Help talk:IPA for Arabic

From Wikipedia, the free encyclopedia
Jump to: navigation, search
          This page is of interest to the following WikiProjects:
Wikipedia Help Project (Rated NA-class, Low-importance)
WikiProject icon This page is within the scope of the Wikipedia Help Project, a collaborative effort to improve Wikipedia's help documentation for readers and contributors. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks. To browse help related resources see the Help Menu or Help Directory. Or ask for help on your talk page and a volunteer will visit you there.
 NA  This page does not require a rating on the project's quality scale.
 Low  This page has been rated as Low-importance on the project's importance scale.
WikiProject Languages (Rated Start-class, Mid-importance)
WikiProject icon This page is within the scope of WikiProject Languages, a collaborative effort to improve the coverage of standardized, informative and easy-to-use resources about languages on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This page does not require a rating on the project's quality scale.
 Mid  This page has been rated as Mid-importance on the project's importance scale.
WikiProject Writing systems (Rated Start-class, Mid-importance)
WikiProject icon This page falls within the scope of WikiProject Writing systems, a WikiProject interested in improving the encyclopaedic coverage and content of articles relating to writing systems on Wikipedia. If you would like to help out, you are welcome to drop by the project page and/or leave a query at the project’s talk page.
Start-Class article Start  This page does not require a rating on the project's quality scale.
 Mid  This page has been rated as Mid-importance on the project's importance scale.


Note that the example for IPA /a/ as the sound in "run" is wrong. Subject to dialectical differences, /a/ is the sound in "half" or "last".

Maybe the prior author intended "ran"?

Why this have hebrew vowels? (I know theres a realtion between the languages, but why shown whit the hebrew alphabet?)
Hi, firstly please sign you posts using four tildes ~~~~ at the end of what you write (and even consider creating an account). To answer your question, the author of the page created it by copying it from WP:IPA for Hebrew and evidently hasn't finished yet (or is unsure how to do it). Cheers Epson291 (talk) 07:44, 21 February 2009 (UTC)
I've removed the Hebrew, but we've got to adjust this for Arabic and probably deemphasize the orthography-based approach since Arabic uses an abjad. Which vowel allophones should we include? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 17:43, 23 February 2009 (UTC)

Here is a chart of the Arabic (and Hebrew) vowel system:

In a broad transcription we could use [a aː i iː u uː] but if this is a guide for non-speakers shouldn't we try to be as accurate as possible? That is, phonetic instead of phonemic. I prefer [ɛ̈ ɪ ʊ] for short vowels, [æː iː uː] for long vowels. Also note that the low vowels have back allophones [ɑ ɑː] around emphatic consonants and /r/.

Also we should note that [e eː o oː] are not native phonemes of Modern Standard Arabic, occuring mostly in loanwords and dialectal Arabic.

But I'm open to using other systems if they work better. (talk) 18:58, 26 February 2009 (UTC)

Arabic is a pretty complicated situation because the actual pronunciation depends on a speaker's native dialect. While we could have a more phonetic system, I'd like to point out that [ɛ̈] is incorrect for short /a/. If anything, it's [ɐ] but even in the most prestigious (i.e. prescriptive) forms, the actual realization of vowels is highly dependant on surrounding consonants in ways that may be too complicated and too varied for our wikipedia standard to reasonably incorporate. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:25, 26 February 2009 (UTC)
Should we make a note of it in the article under the vowels? Otherwise, it is a bit misleading as to which vowel is pronounced. --Anatoli (talk) 23:09, 26 February 2009 (UTC)
Make a note that that MSA pronunciation is dependant on dialect? Sure. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:10, 27 February 2009 (UTC)

If a phonetic transcription will be too complicated by all means go with a broad transcription, ie. [a aː i iː u uː]. I disagree about short /a/ being [ɐ] though. Not in any Arabic I know. But that would be irrelevant in the broad transcription I'm suggesting. (talk) 22:37, 1 March 2009 (UTC)

I am not too strong with Arabic vowels, as I haven't seen too much of IPA used with Arabic but I usually have to decide between /æ/ and /ɑ/ pronunciation (and long counterparts), a quick example: قلب /qɑlb/ - "heart" and كلب /kælb/ - "dog", . The trouble is that even the best Arabic dictionaries will only give fatha "a", kasra "i" or damma "u", without specifics of the accent. Not familiar with /ɐ/ either. Aeusoes1, please give some examples of Arabic vowels related to fatha or variants of /a/, if you can. Anatoli (talk) 23:21, 1 March 2009 (UTC)
According to Arabic phonology, [ɐ] shows up before a word boundary (which I believe means word-finally). I only brought up this particular phone because I think descriptions of Arabic /a/ as being [ɛ̈] may be hearing the [ɐ] since they're almost the same thing.
One way to work on this is to go backwards from pure phonemic transcription (this is the way we approached it at WP talk:IPA for SpanishWikipedia talk:IPA for Spanish. We start with the vowel phonemes /a aː i iː u uː/. Since, for example, a notable feature of Arabic is the retraction of /a(ː)/ to [ɑ(ː)] in the environment of certain consonants, we'll include that in our guide (though we need to be carefulabout the rules governing when this occurs). — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:20, 2 March 2009 (UTC)

It must be Wikipedia talk:IPA for Spanish. From what I gather there are no strict rules only a guide. Also, you use [ɛ̈], [ɐ] and /a(ː)/, not /æ(ː)/. Don't you think هذا has the sounds /æ/? Anatoli (talk) 00:59, 2 March 2009 (UTC)

Please don't confuse phonemic slashes with phonetic brackets. There is indeed a phoneme in Arabic that may be represented as /a/ though I don't know whether this is ever [a]. The /a/ of هذا may be [æ] though that may depend on a speaker's native dialect. WHat do you mean by "From what I gather there are no strict rules only a guide"? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 03:31, 2 March 2009 (UTC)
I meant the choice between /æ/ and /ɑ/, like in my example with "heart" and "dog" - the rule is not set in stone. The emphatic letters always require /ɑ/ but non-emphatic can have both, for example /r/ can be followed by /æ/ or /ɑ/ (even if no emphatic is around). I don't have all the examples ready, though.
If you say "depends on the dialect", then we can't define at least a version of standard Arabic? Say, take al-Jazeera's TV anchors as a guide? Not sure it would be easy either. Anatoli (talk) 04:00, 2 March 2009 (UTC)
I'm worried that we would then be violating WP:NPOV by picking one variety of Standard Arabic over another. We could represent the /a/ phonemes with [æ] unless they're next to emphatic consonants and pharyngeals in which case we use [ɑ] and then wait for people to complain before adjusting it. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 04:37, 2 March 2009 (UTC)
I like the idea, let's use [æ] and [ɑ] (+ the long equivalents), unless someone complains. Yes, I forgot the pharyngeals. Sound [ɐ] is used in Arabic phonology with a reference, need to use that as well somehow. Anatoli (talk) 04:52, 2 March 2009 (UTC)
Sounds good. Should we move [eː] and [oː] down with the other marginal sounds? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:44, 2 March 2009 (UTC)
Yes, please. Anatoli (talk) 11:29, 2 March 2009 (UTC)

I am content with this transcription system except for one thing: [a] (or [æ]) is indeed backed to [ɑ] around the emphatic consonants [tˤ dˤ sˤ ðˤ] and also [q] and I believe [r] but I'm not sure whether this occurs around the actual pharyngeals [ħ ʕ]. The article on Arabic varieties says no. Gotta be consistent here.AlexanderKaras (talk) 00:00, 3 March 2009 (UTC)

[ɑ] is pronounced in حال (condition) ([ħ]), غداً ([ɣ]) (tomorrow), عربي ([ʕ]) (Arabic). So, [ɣ]) can be added to the list. Anatoli (talk) 02:28, 3 March 2009 (UTC)
What about [x]? Wouldn't it be the same? Also, I'm not sure about [r] because I think it may not be triggered by all varieties. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 07:22, 4 March 2009 (UTC)
No, after [x] it's [æ], example: خمسة "five" - /ˈxæmsɐ/. I am not 100% about [r], despite my earlier comment. If I find something, I'll write here. I wonder if Janet C. E. Watson's The Phonology and Morphology of Arabic has anything. Anatoli (talk) 11:24, 4 March 2009 (UTC)
After a second thought, I could use some help with this. Too busy at work and other commitments but when I have a chance, I'd like to sort out the list of consonants that affect the choice between [æ] or [ɑ]. Anatoli (talk) 03:00, 5 March 2009 (UTC)
I've read Watson's book. She concentrates mostly on two specific dialects, Cairene and San'ani, at the exclusion of Standard Arabic. However, what she does say is on page 46 when she talks a bit about emphatic consonants. She contrasts pharyngealized coronals, labials, and /x ɣ/, (where pharyngeal constriction occurs in the upper pharynx) with pharyngeals (where it is in the lower pharynx). Spread from a pharyngeal results in vowel lowering, spread from a pharyngealized oral consonant "tends to result in centralization of [dorsal] vowels and both lowering and retraction of [guttural] vowels." Spread from a pharyngealized oral may affect the entire phonological word (for either dialect) while from a pharyngeal (or uvular stop in Cairene) it is restricted to adjacent vowels only. These differences are predictable

"...from the principles of phonetic realization… and are due to the combinatorial effect of the two places of articulation involved: with the pharyngealized coronals, primary [coronal] restrains the tongue dorsum to restrict the pharyngeal effect of non-primary [guttural] at the same time as non-primary [guttural] tends to retract the tongue tip from the dental region towards the alveolar ridge…; in the pharyngeals, the primary place is in the larynx and therefore the pharyngeal effect of non-primary [guttural] is not moderated in the mouth. The non-primary constriction of the pharyngeals is consequently lower and closer to the larynx than that of the pharyngealized coronals or the uvulars."

Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 06:48, 5 March 2009 (UTC)
Sounds good. Thanks for the quote. I will check recordings again with /x/ but I may have been wrong before saying that it's it's [æ] after /x/. Anatoli (talk) 07:15, 5 March 2009 (UTC)
I believe I've said elsewhere on Wiki that the realization of /a/ after /x/ depends on the speaker's dialect. It's why Ferguson referred to /x/ and /ɣ/ as semi-emphatic. But as for the assertion that "[ɑ] is pronounced in حال (condition) ([ħ]), غداً ([ɣ]) (tomorrow), عربي ([ʕ]) (Arabic)." I must admit that I am a bit baffled. [a] is only pronounced in حال in certain dialects (such as much of Syrian.) I am interested in hearing what speaker would produce an actual [ɑ] in such a word. One may, however, hear words like الحان (melodies) realized as [ælħɑːn] in the dialect of Aleppo. But even there, the retraction is hardly predictable. For example, a word like حتى (even, in order to) will always be realized as [ħætːæː]. As for عربي, the vowel retraction there is caused by the following [r], not by the preceding pharyngeal. Otherwise, one would expect a retraction of the vowel in words like معاناة [muʕæːnæː] (suffering) but such an expectation would be in error. As for ([ɣ]), I doubt you could say that a word like غداً has the same degree of retraction as a word like ضابط ([dˤɑːbɪtˤ]) (officer.) While Arab prescriptivists prescribe a value of [ɑ] for /a/ following [ɣ], this rarely happens in practice. For example, I doubt you could find a native Arabic speaker outside certain parts of Iraq or the Levant who would naturally pronounce a word like غابة (forest) as [ɣɑːbɑ] rather than as [ɣæːbɐ]. Even in readings of acrolectic texts, such as classical poetry, retraction rarely occurs. For example, listen to this recording of Antarah ibn Shaddâd's Mu'allaqa. The first words (هل غادر) are realized as [hæl ɣæːdærɑ]. Szfski (talk) 10:57, 5 March 2009 (UTC)
Do any of you listen to ArabicPod? Ihab Salih is the main teacher and uses about as neutral and steady a pronunciation as is possible; that is, with MSA. I think we'll all readily admit that the (contemporary) Arabic language and all of its dialects contain much more than the three vowels of the triangular system touted by many Arabicists (Classical, mind you). The difficulty with using IPA for MSA is that even there there are differences between country. It's like the daily news on TV in North America: it's still General American, BUT there ARE regional differences; some subtle, some not so subtle, but none such that viewers have called in to complain about the weather forecaster's pronunciation; that is, to my awareness. The basic fact of the vocalic matter can be summed up by a self-quote from the Wiktionary translations request page:

The emphatic consonants almost always alter the following vowel. Same goes with the pharyngeals ح and ع, sometimes غ and occasionally خ. These may vary from dialect to dialect but the important thing to remember is that consonants pronounced with emphasis or tension often make it difficult for the muscles of the throat and tongue to return to their normal position in time before the following vowel is articulated.

