunciation can't come to terms with those short is and us. The long/short a/ā distinction is less of a problem, because there's no obvious alternative.

Is there room for a compromise, I wonder? One way forward might be to use e and o for the short vowels, while keeping the macrons in ī and ū for the long vowels. In any case, some lines of communication ought to be opened up with Wikipedia:WikiProject Iran about all this. --NigelG (or Ndsg) | Talk 13:05, 24 February 2007 (UTC)

As far as I know, the macrons ī and ū are, in fact, used only for the long vowels, like in Sūriyah (سـورية) and Sūrī (سـوري). Or perhaps I didn't understand you well? - Anas ^Talk? 15:07, 24 February 2007 (UTC)

Exactly. My point is that Iranian WPians seem to be happy to use ī and ū for the long vowels; but they don't like using i and u for the short vowels (e and o would be closer to the pronunciation). Left to their own devices, most Iranians would use e and o for the short vowels & i and u (without macron) for the long vowels. I am suggesting that they might well compromise & agree to add the macrons to indicate the long vowels (as in Arabic). All this refers to the strict transliteration in the first sentence of an article, to be used alongside the Persian script. The standard & primary transliterations are less of a problem.

Examples of the proposed compromise: Eṣfahānī = اصفهانی , ḥoqūq = حقوق . This would lead to the transliteration Moṣaddeq for مصدق. There doesn't seem to be an established way of transliterating the ending in words such as خانه : the most generally acceptable method would be khāneh (the -ah ending sounds unPersian to most Iranians).

One final problem is how to deal with the Ezafeh (strictly Eḍafeh) اضافه , as in دوست عزيز ("dear friend"). The most acceptable way would be to use -e: dūst-e `azīz. This becomes -ye after an -eh ending: خانه‌ی ما ("our house") = khāneh-ye mā. NigelG (or Ndsg) | Talk 18:28, 24 February 2007 (UTC)

There is no reason to use a single standard for transliterating all names which are originally written in Arabic-based alphabets. For example in transliterating Persian and Ordu, using characters such as ḥ and ṣ (instead of h and s) is not only redundant, but simply wrong. This manual of style is primarily proposed for transliterating classic Arabic and Islamic names Although it serves this purpose very well, it is quite wrong to generalize it for all Arabic-based alphabets. In my opiniion, it is not even suitable in transliterating modern Arabic names such as Abdel-nasser's given name (because it doesn't reflect the differences in modern Arabic dialects). Also, it is quite wrong to claim that there is a consensus on using this manual of style for all those alphabets (just look at Iran-related or Pakistan-related articles to see that using this guidline for non-Arabic languages, is not popular among wikipedians). Indeed, the history of this manual of style and it's talk page doesn't show much participation of wikipedians (except those who have been active in Islam-related historical articles). Jahangard 18:31, 24 February 2007 (UTC)

If you read the main page carefully, you'll see that there's a clear distiniction drawn between strict, primary & standard transliterations. The current discussion is entirely about the strict (or "scholarly", for want of a better word) transliteration. I think you're really talking about the standard transliteration, which the MOS quite explicitly says should follow the pronunciation.

The entire discussion here relates to the method of transliterating the Arabic/Persian letters used in the first sentence of most Iran-related articles. --NigelG (or Ndsg) | Talk 19:00, 24 February 2007 (UTC)

I know, and my comment is about that. Perso-Arabic letters are shown in those introductions (showing it twice is totally redundant). The main purpose of transliteration should be giving an idea about the original pronounciation. This manual of style was primarily proposed for the classic Arabic names in Islam-related articles (following the classic Arabic pronounciation). This guideline serves that purpose very well (and is almost the same as what scholars and orientalists use for that purpose). However, using it for other purposes , such as non-Arabic languages, is quite wrong. Also, calling it "scholarly" is correct, only when it is used for classic Arabic. For other purposes, the transliteration standards used by the scholars and academisians (e.g. Iranologists) are different. Jahangard

Fine: just quote some reliable, widely accepted sources that we can use as a basis for discussion. You might like to look at this document, for example, which compares a number of systems.