I doubt any Arabs feel any embarrassment over this issue. After all, it's not like it's anything to be ashamed of. However, and again, the various dialects render them (the vowels) differently. But ALL, as far as I know, DO have altered vowels after emphatics. Back to the ArabicPod thing, Salih's vowels are very regular and centralized. This means that there is rarely an /æ/ or /ɐ/. And these are almost always after an emphatic or pharyngeal; e.g., /naʕɐ̽m/. Now, why on earth did I indicate a mid-centralized near-open central vowel after the pharyngeal fricative? I did this to indicate the VERY centralizedness of this vowel. Sounds funny doesn't it? Well, that's the best way I can put it at this moment. Even if we find the "perfect" articulator of MSA even HIS vowels will differ from the norm. MSA phonology is very tricky, though it be more documented than other dialects, etc. At this point I think we should use a broad transcriptional inventory AT FIRST. Just so we can say for certain "OK, these vowels/consonants definitely ARE in MSA." Otherwise you might expect to see an edit-war sometime soon. Not from ME, mind you, as that kind of thing is not my kind of thing. So, to sum up, let's stick to the essentials. What can we all agree on? What are we not comfortably certain about? How can we find out more? Etc., etc., etc.—Strabismus (talk) 02:55, 6 March 2009 (UTC)
Thanks for joining the discussion, Strabismus. I wasn't so sure about the few consonants you mentioned. I remember seeing in a textbook something along those lines, similar to what you said but I can't find it anymore! How about we write what we are sure of and those consonants with "sometimes" can be marked as "sometimes"? Anatoli (talk) 03:24, 6 March 2009 (UTC)

Speaking about ɐ, ɛ, ʊ, ɪ[edit]

  • I was amazed to read claims that [ɐ] sound is prestigious! In Egypt, this is like, rare & probably rural & not prestigious at all! Regularly, /æ/, /ɑ/ are the vowels to be pronounced, either in Modern Standard Arabic spoken by an Egyptian, or Egyptian Arabic.
  • The short [ɛ] sound, or the like, is pronounced, as far as I know, in Tunisia & probably around Tunisia. the short &/or long [ɛ(ː)] are/is pronounced, as far as I know, in Lebanese Arabic & probably around Lebanon.
  • I'm not sure where is [ʊ], or the like, is pronounced. But I'm very sure that in the Standard “prestigious” Egyptian Arabic, it is absent & instead short [o] or/& [ɵ], and long [uː] are pronounced.
In the Standard “prestigious” Egyptian Arabic, short [e] sometimes is substituted with short [ɪ], which could be, as well, a shortened version of long [iː].
  • Always remember: Modern Standard Arabic pronunciation is highly dependent on the speaker's native variety of Arabic & to claim that the Modern Standard Arabic has a one standard pronunciation is bizarre! --Mahmudmasri (talk) 22:22, 8 October 2009 (UTC)

Try to paste some of the Arabic words into this text-to-speech engine: Acapela HQ TTS Interactive Demo (select Arabic). I have checked with some words I am sure of first. Hmm, حال and غداً seem to have [æ] but خمسة and أراد [ɑ]. Anatoli (talk) 09:37, 6 March 2009 (UTC)


Letters خ and غ are out of question, listened to ArabicPod (lesson 16) to confirm, they are both followed by [æ]. Anatoli (talk) 11:10, 6 March 2009 (UTC)
  • [ɑ] before or after خ might be pronounced, but rarely. Regularly, it is [æ]~[a]. --Mahmudmasri (talk) 22:22, 8 October 2009 (UTC)
I should mention, in the interest of completeness, that certain speakers will retract their vowels after these consonants. This is particularly characteristic of Arabic speakers who happen to hail from certain parts of Iraq, Syria and Morocco. It is also one of the preferred ways of reciting the Qur'an. Moreover, these tendencies do get carried over into MSA, as Watson and others have demonstrated. This is one of the reasons why I honestly don't think that a precise IPA transcription of an MSA phrase or word will be possible without making some kind of decision as to which dialect should be used as a substratum. If you need a source on these sounds, though, try Ferguson-
The consonants /q x ɣ/ in arabic generally function in part similarly to emphatic consonants in that in certain sequences the allophones of neighboring sounds are those expected in the vicinity of emphatics. For example, in some varieties of Moroccan Arabic, /a/ before these sounds has the 'emphatic' allophone, but after them does not. This quality of semi-emphasis associated with /q x ɣ/ is probably the explanation for the arab grammarians' term مستعلية applied only to these consonants. -The Emphatic L In Arabic, Language, Vol. 32, No. 3 (Jul. - Sep., 1956), pp. 446-452. Szfski (talk) 13:40, 6 March 2009 (UTC)
Has anybody read the article on Arabic by Alan Kaye in The World's Major Languages? It's rather cursory but I was just wondering. I think I should also state here that there may be danger in deeming (formal) Iraqi Arabic as the be-all and end-all of MSA phonology.
À propos the khā and ghayn, are you (Anatoli Atitarev) sure that what you heard was a plain near-open front unrounded vowel with absolutely no centralization? I recall you telling me that you are not too proficient with transcribing audially what you hear into the IPA:

No, I can't convert sounds to IPA easily but I am searching for an online source describing the Arabic phonology in a more or less standard IPA.

At any rate, it would behoove us to stick to the essentials but not forget those things about which we are not entirely certain. I do think it would be good to include—dare I say—an editorial giving some insight into certain subtleties for which we do not happen to have authoritative confirmation (yet). This may set off a few alarms with the Wikireactionaries but let them be alarmed; it is high time we (individuals) start giving our own two cents here at WP even if they only amount to a single cent.—Strabismus (talk) 19:06, 6 March 2009 (UTC)
Guys, I don't have much to add here, sorry. I expressed my observations, which may not be 100% accurate. I can tell the difference between [æ] and [ɑ]. At any rate, in غداً and خمسة it's not [ɑ]. What is it in your opinion? Anatoli (talk) 02:53, 11 March 2009 (UTC)

This isn't even IPA[edit]

It seems peculiar that this article should claim to be an IPA guide for Arabic when it includes symbols that aren't part of IPA at all. The use of macrons and digraphs seems...well...a bit weird. Szfski (talk) 09:41, 16 May 2009 (UTC)

Since many places incorporate a romanization system, it may help readers to compare that romanization system to the IPA. Perhaps rather than "trans" we should label the column something else. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 16:00, 16 May 2009 (UTC)

L in Allah[edit]

I removed the bit that said that the emphatic /lˤ/ occurs only in Allah. This is not true. It occurs elsewhere in complementary distribution with /l/ as well as in a large number of borrow words. (see The Emphatic L In Arabic, Language, Vol. 32, No. 3 (Jul. - Sep., 1956), pp. 446-452.) Just so no one wonders why I made the deletion. Szfski (talk) 13:35, 24 September 2009 (UTC)

Hmmm, we should have a note of clarification. What are the specifics? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 17:23, 24 September 2009 (UTC)
/l/ has an emphatic allophone [lˤ] in the vicinity of pharyngealized/velarized/emphatic consonants under certain conditions. For example, the word atˤwal (longer, taller) is realized as [ɑtˤwɑlˤ]. However, this "second-degree" velarization can be broken by any intervening instance of /i/ between the velarized consonant in question and the /l/. Thus, a word like tˤawīl is realized as [tˤɑwiːl].
Furthermore, there are plenty of instances in which the so-called "semi-emphatic" consonants غ and خ have been known to cause a phonemic /lˤ/ to occur as in the following examples from Iraqi Arabic
[ɣuːlˤ] غول "A demonic spirit" [χɑːlˤ] خال "Maternal uncle"
[ɣuːlˤiː] غولي "My demonic spirit" [χɑːlˤiː] خالي "My maternal uncle"
[ɣuːliː] غولي "Terrifying" [χɑːliː] خالي "Vacant"
This occurred as follows: first, the semi-emphatic /ɣ/ or /x/ colored the vowel of the word and velarized the /l/. The /lˤ/ was then extended to all occurrences of the word, whether or not there was a pronoun suffix attached. /l/ remained, however, where the following /iː/ was a part of the original word/stem rather than a pronoun marker since /i/ and /iː/ can, as I have already said, be shown to de-velarize instances of surrounding /lˤ/ or prevent surrounding instances of /l/ from being velarized within the stem. All this leaves /l/ and /lˤ/ as phonemic as can be.
Then there are cases of borrowings, such as the Moroccan Arabic word lˤɑmba (lamp) which got a velar pronunciation because the language of borrowing (probably Spanish, Italian or French) had either the vowel [ɑ] or [a], sounds which in Arabic only occur in the vicinity of velars, and which therefore velarized the /l/.
The article I cited goes into more depth on all of these points. Hope this helps Szfski (talk) 03:40, 25 September 2009 (UTC)
It does, though incorporating it into the key might be a pain. When you get a chance, you should edit Arabic phonology accordingly (since that was the source of the contested statement). If it is too complicated, we may just have to not indicate a velarized /l/ in this pronunciation guide. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 04:28, 25 September 2009 (UTC)
I edited Arabic phonology quite some time ago to reflect the phonemic /lˤ/. See note #4 in that article under the consonant chart for information on /lˤ/. I just added a sentence on the allophonic one, no extra sourcing was needed since the current sources support the edit already.
Come to think of it, maybe I'm getting ahead of myself. Allophonic /lˤ/ might just confuse most users of this transcription guide, which is intended for non-specialists. However, phonemic /lˤ/ should be included, I feel, perhaps with a qualifier stating that it most often occurs as a phoneme in forms of the word for God in addition to certain loanwords. I can only think of one cross-dialect MSA minimal pair for phonemic /lˤ/ and /l/: wallāhu [walːaːhu] "He empowered him" and wallāhu [wɑlˤːɑːhu] "And God." That said, I can think of several other minimal pairs for various dialects and regional pronunciations such as Iraqi and Sudanese. Szfski (talk) 17:18, 25 September 2009 (UTC)
Yeah, my concern here is that editors can look here and know when to put /lˤ/. Is there anything more specific than "certain loanwords"? Loanwords from certain languages? Or is it okay to be vague that way? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:59, 25 September 2009 (UTC)

Emphatics and vowels business[edit]

Footnote 1 claims the following:

Both the long and short a phonemes are retracted to [ɑ(ː)] in the environment of pharyngeal and pharyngealized consonants (i.e. the so-called "emphatic" consonants), namely /ħ ʕ tˤ dˤ ðˤ q sˤ/.

I've noticed that this isn't precisely true for the first two phonemes on the list. While I'm not sure if it's [æ(ː)] (I think it is, but can't be sure), the vowel sounds in عامل ʿāmil and حامل ḥāmil are most definitely not [ɑː] and are also definitely more front than [ɑː]. This also applies to short vowels: witness عمل ʿamala and حمل ḥamala. Or فعل faʿala, that most Arabic of verbs. What to do? Lockesdonkey (talk) 17:57, 29 October 2009 (UTC)

While it varies from dialect to dialect (even when speakers are using MSA), our article on Arabic phonology does not include /ħ/ or /ʕ/ in this list, though it's kind of surprising that retraction is weaker with these consonants. We should perhaps omit them in the list unless we find a definitive source including them. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:07, 29 October 2009 (UTC)
Aeusoes- ħ and ʕ do not cause the same kind of retraction because their pharyngealization is not a secondary articulation and they do not require anything other than the root of the tongue to move, allowing for a somewhat quicker release. (Also, in some dialects, these sounds are not pharyngeal at all and are in fact epiglottal, but that's another issue.)
Phonology-geekery side-note: /q/ in many dialects also does not cause as much retraction, since it need not involve the pharynx. In fact /ɣ/ and /χ/ are right there with /q/ in causing inconsistent retraction across dialects. Historically, however, /χ ɣ q/ did cause consistent retraction in the literary register (as they still do in traditional Qur'anic recitation) and were possibly more pharyngealized than they are today, as attested by early Arab grammarians who referred to a series of sounds consisting of /tˤ dˤ ðˤ q sˤ χ ɣ/ as musta'liyya which might be translated as "tongue-raisers." Szfski (talk) 09:22, 30 October 2009 (UTC)
What about /r/? Retraction often (maybe always, I'm not sure) occurs after /r/. Lockesdonkey (talk) 01:34, 31 October 2009 (UTC)
Arabic phonology lists /r/, though some dialects contrast an emphatic /r/ with a non-emphatic one. I'm surprised about /ħ/ and /ʕ/ because my understanding of the vowel allophony is one of pharyngeal spreading ([ɑ] is to [ʕ] what [i] is to [j] and [u] is to [w]). That it may often be epiglottal may help to explain this, though not very thoroughly. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 06:19, 31 October 2009 (UTC)
Btw. Aeusoes1, pharyngealization doesn't just cause a retraction of /a/. It affects the other vowel phonemes as well. Listen to <a href="">this recording</a> I made of a male Palestinian saying the words /suːra/, /sˤuːra/, /tiːn/ and /tˤiːn/ (meaning Qur'anic verse, picture, fig and clay, respectively.) My preferred explanation (as a speaker of Levantine Arabic, anyway) is that the apical and dorsal primary articulations' being changed first causes a delay in the change of the secondary pharyngeal articulation, thereby allowing said pharyngealization to have a more prominent effect on following vowels (in much the way the PIE o-coloring laryngeal */h₂/ caused rounding via secondary articulation in places where the primary articulation of PIE */w-/ did not.) Of course, this does not explain why many dialects have pharyngealization that affects both preceding and following vowels. Such a phenomenon would suggest that the whole word is "pre-pharyngealized" in the way voiced stops are pre-voiced in, say, French but not English. Moreover, in many dialects (including classical Arabic as used in traditional Qur'an recitation,) certain sounds such as /i/ can prevent pharyngealization from spreading past them or depharyngealize surrounding sounds (c.f. [ɑlˤːɑːhu] God as against [bismilːæːhi] In the name of God). The research displays an embarrassing lack of agreement on this matter, as it has since Ferguson brought it up half a century ago. Mind you, it is also possible to analyze Arabic as containing /ɑ/ and /æ/ as separate phonemes, and the pharyngealized obstruents as allophones in complementary distribution with the plain ones.
/r/ does indeed cause retraction, though I'm not as aware of studies examining that phoneme. I can only tell you that, after looking at a wave-form reading of the phoneme in my own speech, this is because /r/ is always somewhat pharyngealized, at least for me. However, I have no reason to think that this is true for speakers from non-Levantine dialect bundles or, for that matter, for speakers of non-Jerusalem Arabic.
I'll talk to some folks I know who might be able to get you a straighter answer than I can. Szfski (talk) 22:29, 1 November 2009 (UTC)
Gosh, this is getting too deep for this talk page. I think we all agree that we should take out the pharyngeals and add /r/ to the list here, correct? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:47, 2 November 2009 (UTC)

I edited the footnotes to remove the pharyngeals and add /r/. I'm not sure if /q/ should be included as well? Also, I think we should include that short /i/, /u/ are realized as [ɪ], [ʊ] as in كتب [ˈkʊtʊb] 'books.'AlexanderKaras (talk) 05:45, 20 November 2009 (UTC)

According to Arabic phonology, /q/ should be included. I don't think we should mention those allophones of the short high vowels, partly because I'm not sure how true it is (so far, I haven't seen any sources that say as such) and it might not be worth it to mark such a distinction in our pronunciation guide. What do you think, Szfski? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:56, 21 November 2009 (UTC)
/q/ does cause retraction, as it is an emphatic consonant. Semiticists don't typically group /r/ in the emphatic series. Both, however, do have the effect of vowel retraction in most versions of modern Arabic. A segment such as *[ræqæ] is not possible for most speakers.
Regarding the short vowels, it varies greatly from dialect to dialect. But some broad rules can still be said to apply to most MSA utterances.
Short /i/
When this sound occurs word-finally at phrase boundaries (which, in MSA, only happens with a few inflectional endings and pronouns), it is a close front [i] with the same quality as its longer counterpart. However, within phrases, when this sound is a case-ending, this is made more complicated by sandhi rules. When the following word begins with the definite prefix /al-/ or certain verbal derivational morphemes, the preceding short /i/ often becomes more open and less front (i.e. more like [ɪ]) and deletes the initial vowel of the following word. So, for example, the phrase "In the childhood of the king" would be realized as [fiː.tˤu.fuː.læ.tɪl.mæ.lɪk] though the phrases "In childhood" and "the king" on their own would be [fiː.tˤu.fuː.læ.ti] and [æl.mæ.lɪk], respectively. Likewise, when it is separated from a word boundary only by enclitics, particularly those few which are affected by lag assimilation, short /i/ generally remains close front.
However, in closed syllables where strict MSA phonology does not allow long vowels to occur (though, today, speakers will often delete the case endings and thereby cause them to occur,) short /i/ is almost always more open and farther back: [ɪ]. This is also generally true in open syllables that do not occur at a morpheme boundary. I'm confidant that if you listen to any non-Qur'anic recording of literary Arabic, you'll find this to be generally true. For example, listen to the first few words of this recording of Imru'l-Qays' Ode. The broad (phonemic) transcription of the first distych is /qifaː nabki min ðikraː ħabiːbin wamanzili bisiqtˤ ilːiwaː bajn adːaxuːli faħawmali/. You'll notice that the /i/ is in fact [ɪ] wherever it is not long, save at word or morpheme boundaries. In fact, pronouncing a short /i/ as [i] instead of [ɪ] in closed syllables may actually impede comprehension. For example, in Palestinian Arabic, /fiːl/ means elephant whereas /fil/ is a kind of cork. If someone tried to refer to said cork with the word [fil] instead of [fɪl], I'd probably think they meant /fiːl/ and were talking about an elephant.
Short /u/
This sound actually has several allophones. When it follows non-emphatic pharyngeals /ħ/ and /ʕ/ it can be fronted to something in the general vicinity of [ʏ], [ø] or [œ]. When it follows emphatics it is often much more open and can tend toward [ɒ]. Elsewhere it can take the shape of [ʊ] particularly in inflectional endings.
I do know where to find sources for most of this (except for some of the sandhi rules, which will require me to look around.) Unfortunately, however, most of the scholarly sources available to me concern colloquial Arabic and not explicitly MSA, and almost none of them actually say that speakers apply their colloquial phonotactics to their MSA utterances to some degree. This would not be such a problem if synthesis weren't prohibited so strictly on Wikipedia.
Furthermore, materials intended for MSA learners seem not to pass muster very often on Wiki. Otherwise, I suppose you could find sources online like this one and others. I believe Alif Baa by Brustad et al. contains some slightly more detailed information for the learner (unfortunately, I don't have it on hand at the moment.) Barring that, you could check out pages 7-9 the first section of vol. 1 of W. Wright's mammoth Arabic Grammar for a brief description (albeit quite imprecise) of the short vowels' allophony. Szfski (talk) 10:34, 21 November 2009 (UTC)

concerning pronunciation of ج[edit]

“Egyptians are wont to pronounce ﺝ as [ɡ] in all situations, even when speaking MSA, and this carries over even into official communications such as news broadcasts and government bulletins & it's considered prestigious.” --Mahmudmasri (talk) 09:29, 11 February 2010 (UTC)
What's the issue, exactly? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 19:17, 11 February 2010 (UTC)
This note /ɡ/ and /tʃ/ are used in colloquial dialects and loanwords. claims that [g] is exclusively colloquial. --Mahmudmasri (talk) 10:39, 12 April 2010 (UTC)

Egyptian Arabic[edit]

I linked the new Egyptian Arabic template here, {{IPA-arz}}. Those of you who know it might want to ensure that the articles are reasonably supported. — kwami (talk) 08:16, 28 October 2010 (UTC)

I hope that the charts aren't confusing or misleading anymore after these edits: [1]. --Mahmudmasri (talk) 01:45, 5 November 2010 (UTC)

I also suggest adding a rule to Arabic script transliterations:

  • Whenever there is an article dealing with Egyptian related subjects containing transliterated words, the letter g should be used, because [ɡ] is the standard pronunciation in Egypt for ج in all cases.

I am suggesting this rule because I often find false transliterations of Egyptian names and organizations which are correctly pronounced with [ɡ], not with [ʒ] or [d͡ʒ]. --Mahmudmasri (talk) 02:41, 5 November 2010 (UTC)

We can have footnotes to that effect. However, we want to be a bit careful here: this key is intended for the reader, not the editor, and shouldn't swamp the reader with what are for them irrelevant details. Instructions for the transcription of Egyptian Arabic really need to be at the MOS, though I don't think a few basic guidelines would hurt here if they're kept out of the way. — kwami (talk) 13:09, 5 November 2010 (UTC)

Recommendation: Split into Standard Arabic and Egyptian Arabic[edit]

Does all of this material about Egyptian Arabic need to be here? If we take the standpoint that French and Arpitan or Mandarin and Cantonese are different languages, then it just seems logical to split them. Egyptian Arabic may be the most widely known vernacular dialect but it's only applicable to topics related to one country - Egypt. I suggest we create a new page for the transcription of Egyptian Arabic and move all the relevant material over there. AlexanderKaras (talk) 14:16, 20 November 2010 (UTC)