Any encyclopedia has to have consistent rules—& WP is no exception. This forum is where the rules for Persian are going to have to be agreed: it would be a great pity if Persian were to be the only language without an agreed system of (strict) transliteration. --NigelG (or Ndsg) | Talk 21:20, 24 February 2007 (UTC)

PS Here's a reference with fuller details on the Library of Congress scheme. --NigelG (or Ndsg) | Talk 21:36, 24 February 2007 (UTC)

Indeed, the sources that you cited show that even for the strict transliteration, the standards which are used for Persian are different from Arabic (for example, for letter "ض", most of them use a variant of character "z", instead of "ḍ" which is used in the Arabic standard transliteration). However, my main point is not about the standards for the strict Perso-Arabic transliteration. I'm saying that in the first sentece of those articles, we should have the proper transcription, not the strict transliteration. The purpose of strict transliteration is to show the correct spelling in the original alphabet (when the original characters are not available). When the original characters are shown in the article (e.g. in the Iran-related articles of Wikipedia), there is no point in strict tranliteration and showing the transliterated version is totally redundant. Instead, the proper transcription should be shown to give the average reader of the article (who doesn't know that language) an idea about the correct pronounciation of the word. For the classic Arabic names in Islam-related articles, WP:MOS-AR is very useful, because, it is a proper transcription for the Qur'anic Arabic. But, for other articles (specially those with non-Arabic titles), WP:MOS-AR is not only useless, but also misleading. Jahangard 00:56, 25 February 2007 (UTC)

Obviously this manual of style is set for the literary Arabic language, and it serves its purpose very well. This language is used all over the Arabic word, in Islamic books and the Qura'n, and almost 90% of all published media (as the Modern Standard Arabic). Now although almost every Arab country has a different vernacular language, like Levantine Arabic and Egyptian Arabic, in an encyclopedia strict literary Arabic should be used. However, I've understood from your comments that it is somehow an inaccurate guideline to follow when it comes to articles with Persian names. I oppose modifying this MoS, because it would confuse editors (more than it already does). There needs to be a solution for this problem; my suggestion is that a separate and different MoS be started for the Persian language. It can, perhaps, be based on the Arabic MoS, with applying modifications to what needs modifying, changing what needs changing, or even removing what doesn't suit the Persian language, its grammar, and the pronunciation of its words. Seeing how I am not really experienced in the Persian language, nor allegedly anyone who has worked on this MoS, I suggest that the members of WikiProject Iran team up and start a Persian MoS. Do you concur? - Anas ^Talk? 12:25, 25 February 2007 (UTC)

Hm, work on a Persian MoS seems to have already been started by Sangak. I support starting and completing this spin-off, as long as there is consensus (there already is, I suppose) on doing so. All the best. - Anas ^Talk? 12:37, 25 February 2007 (UTC)

The details of Persian transliteration (TL) can indeed, & probably should, be discussed there. But there is one fairly fundamental point of principle that ought to be agreed on here: should a "strict" TL be used at all in Persian- & Iran-related articles? Jahangard's assertion that such a TL is unnecessary & redundant is at least open to question. One issue is whether allowance should be made for users whose software cannot render the Arabic script. There are other details, of course, which would perhaps be better discussed on the new forum: eg should initial `eyn be dropped from the TL?; can q & gh be used interchangeably?; etc etc. But we had better clear up the basic principle of having a strict TL as soon as possible. --NigelG (or Ndsg) | Talk 19:21, 25 February 2007 (UTC)