Most of the stuff on Egyptian Arabic is way too detailed for this key anyway, and should be removed to the Eg. Ar. article. (For instance, we don't need the names of the letters spelled out, we can just link the letters themselves, and we shouldn't divide up phonemes according to orthography, but only per their allophones.) Once that clutter is gone, there isn't a whole lot of stuff specific to Egyptian.
Mandarin and Cantonese aren't even close: they have very different sound systems, so there wouldn't be much basis for a merger. With Arpitan, besides the fact that we hardly need a dedicated key, there is one primary difference: no-one who needs to look up French IPA transcriptions is going to be the slightest bit interested in Arpitan, and will only find it annoying. Egyptian, however, is the prestige colloquial dialect of Arabic, and many students of Arabic learn Egyptian. Therefore IMO the combination may be appropriate, esp. considering how little accommodation needs to be given. Instead of Arpitan, consider WP:IPA for Portuguese, where both European and Brazilian standards are given, as well as Galician. — kwami (talk) 18:52, 20 November 2010 (UTC)
Nevertheless, the notes are a bit excessive. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 20:34, 20 November 2010 (UTC)
Agreed. I'll start cutting stuff out; you might have some practical ideas there too. — kwami (talk) 22:26, 20 November 2010 (UTC)
Now that I think of it, MSA and Egyptian Arabic are probably close enough in phonology to justify keeping them together. One change I recommend: I see a lot of material about Maghrebi Arabic that it would probably be best to remove. That information belongs in the articles about their respective dialects. This key is concerned with the two major standards. - AlexanderKaras (talk) 01:31, 30 November 2010 (UTC)
I cut out nearly half. Could probably use some more pruning. — kwami (talk) 07:51, 30 November 2010 (UTC)
Right now [2] I don't think that we can remove more notes. --Mahmudmasri (talk) 16:52, 30 November 2010 (UTC)
Perhaps you could clarify the note on /r/. What does it mean that it's "pronounced in between [ɾ] and [r]" and how will this information affect transcription? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:14, 30 November 2010 (UTC)
In Egypt, the pronunciation of that alveolar consonant is a trill by some speaker and a tap or flap by other speakers. Usually when transcribing that Egyptian Arabic alveolar consonant, the flap/tap symbol is used [ɾ]. So, the point is, if an Egyptian word is to be transcribed, then [ɾ] would be used, instead of [r]. --Mahmudmasri (talk) 20:49, 30 November 2010 (UTC)
Is it actually a tap, or just a one-contact trill? I've found that when a trill is alleged to vary with a tap, it very often does not actually do so, but only varies in the number of contacts. — kwami (talk) 23:17, 30 November 2010 (UTC)
If it varies in Egypt, I think we should pick one and stick with it. r makes more sense to me (which would make the note in question unnecessary), but if we do use {IPA|ɾ}} then we should reword that note. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 01:53, 1 December 2010 (UTC)
So, what you Kwami just said explains what I was trying to simply explain. The Egyptian alveolar consonant is not exactly a trill nor exactly a tap. It is half-way between both. The same thing applies to its pharyngealized counterpart. However, some speakers actually pronounce both of the them as trilled, while others pronounce them as tapped, though the half-way trilled-tapped dominate. Only if there were a symbol representing exactly that Egyptian consonant, half-way between a trill and a tap! --Mahmudmasri (talk) 01:06, 2 December 2010 (UTC)
Okay, but this is getting in the realm of detailed phonetics. Should we use the same symbol for both the emphatic and non-emphatic rhotic?— Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 04:46, 2 December 2010 (UTC)
I think there's some confusion in terminology here. From what I know, a flap (not a tap) is the same thing as a one-contact trill (a tap is a very brief plosive) and either way there's no need to make such fine distinctions here. It's enough to say "in Egypt, /r/ may be [ɾ]" and get on with it. Furthermore I think that whatever symbol we pick for the Egyptian /r/ should also be used to transcribe its emphatic counterpart. As far as I know there is no other difference in pronunciation. - AlexanderKaras (talk) 09:39, 2 December 2010 (UTC)
It's not a matter of weak or strong, it's a matter of articulation. A flap/tap is an active motion of the tongue. A trill is a passive motion: the tongue vibrates in the airstream. Either may be strongly or weakly articulated. — kwami (talk) 20:34, 5 December 2010 (UTC)


One glaring issue: we need to explain where the stress goes.

The rules I learned go something like this: stress goes on the last heavy syllable of the word (ie. ending in a consonant or a long vowel) or barring that, the third-last syllable. These are for MSA; other dialects may differ. - AlexanderKaras (talk) 10:05, 2 December 2010 (UTC)

MSA indeed has no one correct pronunciation, especially for the stress. It is used differently across all the Arab League. In Egypt, for example, when we use MSA, we put stresses just as we put them in Egyptian Arabic, penultimately.
  • An example: the republic الجمهورية DIN 31635: al-ǧumhūriyyah in MSA, in Egypt: [æl ɡomhuːˈrejːæ]. Egyptian Arabic: [el ɡomhoˈrejːæ] (the stress is penultimate).
  • I hear people from other places, such as the Levant, pronounce that word in MSA like: [al ˈʒʊmhuːrijːa] (the stress is initial).
  • I hear people from places like Northwest Africa pronounce it in MSA like: [ɛl ʒumhuːrɪjˈjɛ] (the stress is final).
I don't know how people outside of the region are taught to pronounce MSA. Mostly, I feel surprised that there is a claim that there is only one correct pronunciation for MSA. --Mahmudmasri (talk) 15:59, 2 December 2010 (UTC)
Mostly, in Wikipedia articles, MSA transcriptions are needed for Islamic terms. I don't recommend using IPA for that purpose. Using the transliteration scheme DIN 31635 would be better, doing away with the stress and other realizations of pronunciations of MSA across the Arab League. --Mahmudmasri (talk) 16:13, 2 December 2010 (UTC)
So users would have to learn another transcription scheme that doesn't even indicate stress? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:43, 2 December 2010 (UTC)
No one has to learn anything at all. Just, there is no need to add stress marks in IPA. --Mahmudmasri (talk) 00:56, 3 December 2010 (UTC)
The stress rules outlined above are how MSA is taught to 2nd-language learners in Western universities. Papers are written on MSA stress and syllable weight. The fact that the various Arabic dialects/languages differ in stress assignment is something that we can cover in that article, but it would be a disservice to our readers to leave stress out altogether. — kwami (talk) 09:15, 3 December 2010 (UTC)
I agree. It would be more confusing to leave the stress out altogether than to use the rules for MSA and explain they may, like everything else about Arabic, vary regionally. Also, a minor issue with your transcriptions: the definite article begins with an epenthetic /ʔ/. - AlexanderKaras (talk) 02:13, 4 December 2010 (UTC)
  • Well, if the transcription doesn't at least reflect how the language is used, where the language is the primary official written language, then what's the benefit of using IPA from the first place?
  • What is with syllable weights? According to what I understood from AlexanderKaras, that word الجمهورية would only be stressed as /æl ɡomˈhuːrejːæ/ and I can't remember having heard any Arabic language speaker pronouncing that word with the second syllable stressed.
  • I have to repeat again, MSA has many correct pronunciations. The pronunciations I wrote in my previous comment are all correct and can't be claimed false.
  • Omitting stress marks is better than adding incorrectly placed stress marks. Thanks. --Mahmudmasri (talk) 03:28, 4 December 2010 (UTC)
According to the stress rule you deleted, الجمهورية would be /æl ɡomhuːˈrejːæ/, with stress on the penult. — kwami (talk) 20:33, 5 December 2010 (UTC)
Thanks for explaining, but what is a heavy syllable? What I understood from AlexanderKaras is that /ɡom.huː.ˈrej.jæ/ heavy syllable is /huː/, so that the pronunciation would be: /ɡom.ˈhuː.rej.jæ/, which is incorrect. --Mahmudmasri (talk) 21:14, 5 December 2010 (UTC)
What is "heavy" or "light" depends on the language, but in this case I believe it's anything that has a long vowel, diphthong, or ends in a consonant. ري would thus qualify whether you analyze it as /reː/, /rei̯/, or /rej/. (In some languages, like Urdu, a syllable would have to have both a long vowel/diphthong and a coda consonant, etc.) See syllable weight. — kwami (talk) 22:06, 5 December 2010 (UTC)
I understood from that page that all of the syllables in that word [ɡom.huː.ˈrej.jæ] (in the standard pronunciation of MSA in Egypt) are heavy, with the exception of the last syllable [jæ], which is light. So, the stress in the standard pronunciation of MSA in Egypt would be: On the heavy syllable, but if there were more than one heavy syllable, then the penult syllable would be the stressed syllable — not the antepenult. But what about that example, with the final stressed syllable: (two) republics جمهوريتان/جمهوريتين ǧumhūriyyatān/ǧumhūriyyatayn standard pron. of MSA in Egypt: [ɡom.huː.rej.jæ.ˈtæːn, ɡom.huː.rej.jæ.ˈtæjn]? So the final syllable is super-heavy (CVVC)? --Mahmudmasri (talk) 00:14, 6 December 2010 (UTC)
Yup. Arabic language#Stress. I'd forgotten the details, but I think that section is pretty much what I learned in school. — kwami (talk) 01:10, 6 December 2010 (UTC)
I have edited that section there because it suggested that the stress rules in Egypt are exclusive to Egyptian Arabic language only, rather than in MSA pronunciation in Egypt. All the MSA examples there apply to how MSA syllables are stressed in Egypt, with the exception of the Yemenite stress rules and this MSA example, MAK-ta-bun should be [mæk.ˈtæ.bon]. --Mahmudmasri (talk) 02:25, 6 December 2010 (UTC)

Minor issues[edit]

I've added a brief description of the stress rules in Arabic. Feel free to alter them/add footnotes or whatever you need to do.

There are a few outstanding issues I'd like to address here:

  • Arabic does not permit a syllable to begin with an empty onset. When a word "begins" with a vowel, in reality an epenthetic glottal stop is inserted before it. For example, a word like اسم "name" is actually pronounced /ʔism/. I see this problem a lot with the definite article of Arabic.
  • How should the <ة> tā' marbūṭah ending be treated? It's realized as either /t/ or /h/ depending on context, and can be silent in modern pronunciation. I don't know how French transcription deals with liaison but it's probably a good example to follow.
  • The nisba is transcribed variously as īy/īyah or iyy/iyyah. How should it be transcribed? I was under the impression it was pronounced /iːja/ in MSA at least, but I could be mistaken.

If anyone else has any questions, they can be discussed here. - AlexanderKaras (talk) 02:46, 4 December 2010 (UTC)

Can you explain the context of <ة>? When is it /t/, /h/ or silent? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 02:56, 4 December 2010 (UTC)
  • Allow me to explain a few things:
  • To AlexanderKaras, I didn't have to add [ʔ] because the glottal-stop is canceled if that word with its definitive article is to be pronounced after a word ending with a vowel. Example: to the republic إلى الجمهورية ʾila l-ǧumhūriyyah; standard MSA in Egypt: [ˈʔelæ lɡomhuːˈrejːæ]. This is false [ˈʔelæ ʔæl ɡomhuːˈrejːæ]. Adding that glottal-stop mark would seem like: empire US: /ˈʔɛmpɑɪɹ/. (edited comment) اسم doesn't actually have [ʔ]. For example, the MSA question, What's your name? (s. m.) ما اسمك؟ masmuk(a); standard pron. of MSA in Egypt: [ˈmæsmok(æ)]. Words which have initial glottal-stop [ʔ] have their spelling starting with إ with a hamzah underneath the ʾalif.  The ʾalif  of اسم is called hamzat waṣl, while the other ones with a hamzah underneath /ʔi-/ ʾi-  or above أ /ʔa-/ ʾa-  are called hamzat qaṭʿ.
  • ة is never pronounced as final /-h/. It is pronounced as /-a, -at/ only. In many transliterations, a final -h is added, so that the Arabic text would be 100% recovered.
  • It is /-at/ in the genitive state, example: the university's library مكتبة الجامعة maktabatu l-ǧāmiʿah standard MSA in Egypt: [mækˈtæbæto lɡæːˈmeʕæ].
  • The relative endings (nisbah) aren't pronounced /-iːj, -iːja/ -īy, -īyah. The correct pronunciations of them are: /-ij, -ijːa/ -iyy, -iyyah. --Mahmudmasri (talk) 03:45, 4 December 2010 (UTC)
Do you have a source? Because, like I said, I've seen it transcribed both ways. The end commonly seen in Arabic names (signifying a geographical origin, a form of nisbah) is not usually transliterated īy except in very pedantic usage. Bear in mind that Egyptian Arabic pronunciation is not the same. For example, I would transcribe that same phrase as [ˈmæktæbæt‿ ʊlˈdʒæːmɪʕæ].
(For the benefit of our non-Arabic-speaking friend, the word مكتبة "library" is pronounced [ˈmæktæbæ] in isolation; notice the linking). So the letter ة is pronounced /t/ when linking two words. Followed by a pause, it's silent (historically, it was [h], which explains the spelling). So like I said, it functions much like liaison in French if you know anything about that. - AlexanderKaras (talk) 10:44, 5 December 2010 (UTC)
  • We can't be sure 100% how exactly the language of Koran was pronounced historically and that's not the case here. We are dealing with how the language is being used today.
  • See that PDF [3], it has the transliteration of the relative state: DIN 31635 and others transliterate it that way (-iyya, closer to how it is pronounced), with the ambiguous (-īya) only restricted to ALA-LC. See also [4] (P. 181, number 25.4) and [5] (p. 60, X).
  • The masculine ending is always transliterated and mostly pronounced as /-iː/, even if it's not pronounced as /-iː/, it is transliterated as -ī. Normally, that ending is pronounced as /-i(ː)/. The /-ij/ pronunciation is rarely used (for the masculine form) and if used, mostly by clerics who stress too much on the classical pronunciation. (edited) Example: Egyptian, (s. m.) مصرى miṣrī; (s. f.) مصرية miṣriyyah. Standard MSA pron. in Egypt: [ˈmesˤri(ː)]; [mesˤˈrejːɑ, -æ]; Egyptian Arabic: [ˈmɑsˤɾi, ˈmɑs-]; [mɑsˤˈɾejːɑ, mɑsˈ-], respectively.
  • Yes, I know about liaison (linking), but for simplicity, it's not used.
  • A better transcription for what you typed above, [ˈmæktæbæt‿ ʊlˈdʒæːmɪʕæ], is: [ˈmæktæbætʊ‿lˈdʒæːmɪʕæ] ([ʊ] is the ḍammah of مكتبة), which would be pronounced in Modern Standard Arabic in Egypt as: [mækˈtæbæto‿lɡæːˈmeʕæ] (notice where are the stresses), while in Egyptian Arabic as: [mækˈtæbt] or [mækˈtæbet el ˈɡæmʕæ] (notice where are the stresses; no epenthetic [ʔ], in both MSA and E). In my previous comment, I wasn't transcribing the Egyptian Arabic pronunciation, I was transcribing the standard pronunciation of Modern Standard Arabic in Egypt. --Mahmudmasri (talk) 15:17, 5 December 2010 (UTC)
I suppose I might as well respond to you point by point; after all you did take the time to type up all that.
  • I'm aware we are transcribing how the language is pronounced in modern usage; that's why it's for MSA, not Qur'anic Arabic.
*Do you have a source on the pronunciation of the nisbah, preferably in IPA? I would just like to make sure it is correct.
  • I wasn't implying we use linking in transcription; it was only there for didactic purposes.
  • I know the glottal stop is not an inherent part of the article; it's just inserted because a word cannot begin with a vowel. In Arabic the difference between epenthetic and phonemic glottal stop is known as همزة الوصل hamzat ul-waṣl and همزة القطع hamzat ul-qaṭʿ, I believe.
  • I'm afraid I don't understand what is the point of transcribing MSA as pronounced in Egypt. Since this guide is meant for the broadest possible use, it necessarily has to gloss over some fine points and present a somewhat fictitious ideal, just as Wikipedia:IPA for English does. It can also be used for Egyptian Arabic transcription, but that is a different issue altogether. - AlexanderKaras (talk) 09:10, 6 December 2010 (UTC)
Do you have a source on the nisbah
Thanks for your time. Well, I can't find now sources for how the nisbah is pronounced. I was transcribing in MSA-Egypt just to show you that the pronunciation really differs from what I read in Wikipedia articles and what is claimed to be a uniform standard across all north Africa and west Asia. I have suggested before that MSA transcription should be broad, // as in English, also, I have already suggested that stress marks should be omitted, but my proposals were refused. Languages of the third world and especially of our region, are not well documented or studied. Moreover, the different languages/dialects of the region are politically (and/or religiously) treated as slang or accents of the (most eloquent) Koranic language, that's why I don't think we would succeed in finding a source demonstrating the realistic pronunciation of MSA thoroughly in extreme details. I feel very irritated that the transcription is made in a very biased and a different pronunciation of how MSA is pronounced in Egypt (also other regions) and I even feel more irritated when our Egyptian pronunciation of MSA is claimed to be unauthentic and deviant from the standard! The pronunciation of ج is pronounced as [ɡ/ʒ/d͡ʒ]. The affricate [d͡ʒ] isn't more common nor more correct than [ɡ/ʒ]. The same applies to ظ [zˤ/ðˤ], خ [x/χ] and غ [ɣ/ʁ]. None of those pronunciations could be claimed more standard or more correct over others. Vowels have different values across the region as well. Stressed syllables differ... --Mahmudmasri (talk) 16:48, 6 December 2010 (UTC)

Typical Error, Broad Transcription Can Fix[edit]

I want to demonstrate a typical Arabic transcription failure I see in Wikipedia articles. For example, jihad (ǧihād) article has جهاد transcribed phonetically [dʒiˈhæːd]. This pronunciation is totally fake and not even phonetically correct to any region of the Arab League. The region which uses the affricate [d͡ʒ] (Arabian Peninsula, Iraq, Algeria) they don't pronounce that near-open vowel [æ]. In west Asia, those accents which differentiate between a front and a back vowel (retraction), have these two vowels [a], [ɑ]. Those who don't, have only [a~ä]. In Egypt, which has the near-open vowel [æ] doesn't have the short [i] (initially or medially) nor the affricate [d͡ʒ]. Northwest Africans don't have [æ] nor [d͡ʒ], with the exception of Algeria which has [d͡ʒ] but still, not [æ]. Northwest Africans contrast between the retracted and non-retracted vowel: [ɛ~e] and [a]. The broad transcription of the word could be something like /ɡɪˈhaːd,  ʒɪˈhaːd,  dʒɪˈhaːd/. A phonetic transcription based on MSA-Egypt would be [ɡeˈhæːd]; MSA-Levant [ʒɪˈhaːd]; MSA-Arabian Peninsula and Iraq [d͡ʒiˈhäːd]; MSA-Algeria [d͡ʒiˈhɛːd], MSA-Tunisia [ʒɪˈhɛːd]. --Mahmudmasri (talk) 00:37, 7 December 2010 (UTC)

My problem with broad transcription is that it leaves out many phonetic details that will leave a non-Arabic speaker with an inaccurate view of how the language is pronounced. We are not trying to cover ever single dialect here; we are trying to show a (admittedly idealized) form of MSA pronunciation. This is the pronunciation taught in Western universities and is based on the rules for Qur'anic recitation.
I would like to ask you first, though: Where do you get all this material on dialects? - (talk) 05:16, 9 December 2010 (UTC)
You confuse accents of literary Arabic with spoken dialects/varieties of Arabic. A video of two Moroccans speaking literary Arabic with their Moroccan accent. This is a typical situation only found on pan-Arabist TV shows (Aljazeera is one). This is another link for a Moroccan singer speaking and singing in Moroccan Arabic. Do you feel any difference between both of the languages in both videoclips? If you failed to feel the difference, then it's a huge erroneous problem!
I would like to repeat what I wrote before, just for sake of emphasis: What is the usefulness of transcribing literary Arabic in a way different from what people in the Arab League pronounce? What if someone insisted that English language should be phonetically transcribed in an Egyptian accent? Would it really make sense to you? This is similar to what you were telling me on the transcription of literary Arabic as pronounced by westerners. The way literary Arabic is transcribed in Wikipedia isn't anyone's pronunciation in the whole Arab League. --Mahmudmasri (talk) 19:53, 10 December 2010 (UTC)
We typically choose the standard form of a language. That's true for French, Spanish, German, Mandarin, Hindi, etc. Why special consideration for Arabic? — kwami (talk) 20:29, 10 December 2010 (UTC)
The simple answer is, because there isn't a single phonetically correct standard pronunciation for literary Arabic. Also, the claimed standard used here isn't phonetically correct to any place across the whole Arab League, the main users of literary Arabic. --Mahmudmasri (talk) 00:07, 11 December 2010 (UTC)
Actually, what you're saying holds true for any language. Barely any native English speakers use Received Pronunciation on a daily basis, yet it's (a) standard. It's a useful abstraction of how people actually speak. Nobody is claiming all Arabs talk like this, all the time. As well, the standard pronunciation is emphatically not "how Westerners speak Arabic." It's merely the pronunciation that is taught to them. Again, a useful analogy is RP. - AlexanderKaras (talk) 23:00, 11 December 2010 (UTC)
English transcription is broad, representing diaphonemes. Rarely, if ever, English language speakers pronounce in RP, not even the British, even BBC-English is quite different from what RP phonetically represents, but no problem because the transcription didn't claim to be 100% phonetically accurate. The Arabic transcription used here is written between square brackets, claiming that this is a phonetically accurate pronunciation for a certain place at least, while it's not phonetically correct to any place on earth. At least continue transcribing the same way you are doing it, but not in between square brackets, because those pronunciations are fake. I have already written three different transcriptions for the word جهاد to be broadly transcribed, /ɡɪˈhaːd,  ʒɪˈhaːd,  dʒɪˈhaːd/. Writing a broad transcription is better than writing a false transcription. --Mahmudmasri (talk) 01:14, 12 December 2010 (UTC)
We've had this debate with other languages. The problem is that while English speakers speak English, they generally do not speak Arabic, and so can not be expected to process a phonemic transcription. — kwami (talk) 01:53, 12 December 2010 (UTC)
I'm interested to read those discussions and see their arguments. Would you please give me links? --Mahmudmasri (talk) 03:03, 12 December 2010 (UTC)
So far, I read discussions about German, Dutch, Catalan, Burmese. I understood from all of them that their problem was to cover more accents. But our problem here is that the phonetic transcription isn't anyone's accent! --Mahmudmasri (talk) 03:23, 12 December 2010 (UTC)
Despite what you might think, the usage of brackets does not prevent us from utilizing a diaphonemic IPA transcription system. This was/is done for Spanish, which has dialects with different phonologies but largely similar phonemic inventories. As with Spanish, it is possible to have a system that is both phonetic and diaphonemic.
With that said, it's my understanding that the variability of MSA pronunciations around the Arabic-speaking region is different enough and in such a way that we would be unable to form such a diaphonemic transcription as the discussion on stress assignment illustrates. I could be wrong, though. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 05:20, 12 December 2010 (UTC)
That doesn't mean we can't prescribe a standard. The pronunciation of MSA probably does not vary more regionally than that of English. Also, I would like to see sources for those transcriptions. - AlexanderKaras (talk) 06:04, 12 December 2010 (UTC)

The "O" sound[edit]

The "O" nearly always in Arabic transliterations has the sound in "got" not "more".

John Cengiz talk 14:08, 8 January 2011 (UTC)

Do you have a source? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 18:02, 8 January 2011 (UTC)
Mohamed? Mostafa? Omar? Osama? I can't think of any Arabic male names where the "O" has the sound in "more". John Cengiz talk 18:20, 8 January 2011 (UTC)
That's because we don't have [o] in English. But you haven't given us an Arabic word with [ɒ]. — kwami (talk) 21:47, 8 January 2011 (UTC)
I don't want examples, I want sources. Unless you're talking about English pronunciation of Arabic words/names. That would be irrelevant. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 22:00, 8 January 2011 (UTC)
Mohamed, Mostafa, Omar, and Osama all have the [ɒ] ("got") sound in Arabic and translit to English. Whether it's represented by an [o] or an [ɒ], the English and Arabic is the "O" in "got". Maybe it's the "nearest English equivalent" box that needs fixed but this is an elementary confusion on the table, that needs sorted. John Cengiz talk 22:28, 8 January 2011 (UTC)
It may indeed be an issue of English examples. Keep in mind that Mohamed has the the vowel of moe in the speech of many Americans and the vowel of got is often unrounded for Americans (making it not even an o sound at all). — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 22:38, 8 January 2011 (UTC)
When an American says "hot" it may sound like "hat", I think, but I don't have an American accent. When an American says "Moe" for the "Mo" in "Mohamed", they're wrong, in at least the correct pronunciation. The "nearest English equivalent" bit needs work, it is not accurate. You've got "soot" or "soon" for the "U"'s and "more" for the "O"'s. ??? See the IPA for English, it's got a "ɔː" with "thought". And the Arabic table doesn't even have this sound yet. John Cengiz talk 23:19, 8 January 2011 (UTC)
Okay, but what's a closer equivalent? Thought seems to be closest (especially if you're Australian). — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 23:25, 8 January 2011 (UTC)
Yeah, "dawn" or "thought" are more examples of the sound. John Cengiz talk 23:36, 8 January 2011 (UTC)
What does this have to do with Arabic? We're transcribing Arabic here, not English. — kwami (talk) 07:52, 9 January 2011 (UTC)

The Roman letter o here is used to transcribe [o] (Close-mid back rounded vowel), different from how it is used in the English orthography. Names of Arabic origins are pronounced differently by English language speakers.