From my point of view, a strict transliteration is not required, but it is rather recommended. Using it is not always redundant; for example, the article on the city of Cairo needs the strict transliteration, al-Qāhirah, because it is completely different from the primary transliteration, which is Cairo. In cases where the strict transliteration is redundant, or very close to the primary transliteration, use the strict only, like Al-Kindi. Nigel, the MoS guidelines are very suitable for the Arabic language (in 90% of the cases); I see it fine as it is, but if there are some things that don't work well with Persian, discuss and change them in the Persian MoS. Simple. Cheers! :) - Anas ^Talk? 12:13, 26 February 2007 (UTC)

Actually, it may surprise you to hear that I don't have any particular axe to grind—apart from preferring to see a consistent approach used on WP wherever possible. Would you be happy to see Persian articles using x and š where Arabic uses kh and sh, for example? That is certainly one approach being discussed at the moment. --NigelG (or Ndsg) | Talk 16:34, 26 February 2007 (UTC)

I'm afraid I didn't understand you well. I've also lost track of the discussion, could you explain your comment please. I agree that WP should aim for consistency, but if what applies to Arabic doesn't apply to Persian, then maybe consistency can not be achieved. - Anas ^Talk? 17:02, 26 February 2007 (UTC)

Persian (continued)

[The last thread was getting a bit long, so I'm continuing it here.]

My point about consistency was the following. There are a number of sounds which are roughly the same in Persian & Arabic: these include uncontroversial ones such as b, d, t, etc; & some others, such as kh, sh and q, which Iranian authors might decide to transliterate/transcribe differently from the accepted Arabic equivalents, even though the sounds are the same (they might choose to use x, š and gh, respectively). I was merely raising the question: should it be a WP guideline that both languages use kh, sh and q for these letters? --NigelG (or Ndsg) | Talk 17:27, 26 February 2007 (UTC)

I see. In Arabic none but the š are correct, but it is less commonly used (AMOS was based on ALA-LC Romanization, I think). In my opinion, consistency is always the better and preferable path to take, especially when it is correct in both language pronunciations. Lest we forget, most of the readers in WP are English, and might not really know how to correctly read the aforementioned alternatives; this is why the AMOS was based on ALA-LC, because it is both correct and user-friendly. - Anas ^Talk? 17:56, 26 February 2007 (UTC)

Another issue is whether a strict transliteration (TL) is needed at all, even in the opening sentence. Jahangard, for instance, thinks not: he favours a "correct transcription"—a phonetically correct version for foreign readers unacquainted with either Persian or IPA. Actually, IPA might be a better approach (given that it's quite widely used in WP): that's something to consider.

What it boils down to is this: is it important to be able to reconstruct the Persian spelling from the TL given in the opening sentence? If so, the TL is going to have to use diacritics to distinguish between the letters pronounced T ( ت ط), H (ه ح), S ( س ص ث) and Z(ز ض ذ ظ ). If this 1:1 requirement is dropped (as Jahangard points out, the TL is redundant if the Persian script is there), the solution is much simpler: you just give the pronunciation hāfez for حافظ and leave it at that. At least it would stop people saying hafīz!

One final point: the assumption that all Persian-speakers can read the Persian script is false, particularly for 2nd-generation émigrés. <weasel>A sizeable number</weasel> of WP users looking up Persian- & Iran-related articles will fall in this category, & their interests ought to be considered too. --NigelG (or Ndsg) | Talk 11:41, 27 February 2007 (UTC)

I honestly think a strict transliteration is necessary sometimes, and when it is redundant then just use it by itself. If what you mean by "Persian script" is the original Persian script (e.g. ايران), then of course it is not enough, which is why the strict transliteration is utilized. I think the idea of a strict transliteration is different somehow from IPA; it is used to show how a word is equivalently spelled (transliterated) in English from the original language and how it is pronounced. Both, a strict transliteration and an IPA pronunciation, are not required. And actually strict transliterations are used by not more than 0.5% of the articles with Arabic names.

I'm not really getting your point; do you mean to say that a strict transliteration can be replaced by a pronunciation? If you think about it, it's the same thing, and I actually prefer a transliteration because using transliterations based on guidelines leads to uniformity and consistency. I'm sorry, I'm having a busy day, so excuse me for my quick response. - Anas ^Talk? 17:03, 27 February 2007 (UTC)

No, a strict TL is not the same thing as a pronunciation—certainly not in Persian, at any rate. As I've just explained, 4 different Arabic letters are all pronounced as /z/ in Persian.

I'm not making any specific recommendations here, just trying to elucidate the main issues. The details are being discussed over at WP:MOS-PE. --NigelG (or Ndsg) | Talk 18:40, 27 February 2007 (UTC)

So, I assume that every letter has its own transliterated form, right? This means using a transliteration is more accurate, since as you indicated, four letters have the same pronunciation; we can distinguish these letters from each other by using a transliteration. As for WP:MOS-PE, it seems that work is being done by a specific group of users; I'm not so sure these users represent the Wikipedia community. Shouldn't everything be discussed before working on the new MoS; I noticed that the method or standards that the strict transliteration (a debatable section) was based on were chosen directly and not discussed at all. - Anas ^Talk? 19:30, 27 February 2007 (UTC)

Well, there's no harm in letting them discuss the details on WP:MOS-PE & come up with a proposal that suits them. The proposal can then be discussed with the WP community (or at any rate those who have an interest in these matters!) here or on their Talk page. --NigelG (or Ndsg) | Talk 20:56, 27 February 2007 (UTC)

PS Doesn't anyone else want to join in?! --NigelG (or Ndsg) | Talk 20:58, 27 February 2007 (UTC)

I'll be there. :) - Anas ^Talk? 11:31, 28 February 2007 (UTC)

Alphabetization

The discussion on the use of solar letters seems to be fairly abstract and seems to not make reference to actual usage in the Arab world. Perhaps a reference to practical usage might be helpful. Usage of Arab names in English in the United Arab Emirates (where I live and teach in a university just for Emirati citizens) is fairly consistent -- and the usage here fits with usages in Egypt, Saudi, Oman, and Yemen (I've checked with various Arab faculty and just double-checked a variety of official government & media style guides for these countries). These conventions are used by all of the major English language newspapers in these countries, has been developed through extensive local consultation, and serve as the standard in the UAE for all government offices. The resulting consensus is as follows:

(1) sun/moon letters are not considered at all as they confuse the issue of the structure of the name -- yes, the pronunciation changes, but the "al" is an "al" is an "al" (or an "el-"). To suggest that spelling should change due to pronunciation would be like suggesting that "Louisville" be spelled "Loo'vul" in Wikipedia because it is pronounced that way by the people who live there. (To be honest, after many years here in the Middle East working among Arabs, today on Wikipedia was the first time I had ever seen the use of a sun letter in transliterating an Arabic proper name!);

(2) the al- or -el are always given as a part of the family/tribal name [sometimes with the hyphen, sometimes without] in formal usage -- it is this way on passports of all Emirati too (their passports use both Arabic and English). Conversationally one might refer to an entire family or tribe omitting the al-/el- (such as talking about the Maktoums), but reference to an individual always uses the al-/el-;

(3) alphabetization, when it is done by family name, is done by the al- or el- , not by the name that follows. Truth be told, alphabetization here is more often done by first name rather than family/last name, even in the phone books;

(4) secondary references to people referred to initially by their full name, e.g. Shaykh Zayed bin Sultan al-Nahayan, vary. First names predominate in the Gulf, so for the "royalty" of the region a title and a first name is correct: Shaykh Zayed is referred to as Shaykh Zayed. Shaykh Mohammad bin Rashid al-Maktoum would never be referred to as "Maktoum" but possibly as Mohammad bin Rashid, if not Shaykh Mohammad. For everyday sorts of people, a secondary reference is likely to be to the family name, e.g. al-Futtaim or al-Jabbar. Note that the al- or -el is lower case.