(Mohamed, Mostafa, Omar, Osama) would be typically realized and pronounced in English as: /məˈhɑːməd, məˈstɑːfə, ˈoʊmɑːɹ, oʊˈsɑːmə/  or (Mohamed, Osama) /moʊˈhæməd, oʊˈsæmə/, with other possible pronunciations...

The previous names aren't pronounced that way by Arabic language speakers. The first three names are missing specific consonants not pronounced by English language speakers. --Mahmudmasri (talk) 12:54, 9 January 2011 (UTC)

Names of dictators[edit]

Maybe someone could add a transcription for Hosni Mubarak and Zine El Abidine Ben Ali? With how much their names have been in the news, it would be nice to know if the media is getting them right.— ʀoyoтϵ 04:21, 30 January 2011 (UTC)

I have added it for Hosni, but I don't know Tunisian Arabic pronunciation, so I can't add it. Ask a Tunisian. --Mahmudmasri (talk) 01:42, 3 February 2011 (UTC)

Unnecessary notes?[edit]

I am taking attention to reducing the number of footnotes in some of these IPA for X pages. These pronunciation keys are designed primarily for readers wanting to understand the language-specific IPA transcriptions they encounter in Wikipedia articles. We shouldn't swamp them with irrelevant details.

I would simply remove the notes and duplicate them here to allow other editors to find the appropriate article space for them, but I'm not exactly sure which ones are relevant. I'm not even sure if we are transcribing Arabic in a uniform manner across the IPA. What's the situation? 00:48, 29 December 2011 (UTC)

To Aeusoes1. Unfortunately, those are the least notes we can have for such a variety of pronunciation for Literary Arabic, yes! In is a way that you are undermining. The questions of the user underneath this is just an example for people who would always come and wonder why aren't we mentioning their regional pronunciation as well. --Mahmudmasri (talk) 23:57, 30 March 2012 (UTC)
But how are we transcribing Arabic at Wikipedia? Are we doing it in a regular way across Wikipedia? If not, what is the criteria for the differences? Is the process clear enough for editors who want to contribute?
I'll give you an example. "In Egyptian Arabic, some speakers lack certain emphatic consonants altogether." This is unhelpful for everyone: readers can make no use of it, editors are given no lead as to how to transcribe the names of Egyptian places or people, and anyone interested in regional phonology (a group we shouldn't be oriented towards, but seem to be anyway) gets nothing out of it. How is it that this note is necessary for the transcription of Arabic at Wikipedia? — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 02:45, 31 March 2012 (UTC)
  • I removed the pharyngealization note. I also edited the notes and summed together sounds which are never phonemic to native speakers, so it was unnecessary to split them.
  • The criteria for contributing is whether the user is going to transcribe a Literary Arabic term (as the Islamic jargon), he must use one of the Literary Arabic standards. Across Wikipedia articles, a broad transcription (especially for the vowels) is usually used for that, although some users like to use the east central Arabian pronunciation, I'm not sure what's the significance of this pronunciation to them since the Egyptian and the Levantine pronunciations are much more popular in the Literary Arabic media.
  • In the case of transcribing names of modern places and modern personalities, they must be transcribed according to the local variety of Arabic as they are normally and commonly pronounced. Attempting to use a constructed Modern Literary Arabic pronunciation in these cases makes the pronunciation fake and artificial, sometimes even funny.

--Mahmudmasri (talk) 01:50, 3 April 2012 (UTC)

Your answers and edits have me more concerned about this guideline.
  • What is the deal with the tildes? I understand what dʒ~ʒ means dialectally (some varieties pronounce one, others the other) but how does this manifest in our transcription scheme? This is what I mean by "criteria for the differences." If we really just let editors pick whichever Literary Arabic standard they see as best, which you seem to be implying, then it negates the whole point of having this guideline.
  • Grouping sounds together just because they are non-contrastive also misses the point. Non-technical readers aren't going to care whether a sound is phonemic or not; they want a clear guide on how to pronounce words in Arabic. This is why [b] and [β] are separate entries at WP:IPA for Spanish, even though there is no contrastive distinction.
  • With English placenames, "local" pronunciations don't link to WP:IPA for English since the English guideline doesn't help readers with the kinds of transcriptions found with local pronunciations. Those local place names also typically are a supplement to the more general English pronunciation because even Englishmen, for example, talk about New York. This is why I disagree with the notion modern people and places should be transcribed in more local Arabic varieties. Putting the pronunciation in local varieties is fine, but I don't think this guideline should try to encompass those varieties. Even if we don't link to here in these "local" transcriptions, we need to be very clear about what we mean by "local." Is it in a particular city? Region? Nation? — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 02:39, 4 April 2012 (UTC)
  • I'll revise again the table.  "Local" here means, for example, Egyptian Arabic, Lebanese Arabic... The Egyptian Arabic and the Lebanese Arabic aren't only spoken in Cairo or Beirut, they are spoken by the majority of their respective regions (Egypt — Lebanon and parts of Syria). There's something you can't figure out regarding the case of Arabic speakers when they pronounce names. Even if they were in a situation when they have to speak in Literary Arabic, they always pronounce names in their regional languages. This means that this name which is of the Arabic origin, محمد mḥmmd is pronounced by Egyptians as [mæˈħæmmæd], not [moˈħæmmæd, mʊˈħæmmæd, muˈħammad, mʊˈʜämmɐd, mħɛmmɛd]. This is different from the case of English speakers, in situations when they should speak in Literary English, they pronounce names in standard English of their region (ex: General American/Received pronunciation), not in their local accents of the cities where they live.
  • The case of Spanish is also different, because the phoneme /b/ is pronounced by the same Spanish speakers as [b] in some cases and in other cases as [β]. The case in Arabic is that they either use one pronunciation of these only, all the time, [d͡ʒ], [ʒ], [ɡ], [ɟ]  ([ʒ] or [ɡ] are the most common), so grouping them reduces unnecessary distractions for someone who wants to quickly understand what do the symbols mean. It also makes him understand that using one of the sounds is correct and dialectal. You may wonder now why didn't I group [ɡ] with [d͡ʒ~ʒ], it's because the affricate and the fricative are realized by Arabic speakers as the same sound, they don't feel a difference, even in loanwords, but they feel a difference between [ɡ] and [d͡ʒ~ʒ] which are also used in loanwords. So, /ɡ/ and /d͡ʒ~ʒ/ are phonemic in loanwords for all Arabic speakers, while [d͡ʒ] and [ʒ] aren't phonemic in all cases. For example (I won't indicate the stress and the vowel length in the following examples) Michael Jackson is either pronounced /majkel ʒakson/ or /majkel d͡ʒakson/, but never /majkel ɡakson/ - Jaguar /ʒagwar/ or /d͡ʒagwar/, never /(d͡)ʒa(d͡)ʒwar/ or /ɡagwar/.
  • The case of Literary Arabic language isn't Wikipedia's problem. It's a real issue with Literary Arabic itself. Its pronunciation/stress vary widely from region to region. Each region has what its people consider standard.

--Mahmudmasri (talk) 00:48, 5 April 2012 (UTC)

Hmm, it gets quite messy then. Would we be able to do something like at WP:IPA for Portuguese where we correlate sounds between varieties? If so, how many columns would we need? — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 02:01, 5 April 2012 (UTC)
I thought of the same idea, but I see it impractical, because the difference in pronunciation is unrelated to sovereign borders, but rather regions, regardless of which sovereign states have parts in that region.

--Mahmudmasri (talk) 17:32, 7 April 2012 (UTC)

A similar problem exists with WP:IPA for Portuguese. The "Brazilian" variety is one of many and the European variety is basically the standard. Think of this as picking a representative variety of these broader "national" dialects. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 22:39, 7 April 2012 (UTC)
I still see the situation of Arabic language to be different and therefore it would not be acceptable by Arabic speakers, but however, because you insisted, I'd say, let's have the following four flags of: Egypt, Lebanon, Tunisia, Saudi Arabia.

--Mahmudmasri (talk) 23:49, 9 April 2012 (UTC)

ط is not pronounced [tˤ][edit]

As an Arabic speaker and linguist, I was interested to see what the IPA phonology for Arabic was - I was particularly shocked when I saw that, to name one of the many errors, ط was described as being a pharyngealised [t]. This is definitely not correct - in Iraq, the ط is normally pronounced as something resembling a voiced dental plosive with the middle of the tongue touching the top two incisors, rather than the tip, and there is definitely no constriction of the pharynx during speech. A few other Arabic letters: ظ ([ðˤ] and also, bizarrely written as [zˤ]) and ض ([dˤ]) have been written as pharyngealised, even though, through my experience as a native Arabic speaker, they are not. What are your views on this issue? I would be interested to hear from some North-African or Peninsular Arabic speakers!

J. Al-Khalili (talk) 18:11, 20 March 2012 (UTC)

Dear Khalili. Have you noticed that this table is for the pronunciation of a very wide region? In this wide region the pronunciation varies greatly, not to mention the non-Literary Arabic pronunciation. In Egypt, Sudan and the west Levant, we do pronounce the consonants you mentioned the same way you questioned and this is the norm in Literary Arabic. The plosives are alveolar, not dental for these regions; ظ is pharyngealized voiced alveolar fricative. In the Arabic phonology article, it is stated in details that there are differences. --Mahmudmasri (talk) 23:51, 30 March 2012 (UTC)

Dear Mahmudasri,

Thank you for your insightful contribution to my question. I am pleased, to say the least, that I have been proven wrong, especially by somebody, who thankfully, has a fluent grasp on the dialect of Arabic spoken in places other than Iraq. In Syria, I know for a fact that the consonants are not pharyngealised, as in Egypt and Sudan. May I ask from where in the Middle East do you come?