This actual usage would seem to be appropriate usage for Wikipedia. It is wide-spread within the Arab world, serves as the base for media world-wide, and was developed by Arabs for referring to themselves in English. So, I would recommendation the adoption of the al- or el- prefix, not creating a difference in transliteration with sun letters for proper names of people and places, and alphabetizing under al- or el- as appropriate. Cyg-nifier 18:22, 20 March 2007 (UTC)

I do agree with much of what you have suggested. Solar letters are, in my opinion too, not appropriate to use in Wikipedia, and they are usually not. They are, however, used in strict transliterations, which needs to indicate the solar letters because that is one of the purposes of a transliteration — show how a word is spelled with the corresponding letters of another language, which is in this case English. I'd have to disagree with the third point; it would seem inappropriate to alphabetize names according to the al-s or el-s; it's like placing Robin van Persie under V, which methinks is illogical and unreasonable. By the way, I've placed this discussion under a subsection. Cheers. - Anas ^talk? 18:43, 20 March 2007 (UTC)

I'll report back tomorrow with some research on alphabetizing practices actually used here in the Middle East. In the meantime, I've a curiosity question: why are all the references here to "solar" letters? Every person I know who teaches Arabic here in the UAE, in Yemen, and in Egypt (and all of my students who are nationals) refer to the letters as "sun" letters and "moon" letters. My understanding is that the letters that cancel out the l in al- and double the initial letter of the noun are called "sun" letters poetically because the Arabic word for sun /shams/ is an example of this usage. The letters that keep the full "al-" are then called "moon" poetically because the Arabic word for moon /qamar/ is an example for this use. "Solar" seems to suggest (to me anyway) something related to solar/lunar calendars or something more scientific rather than poetic. Cyg-nifier 18:48, 21 March 2007 (UTC)

It is more accurate to say solar letters than sun letters, because in Arabic it is āḥrof shamsīyyah and āḥrof qamarīyyah, not āḥrof (al-)shams or āḥrof (al-)qamar. However, both names are accepted and correct. Sun letters are more commonly used though. Wikipedia's article on the letters uses Sun letters and a Google search yields more results. It's not a big deal anyway. I think I have a yellow pages directory somewhere. I'll try to look for it later and report back. - Anas ^talk? 22:10, 21 March 2007 (UTC)

OK. My report back on alphabetizing. I've done a sampling of universities in the Middle East which are English-using and intended for either all or mostly nationals, government agencies, and academic journals that are in English, about the Arab world, and based (or co-based) in the Middle East. Where possible I have looked at actual style guides; where not, I've reviewed publications by a university/agency/journal to extract their "house" practices. The results were over-whelmingly consistent: family names keep the al-/el-/ibn- attached to the rest of the family name at all times (whether there is a hyphen or not) and when alphabetizing was called for, whether in an index, bibliography, list of faculty or authors, the al-/el-/ibn were an integral part of the list. The only style guides I could find that did not follow this practice originated outside of the Middle East (such as for The Guardian).

So, the following is just a few sources, but are representive of virtually all I found. Here in the UAE both the Ministry of Information & Culture and the Ministry of Communications have a style that alphabetizes with al- and el- (and the odd ibn-). Zayed University, UAE University, and the Higher Colleges system (with Dubai Men's/Dubai Women's, etc.) all alphabetize using the family name with al- or -el intact as well as ibn when it is linked with the family name (as w/ Ibn Khaldun and Ibn Battuta); these universities are exclusively for nationals, who tend to be reasonably vocal if anything smacks of slighting dignity, so I doubt we'd be using systems that could possible be seen as inappropriate. The library catalog for these universities use the al- and el- for cataloging works by Arab authors and alphabetizing the catalogs (this is true whether in English or in Arabic). This practice is also used by the library in Alexandria (bibliotheca alexandrina). American University in Cairo, which is a major center for teaching Arabic as well as a major Egyptian university, does also. The yellow pages phone book in English and in Arabic keeps the al-/el- (no one ever seems to claim every having seen a print copy of a white pages in existence, but everyone consulted -- from Arab faculty/students to media/journalism faculty assumes they would be alphabetized by first name).

I looked at the journals Arab Media Society and Arab Studies Quarterly; both use al- and el- and alphabetize using them. Publishers Dar El Ilm Lilmalayin (in Beirut) and Dar Al-Fikr (in Damascus) do as well -- and their practicer seems to be to consistently use the hyphenation as a means to hold the al-/el- with the name. This also makes sense in seeing that in all cases, the citation references (whether parenthetical or as in-text references) to people with al-/el- constructed family names are to be with the al-/el- intact, for example, al-Futtaim or al-Marri. These citation references then form the base for a list of references/bibliography, which would thus naturally keep all the al-/el-'s together alphabetically. Similar references are the standards for publications from the League of Arab States and for ALECSO (Arab League Educational, Cultural, & Scientific Organization) as well as from the UN and UNESCO. Also the Center for Arab & Middle Eastern Studies (Beiruit). As a side note, Edward Said's practice fits with this.

The Middle East Media Guide is the primary source for identifying media outlets in the region -- the 2006 issue (right here in front of me) is especially interesting as a resource. Al- and El- move with the names of people, news agencies, newspapers, etc. at all times. And, yes, the A's look to be the length of listing for Smith or Jones in the New York City phone book. It is also clear that all newspapers and television stations that use "al- or el- ALWAYS have this united with the rest of the name: Al Jazeera, for example, is the organizational name and not Jazeera, al. There are dozens of such newspapers across the Middle East. Obviously when the al- or el- is part of the formally/legal title of an entity such as a newspaper, then it would always need to be capitalized as well. Surely if they see the Al- as an integral part of the name, it makes sense for us to assume this as a practice. ---As a linked issue --> it also is clear in looking at practice that the al- or el- would NEVER be separated in references to place names. Al Quds (Jerusalem) would thus not be alphabetized in English as "Quds, al" -- street names and city names clearly carry the al- & el- with them too. Certainly in Dubai, Abu Dhabi, and Sana'a (the cities of which I'm most knowledgeable) no one would know how to find a street or city on a map or in a directory if the ones using al- had the al- separated from then.

These are just a few examples, but as I said, this use is over-whelmingly consistent. The al-/el- in virtually every place I check was used with no difference due to sun/moon letters (with the sole exception of a few German libraries/dictionaries), the al-/el- remained intact with the family/place name, and alphabetization was with the al-/el-/ibn intact. These are all pretty significant and credible sources -- and all are in the Middle East. It would seem if these practices are so widespread here when the language of use is English, then it is appropriate for them to be the practice of Wikipedia. This suggests that the "al-/el-" is not functioning just as a parallel to the use of the definite article "the" in English, but that its use has altered -- Arab place and person names that use do carry the al-/el-/ibn- with them into English actually become an integral part of the name itself and don't function as a definite article "the" -- and thus can't be separated off from the name in text. This seems a natural part of language evolution and very definitely reflects current journalistic and academic practice. (As a side note, I teach in a college of Communication & Media; my journalism and pr colleagues, who I've consulted over the past few days, assure me that this is the practice that is being taught to our undergraduates and to the many professional journalists who do training with us, and that it reflects practice throughout the Arab world for English-language media. It is, according to them, considered to be the appropriate modern practice --and believe me, we have protocol officers out the wazoo here who would obliterate us if this practice were not considered "halal".) warmly, Cyg-nifier 17:10, 22 March 2007 (UTC)