All the best (talk) 10:34, 11 April 2012 (UTC)

Hi again Khalili, as it's apparent, I'm from Egypt. Regarding the Syrian pronunciation, I meant, the west Syrian pronunciation, because the east shares similarities in pronunciation with the Iraqi pronunciation. Even though the pronunciation of the most of the Arabian Peninsula including Iraq has the emphatics as velarized [ ˠ ], there is a common practice to transcribe the emphatics with the pharyngealized mark [ ˤ ] which is the more common pronunciation elsewhere. --Mahmudmasri (talk) 20:20, 14 April 2012 (UTC)

Hi Mahmudmasri, Thank you for replying so promptly. There is also the issue of ط not being an alveolar plosive at all. It is, to say the least, dental, but the tip of the tongue extends quite a way out of the mouth. I was wondering if it would be possible to create a whole new IPA symbol to account for this apparent difference, because, I think that there should be different IPA symbols for the various Arabic letters depending on where in the Middle East it is spoken. What do you think? (talk) 17:38, 9 May 2012 (UTC)

[t̪ˠ] that's how I understand it's pronounced in the Persian Gulf: velarized voiceless dental plosive, which is written with ⟨ط⟩. But, as you described it as a voiced dental plosive, it would be [d̪ˠ] which I think you meant to say voiceless instead of voiced, because this is what is written with ⟨ض⟩; ‎ ⟨ظ⟩ is [ðˠ], and so on... --Mahmudmasri (talk) 06:04, 10 May 2012 (UTC)

No, I meant voiced, because it is voiced, and I think it is [d̪]. (talk) 18:59, 15 May 2012 (UTC)

Then, how is ⟨ض⟩ pronounced there, colloquially and in Literary Arabic? Does that mean that both ⟨ض⟩ and ⟨ط⟩ are pronounced the same [d̪ˠ]? --Mahmudmasri (talk) 07:47, 16 May 2012 (UTC)

الإطباق (al-itbaq)[edit]

(ص),(ط)and(ظ)with (ɫ) Contain (itbaq)

And is different from the (س), the (ذ), and the (ت, or د)

Even after the (tafkheem (ˤ mark))

As was supposed to use is following the mark (ˠ)

Such as the (ل) in (الله)

أبو السعد 22 (talk) 10:38, 30 April 2012 (UTC)

I don't understand what you are trying to say. Can you be more explicit? — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 11:33, 30 April 2012 (UTC)

(ص)sad not (sˤ) Completely , (ط)taa also not (tˤ) and (ظ) not (ðˤ)

What is the mark of (itbaq)?

Such as that on the (ل)lam

here word of (الله) in IPA (alˠːaːh)

example for (itbaq): ص/̴sˤ/, /̴tˤ/ط, /̴ðˤ/ظ /ɫˤ/ل

The intended it's ص/sˤˤ/, ط/tˤˤ/, ظ/ðˤˤ/ /lˤˤ/ل in (الله) and (اللهم) only

who are you from? — Preceding unsigned comment added by أبو السعد 22 (talkcontribs) 14:40, 30 April 2012 (UTC)

You seem to be having difficulty with English. I recommend that you participate in the Arabic language Wikipedia. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 16:11, 30 April 2012 (UTC)

but some times only, Because I use Google Translator with knowledge in EnglishBecause I use Google Translator with knowledge in English.

Also What is the mark of (itbaq)?

أبو السعد 22 (talk) 19:31, 30 April 2012 (UTC)

Sorry, we can't understand you. Google Translate is not good enough! — kwami (talk) 12:03, 1 May 2012 (UTC)

Who are worders from?

 — Preceding unsigned comment added by أبو السعد 22 (talkcontribs) 21:41, 1 May 2012 (UTC) 

ڪ missing from the list.[edit]

Hello, the letter ڪ is missing from the list. — Preceding unsigned comment added by (talk) 00:28, 20 July 2012 (UTC)

Is it used when writing Arabic? — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 04:12, 20 July 2012 (UTC)
It appears that ڪ is the initial and medial form of the letter ك (Kaaf) [Ref 1]. My book [Ref 2] says that Kaaf is always written like ڪ; the shape has probably changed in the last 150 years.

[1] [2] (talk) 00:08, 21 July 2012 (UTC)

So it's just a positional variant of a letter we already mention. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 02:27, 21 July 2012 (UTC)
In Arabic language, especially in the Quranic texts, it's just a different style for the ك and is treated exactly as ك. --Mahmudmasri (talk) 22:04, 22 July 2012 (UTC)

Move discussion in progress[edit]

There is a move discussion in progress on Wikipedia talk:IPA which affects this page. Please participate on that page and not in this talk page section. Thank you. —RMCD bot 00:56, 20 August 2012 (UTC)

Spacing and diacritics[edit]

Regarding Lfdder stubborn revert. Normal spaces are too thick for what is needed to separate brackets and tildes from diacritics and letters. --Mahmudmasri (talk) 08:43, 17 August 2013 (UTC)

Like I said, these don't display. Try using &thinsp;. — Lfdder (talk) 08:46, 17 August 2013 (UTC)
How come they don't display? They appear on all computers, browsers and mobile phones I tested. The space I used (&#x202f;) is a non-break thin space, what you want me to use is a thin break space. --Mahmudmasri (talk) 11:25, 17 August 2013 (UTC)
Not seeing it in either Firefox or Chrome on OS X nor on my phone. Renders properly elsewhere so it might have something to do with tag placement. — Lfdder (talk) 12:14, 17 August 2013 (UTC)
Frankly, dʒ ~ ʒ, ðˤ ~ zˤ, etc. should be listed separately. This is an IPA key; not a phoneme table. — Lfdder (talk) 12:16, 17 August 2013 (UTC)
For practicality, they shouldn't be separated. Those who are familiar with Arabic would understand why. --Mahmudmasri (talk) 07:52, 18 August 2013 (UTC)
The letter and transcription would span 2 rows. — Lfdder (talk) 07:56, 18 August 2013 (UTC)
If you mean like how f ɸ or j β appear in Help:IPA for Turkish, Azerbaijani and Turkmen, then it is still impractical, because there are no simple national boundaries for the use of the consonants as that highly simple case. --Mahmudmasri (talk) 08:06, 18 August 2013 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────No, vertically, like for r here. — Lfdder (talk) 08:08, 18 August 2013 (UTC)

The numbered notes would still have to be next to the consonants, or if you don't want them sticking to the consonants, then you would have to put the bracketed numbers in another column next to the consonants. --Mahmudmasri (talk) 08:20, 18 August 2013 (UTC)

Short vowels: [ɪ ʊ] (reliable sources for Egyptian Arabic) vs. [e o] (Mahmudmasri's claim)[edit]

OK, I tried to deal with this years ago, but I see that Mahmudmasri has far more patience than me. I have not seen one single linguistic source that claims that the short allophones of /i u/ are pronounced [e o] in Egyptian Arabic rather than [ɪ ʊ]. Mahmudmasri swears up and down that [e o] are correct and that he knows this as a native speaker, but all the linguistic sources I've seen use [ɪ ʊ], and the clips I've heard have vowels that sound more like [ɪ ʊ] except that [e] appears before [j] (possibly likewise for [o] before [w]). The page Egyptian Arabic phonology has no sources whatoever backing up the assertions of the sounds [e o], which are entirely due to Mahmudmasri. I believe that Mahmudmasri's claims are OR unless he can find sources backing up his assertions. He claims that all sources asserting [ɪ ʊ] are "wrong" or "artificial" but I think it's more likely that he simply doesn't quite know what canonical [e o] sound like. Benwing (talk) 07:22, 16 October 2013 (UTC)

An already old discussion. Also see Talk:Egyptian Arabic/Archive 5#transcription. You chose a bad title, it's not me against reliable sources. You could have just asked more about what was already discussed a long time ago. Your sources were clearly not that reliable since they contained lots of outdated claims not entirely reflecting contemporary Egyptian Arabic vocabulary. It's like citing information about contemporary English language from something speaking about early Modern English. By the way, I didn't swear :) It's not my fault that you can't find sources about contemporary Egyptian Arabic, however, I'm aware that non-western languages are much less studied. It's a pity. What to do?
To cut it short for those who have no time to read the archives. Short [ɪ ʊ] exist for contemporary metropolitan Egyptian Arabic pronunciation, however, that's the case when they are pronounced in syllables near other syllables containing [iː] or [uː]. Otherwise, they are [e] or [o].   [i u, ɪ ʊ] can be pronounced by careful speakers in loanwords. There are some exceptions that nativized words are always pronounced with [i u] that should have otherwise be pronounced as [e~ɪ, o~ʊ]. --Mahmudmasri (talk) 05:21, 16 November 2013 (UTC)

Hejazi mid vowels[edit]

@عربي-٣١: [ʕrbi-31]; It's just that this picture claims that the vowels are close-mid. --Mahmudmasri (talk) 23:29, 26 July 2015 (UTC)

Literary Arabic vowels[edit]

@عربي-٣١: [ʕrbi-31]; The mid vowels /eː, oː/ do not exist in Literary Arabic words and are not converted to /aj, aw/ in loanwords within Literary Arabic. ـيه is a standard transcription and a normal practice for writing final // in at least Egypt. English approximated sounds are just approximations. Don't remove them. --Mahmudmasri (talk) 02:24, 16 August 2015 (UTC)

well the first part you were talking about i agree with, i thought it was the phonemes that corresponds from modern day dialects to Standard Arabic, but if we were talking only about Standard Arabic then the corresponding phoneme for ظ is only /ðˤ/ and for ج is only /dʒ/ even modern-day educated egyptians nowadays pronounce it that way.even the use of /æ/ and /ɑ/ instead of /a/ for modern Standard Arabic is wrong since it mainly has one phoneme for that vowel and sounds highly influenced by the Egyptian Arabic pronunciation. The English approximated sounds don't mean any sense since the pronunciation of the emphatic ض ص ط ظ is not the same as pronouncing the plain د ذ ت س at all and it causes a confusion for english speakers, at least write hard d, hard th (as in the) etc.... --عربي-٣١ (talk) 09:59, 18 August 2015 (UTC)
Up till this moment in 2015, we normally pronounce ج as [ɡ] in all cases, unless reciting Koran or pronouncing Islamic related terms.   ظ is normally pronounced [zˤ]. The (near-)open vowels are actually two, in all cases. The pharyngealized consonants are sometimes pronounced with no pharyngealization, in Literary Arabic, by many speakers. The difference is usually distinguished by the open vowel [ɑ]. --Mahmudmasri (talk) 11:51, 23 August 2015 (UTC)


As I understand help pages are supposed to help. But this page in its current version is very overloaded with unnecessary minute details and hardly can help. To use this page one has to know Arabic dialectology thoroughly. In most cases at least two variants are given, thus every time one has to bother one's head with the unnecessary choice what variant to prefer. Moreover, the page clearly states that it is about MSA not about dialects. As I agree that there may be local variants in pronunciation of MSA (like /dʒ/ vs /g/ vs /ʒ/ for jim), better to represent only such important variations. The transcription must be phonemic not allophonic. The bulk of the page belongs to the articles about Arabic phonology and the dialects, not here. Not to mention that the page lacks any sources.--Lüboslóv Yęzýkin (talk) 10:48, 24 November 2015 (UTC)

You can use the first variant for the consonants, always, but the vowels, unfortunately you should have some knowledge about how the language is used, because there are no simple rules to mention, however there are helpful footnotes.
Literary Arabic phonology isn't simple and there is no preferred standard and there are sounds that are considered foreign for one place but not for the other. It is the problem of the language, not the help page. Literary Arabic isn't based on any spoken language. This is how Literary Arabic is used in the Middle East. The page indeed ignored to put in the table the velarized variants of the emphatics. The only symbols that I believe have to be removed, are [rˤ, ä, e̞, o̞, äː, e̞ː, o̞ː], because they are extremely detailed and I don't believe that [rˤ] is used in Literary Arabic pronunciation anywhere. We can reduce the page to the Egyptian pronunciation, but it won't be satisfying, would it? The problem with Literary Arabic is that it is shared by too many peoples with many hugely different local vernaculars, even more than the case for Standard Mandarin! --Mahmudmasri (talk) 00:20, 26 November 2015 (UTC)
Probably, we should clear the definition. There is a standard, it doesn't matter how strictly it is followed (at least in Arabic language courses in any university around the world it's used definitely). Its code is not <ara> but <arb>. These are two different things. As well as <zho> and <cmn>. We do not use every Sinitic topolect (called traditionally "dialects") in "IPA for Chinese", so it is not expected to use every Arabic topolect (again "dialects") here. --Lüboslóv Yęzýkin (talk) 10:52, 26 November 2015 (UTC)
When and how did the help page suggest to use vernaculars? Speaking of standards, British and American English have pronunciation differences, yet both are standards and sticking to one form won't label one as deviating from the standard! --Mahmudmasri (talk) 20:04, 26 November 2015 (UTC)

@عربي-٣١: your edits about the great details of vowel values are unnecessary and if you review other IPA help pages, you'll notice that the letter "a" is used for any vowel at the front, all the way near the center. The letters "e, o" are also used for any vowel from the close-mid to the middle. It's almost a policy to simplify and that's how usually it is done in Wikipedia, unless the language has no IPA help page. Your additional detailed vowels are likely to be removed by other users. And regarding the narrow versus broad transcription, I raised the issue before here, but no one cared or agreed, but the transcription is mostly used for the narrow pronunciation. --Mahmudmasri (talk) 00:20, 26 November 2015 (UTC)

@Mahmudmasri: You, as a native speaker, may know the situation with MSA much better than me, but from my point of view there is at least a desirable standard pronunciation, akin to RP in English. I cannot confirm how strictly this standard is followed at least by Arabic TV news channels (where one is expected to hear MSA), as my comprehension of spoken Arabic is very shallow not to say close to zero (frankly, I do not need this skill in my life at all). But this standard is taught to every learner of MSA and this standard is described in many sources. If I remember right, any Wikipedia article including help pages should follow some sources, thus anybody could confirm oneself the written text. A personal opinion of a speaker, even who knows the language above average could not be seen as reliable. Especially, in such a delicate field as phonetics, where different opinions are always expected. Plainly put, I and everybody cannot confirm either Arabic speakers who wrote this page are wrong or right.