Wow! Amazing research! This most definitely settles the issue. I found that yellow pages book I was talking about — was Arabic, didn't help much. OK, now that we've set the records straight, we'll need to modify the actual guideline. Any thoughts before I change anything? - Anas ^talk? 21:16, 22 March 2007 (UTC)

standard transliterations vs. transcriptions/anglicizations

this page needs to disambiguate cleanly between transliteration standards we recommend using at the introduction of a term, viz. ISO 233, DIN 31635 (Qur’ān), and a suggested standard anglicization (Qur'an). It is flawed and misleading to refer to "Qur'an" as a "transliteration". The guideline should state: if you want to transliterate, use ISO or DIN. If you merely want to transcribe / anglicize, use our proposed standard transcription unless there are good reasons to prefer another spelling. dab (𒁳) 15:27, 14 April 2007 (UTC)

I'm also confused about the guidance the page intends to give. I just moved Al-Ma'arri (ambiguous and using the character ' which displays vertical for most people, something that wouldn't be found in a well-edited published book) to Al-Maʿarri—was that a goof? If there's a Wikipedia policy not to correctly transliterate the names of articles about Arab persons, I missed it & would like to be pointed to it. Wareh 21:39, 18 October 2007 (UTC)

Yes that was a goof, according the this policy it should be Al-Ma'arri. There are too many people who don't understand the difference and getting certain editors to follow any standard is work. On that page, the first line in bold should have the most accurate transliteration of the name. Cuñado ☼ - Talk 02:10, 19 October 2007 (UTC)

Well, I'm sorry for rocking the boat, but let me point out some respects in which this article is a mess. (1) "The current proposal for the standard transcription from Arabic to Roman letters is found below." That would be very valuable, if the link were not broken (and I sure can't figure out where it should point). (2) The only plain and concrete guidance seems to be in the Examples. But according to this "standard," it should be Al-Ma`arri, not Al-Ma'arri. Please note for the record that today Al-Ma`arri is a redlink—and that's for an article that has twenty-one existing redirects. If folks have made 22 article names, and not a single one of them is the one recommended by this page (as far as I can tell), then it seems pretty clear that this (not particularly correct, scholarly, or rational) standard has not been effectively imparted to Wikipedia's collective faculty of understanding. (3) It sure doesn't help that the only example proffered in this important introductory section of a "correctly" named article with an apostrophe of any kind is Ta'if (correct by "Examples" standard), but the article itself is presently located at Ta’if (using a character nowhere recommended for scientific or informal transliteration!). I hope no one will see laziness or the wrong motives in my decision to leave Al-Maʿarri with its overly-correct name. I humbly suggest that anyone thinking about moving it should set a higher priority on sorting out this policy better. Finally, at the risk of adding to an already negative diagnosis, I find it utterly unhelpful that Wikipedia:Naming conventions (Arabic) is being maintained for "historic interest." If it's truly obsolete, it should have been harder for me to stumble on first—maybe converted to a redirect and relocated in some subarticle archive space. (And what is Wikipedia talk:Manual of Style (Islam-related articles)/names but an even obscurer alternate version of this page where someone who "hates" strict transliteration leads the discussion?) Wareh 03:38, 19 October 2007 (UTC)

P.S. The proposal here against ʿ for ʿayin and ʾ for hamza was a year ago. Since then browsers have done better and better at dealing with Unicode, bigger character sets, etc. These are the correct and typographically attractive characters. Is there any support for (considering) switching this page's advice to recommending their use, especially since this old "system" has not gotten much traction anyway? Wareh 03:57, 19 October 2007 (UTC)

You're preaching to the choir. There are dozens of editors that don't care about transliteration and just do whatever they want. That results in most articles being transliterated wrong. The issue is not so much whether or not browsers can support it, but whether or not editors will use whatever standard we decide to use. The current version is simple and uses characters on the keyboard. The number of redirects doesn't matter, see Muammar al-Gaddafi. I didn't notice that the character should have been an `ayin, I didn't look at the Arabic, my fault. It should be Al-Ma`arri then. Cuñado ☼ - Talk 14:47, 19 October 2007 (UTC)

If the problem is that no one respects the simple right-on-the-keyboard policy recommended here, then what purpose does it serve? If I'm hearing you correctly, the only people who read Manual of Style and Naming Convention pages are the same people who do not need a character to be located on their keyboards in order to use it. I'd think committed editors of the sort who want to help implement good policies would be more inspired to go around changing article names if the results were more satisfying. I'd recommend strict transliteration, with half-rings, minus all diacritics for article titles. Right now, '-for-ayin titles like Al-Ma'arri (before I moved it) and Muhammad 'Ali Bay al-'Abid are so widespread that it would seem more sensible to go ahead and move such articles over to the use of half-rings (which are unlikely to be done wrong, precisely because they require more effort to encode). Moving to Muhammad `Ali Al-`Abid is such a marginal improvement that it's hard to work up the inspiration to do it. Of course, there should be plenty of redirects as needed. What do you think? I don't want to do any more moves if I'll just be upsetting people who believe in keeping the rules stated here. Again, isn't the first step to improve & clarify the recommendations here? I really do think that if the page gave lucid and reasonable advice (& to the point: this is how to transliterate Arabic for a page title, this is how to transliterate it when giving Arabic in the article—period—the other niceties can be saved for Romanization of Arabic, not here), it would attract a larger cadre of editors itnerested in implementing its guidelines. Wareh 19:07, 19 October 2007 (UTC)

A new problem has arisen in the case of 'Alawi, which had been named Alawite. Other editors argue that the latter term is the "common" English usage. I have countered that the former is what is used by scholars of the topic in most authoritative sources. A parallel can be seen in teh conflicting spellings of Shiite and Shia. Is there any formal policy on these kinds of discrepancies?Jemiljan (talk) 00:24, 8 December 2008 (UTC)

Well, I visited some of these pages from library computers running Internet Explorer, and I realized the world does not appear ready for the half-rings yet. I still feel that the existing proposals are an unsatisfying compromise, but I'm no longer confident in what direction improvement lies. Perhaps any progressive changes are premature, and I can't immediately come up with better advice than the current proposal, basically "strip the word of all diacritics, use ' for hamza, use ` for `ayn." There are still all kinds of grey areas (idafa and assimilation, e.g. `Abd al-rahman vs. `Abdu-r-rahman vs. `Abd ar-rahman), but I no longer view it as a tractable mess. Wareh 19:26, 7 November 2007 (UTC)

I believe it was settled already that `Abd ar-rahman is the correct form. The rationale is that assimilation makes no difference to accuracy, but helps pronunciation from non-native speakers. Library of Congress, the standard that most resembles the WP standard, does not use assimilation. Cuñado ☼ - Talk 20:52, 7 November 2007 (UTC)

Transcription or transliteration?

The summary box at the beginning of the article states the following:

"This page in a nutshell: Arabic names on Wikipedia should use a standard transliteration of Arabic, unless a primary transliteration exists. A strict transliteration should generally not be used."

But in the article itself--and the tables in the article--the terms used are "Primary Transcription", "Standard Transcription" and "Strict Transliteration". Shouldn't the box be changed to reflect what the article actually says?

Should be: "This page in a nutshell: Arabic names on Wikipedia should use a standard transcription of Arabic, unless a primary transcription exists. A strict transliteration should generally not be used." —Preceding unsigned comment added by Editor2020 (talk • contribs) 03:02, 27 January 2008 (UTC)

Bold Arabic letters

To my eye bolding the Arabic script of the name in the lead looks odd -- it draws unnecessary attention, and we don't bold any other script either. Is there a reason for this? Jpatokal (talk) 22:21, 30 May 2008 (UTC)

The Arabic seems to be much easier to read when it's bold. However, several editors have fought against that, so go ahead and remove it. Cuñado ☼ - Talk 16:07, 31 May 2008 (UTC)