I never denied that there is a standard form which is taught to all of us. I agree and it's Wikipedia's policy that any claims written should be verifiable. However, where are the studies which detail Arabic pronunciation? They are very few and sometimes imprecise. They are put in the Arabic phonology article, however. --Mahmudmasri (talk) 20:08, 26 November 2015 (UTC)

The transcription by Thelwall & Sa'adeddin used in The Handbook of the IPA is not ideal, especially in the vowel part, so I do not propose use that here (though we could possibly adjust it). But I know other sources, where one could see transcription of MSA. For example, in Brill's Encyclopedia of the Arabic Language, the article "Phonetics" (with subjoined references). It uses some allophonic variants for vowels, so it is not barely simplistic. But the allophones there are not too numerous, so anybody with basic skills in phonetic transcription could apply them here.

Following those references would be helpful, but which standard did they stick to, a Persian Gulf, a Levantine or an Egyptian or what else? --Mahmudmasri (talk) 20:13, 26 November 2015 (UTC)

I understand that Arabic is a champion of pluricentric languages, there is not a prestige pronunciation which everybody aims to follow (as Parisian in French or Pekingese in Mandarin). I understand that Arabic native speakers when trying to speak MSA use their native accent more or less: some speak with a heavy accent, some with a more neutral. But nevertheless, there is nothing unique here. Such a language with unstable pronunciation exists, this is English. The two supposedly standard pronunciations, RP and General American, are as similar as MSA in that respect. Everybody speaks English with an accent, there are hundreds of accents, thousands of allophonic variants, but nobody transcribes them all, except for in scientific publications. Every dictionary, textbook, etc. use either RP or GenAm. Here in Wikipedia, some mix of the two is worked out. So we have dealt with English transcription somehow.

Quoting you, "some with a more neutral", that applies to one's society, for example, speaking Literary Arabic with a neutral Egyptian accent to Egyptians, which won't be neutral to a Saudi Arabian. --Mahmudmasri (talk) 20:11, 26 November 2015 (UTC)

So I propose to apply such a principle to MSA (and I believe this should be done with every "Help:IPA" page). There are many sources, we use them, one or another. Those who wants to describe Arabic dialects, just create "Help:IPA for <dialect> Arabic". Everybody's happy.

Yet, there are endless situations where you want to use a notation for a dialectal word, but can't create a separate help page. The most prominent case is when transcribing names. No Arab pronounces any Arabic name in Literary Arabic phonology. Don't you see that this is a special case not given in your examples? مرزوق Marzūq (commonly spelled, Marzouk) is normally pronounced [mæɾˈzuːʔ] when it should be [mɑɾˈzuːq] in Literary Arabic, even if you were speaking a whole speech in Literary Arabic, you would still pronounce the names in the national dialect. --Mahmudmasri (talk) 20:20, 26 November 2015 (UTC)

I'm not rushing the issue, and I'm not ready to participate myself, but someone somewhen should be bold and clear up this page for good. --Lüboslóv Yęzýkin (talk) 10:42, 26 November 2015 (UTC)

I agree with Luboslov. This "help" page is not helpful at all, and certainly not reader-friendly for our casual readers. This page should just give the standardized (i.e. the "artificial", non-region specific) pronunciation of Modern Standard Arabic. At least four new pages (and accompanying templates for use with IPA in the respective articles) should be made: Help:IPA for Egyptian Arabic (currently a redirect back to this page), Help:IPA for Maghrebi Arabic, Help:IPA for Levantine Arabic and Help:IPA for Peninsular Arabic to correspond with all the major dialect families given in Varieties of Arabic#Regional varieties. The sub-regional differences that would be need to be noted in those would be much more manageable, cleaner on the page, and approachable by our readers.--William Thweatt TalkContribs 11:05, 26 November 2015 (UTC)
It is worthy to read my responses to Lüboslóv within his sentences, above. --Mahmudmasri (talk) 20:20, 26 November 2015 (UTC)

@Mahmudmasri:: I do not much like point-by-point rebuttals, and the wiki engine is not much comfortable for this, so I'll again answer in one piece. For now the page covers the vernaculars. For example, such allophonic variants as /ħ ~ ʜ, r ~ ɾ, x ~ χ/, etc. are clearly linked to and explained by their presence in the vernaculars. I do not much know about "labels", but I am sure that pronuncing /x/ insted of /χ/ and vice versa does not "label" anybody. I'm sure hardly any speaker notices such minute and purely allophonic variations. We do not list all regional allophones of /r/ in practical transcription of English.

2nd. I gave an example from Brill's Encyclopedia. I think it is quite a reliable and authoritative source. This page is not called "Arabic phonology" where we can and should provide many details and many sources. This page purely for practical purposes: an editor could transcribe Arabic words into the IPA, and the reader could understand what these IPA "odd" symbols mean.

3rd, 4th and so on. The question "which standard" is pointless, because there are not standards but one standard. The standard that is used in sources (I gave some). It does not matter how a particular speaker realizes Arabic words in his speech. Nor it matters how different realizations are perceived. It is a problem of applied phonetics, dialectology and sociolinguistics, but not of abstract and neutral practical phonemic transcription. Speakers of various English dialects can pronounce John in various ways, but we transcribe the name /dʒɒn/ or /dʒɑn/ (for the US). Thankfully there is not two variants of MSA, but one, so we simply transcribe Marzūq /mærˈzuːq/. I do not know where you got your variant, but it's not sourced. And that is the problem. The plethora of variants is not sourced. These are just personal opinions of speakers of what they think how they pronounce words. The "sourcing" like "I am a native speaker, I think I pronounce words thus and thus, hence I transcribe thus and thus" is not acceptable and verifiable.--Lüboslóv Yęzýkin (talk) 12:38, 27 November 2015 (UTC)

The examples you are providing are incomparable, because the name John is actually pronounced as such and the help page for English transcribes phonemically, for example, the rhotic is just /r/. I am simply not giving personal opinions, you are trying to mandate things not aiming at reaching solutions. Well, ask yourself, why would a Saudi Arabian spell his name as Marzoug/Marzoog? The simple answer is, they pronounce the same name differently [mærˈzuːɡ] and so spell it differently. Finally, no, there are many standards of Literary Arabic and more different than the British and the American are. Your claim demonstrates a very limited knowledge of Literary Arabic, and briefly: مكتبة is standardly pronounced in Literary Arabic in Egypt [mækˈtæbæ], but [ˈmaktaba] in the Levant. The sound the letter ح represents, is pronounced by the Persian Gulf standard as [ʜ] and that's not vernacular pronunciation. The same applies to خ: [χ]. Both are standardly pronounced in Egypt [ħ, x]. The Persian Gulf standard of ج is [d͡ʒ~ʒ] in Literary Arabic, but in their vernaculars the same words can also have [j]. The help page didn't mix these allophones because it is focused on Literary Arabic. The front plosives /t, d/ are alveolar in Egypt, but dental in the Persian Gulf. Those were just examples, not all differences...
  • To make the discussion fruitful, do you suggest removing most or all the notes on the consonants and vowels? But, you'll have to deal with people who come every now and then and mandate: they are dental not alveolar; they are uvulars not velars; the emphatics are velarized not pharyngealized...
  • How can we transcribe names which are the most transcribed words? They definitely need the help page to have minimum information about their sounds that many not exist in Literary Arabic. For example, the male name حميدة [ħeˈmeːdæ] couldn't be transcribed if [eː], which isn't Literary Arabic, is removed.
  • You may also need to review that Wikipedia:Ignore all rules, because you made me feel that you are obsessed with having to cite each and every utterance, knowing that Literary Arabic isn't as much studied in details as western languages are.
--Mahmudmasri (talk) 19:32, 27 November 2015 (UTC)

@Mahmudmasri: Modern Standard/Classical Arabic IPA should only be presented phonemically not phonetically, writing عربية [ʕɑrɑˈbɪjjɐ] phonetically sounds wrong to my ears but it's the Egyptian prounciation it should be written only phonemically /ʕaraˈbijja/ if you were speaking MSA or Classical Arabic in Saudi Arabia and pronunced عربية as [ʕɑrɑˈbɪjjɐ] they might even correct your pronunciation! so Arabic iPA should only be presented phonemically and stop changing it to phonetically based on Egyptian Arabic since not all Arabs are Egypt. Arabs originally came from what is now Saudi Arabia and Yemen and none of these countries have the realization of /a/ as [ɑ] in عربية, plus the example مية [ˈmej.jæ] is just wrong, since it's dialectal and not MSA or Classical. Maybe we should Start making Dialectal IPAs for regional words that don't exist in MSA. since our realization of phonemes and phones is clearly different from a country to another. Many Egyptians pronounce ثقافة as [sɑcɑːfa] does that make it phonetically right to be pronounced like that in MSA? only phonemic transcription should be used! and you don't have to control the page and base it on Egyptian pronunciation. that's not what Wikipedia is about. The transcription must be phonemic not allophonic. -- عربي-٣١ (talk) 10:25, 29 November 2015 (UTC)

Again, I didn't create the policy to transcribe Literary Arabic phonetically, and [ʕɑrɑˈbɪjjɐ] was the Persian Gulf pronunciation and is also cited in some sources. The Egyptian pronunciation is [ʕɑɾɑˈbejjɑ]; ثقافة is [sɑˈqɑːfɑ] vernacularly or literarily [θɑˈqɑːfɑ]. --Mahmudmasri (talk) 21:32, 29 November 2015 (UTC)
The word مية was an example of dialectal words because names are pronounced in dialects only and it was an example for that situation, although it's not a name. --Mahmudmasri (talk) 21:40, 29 November 2015 (UTC)


  • The problem, and I repeat it the third or fourth time, is a lack of sources. What you have described about regional variations of ح and خ in MSA (emphasis) is not verifiable. I could only believe personally you and your perception and description. I (and others) cannot check it myself. You may call any concerns with WP:V rule an obsession, but it is the norm to ask for sources. I'm not doing something extraordinary. Phonetics is not worse in that respect that one can ignore the sourcing of phonetic descriptions. It is not general knowledge which does not require sourcing. If you have a third-party source which clearly states that reporters in official TV news channels (the foremost place where everybody can hear MSA) in Saudi Arabia pronounce ح = [ʜ] and خ = [χ], while in Egypt these are [ħ] and [x] respectively, and this distinction is not allophonic and is important, hence we could add it freely. The "ignore rules" rule is rather about when rules prevent improving articles, but this is not a case.
  • Side note: I've frankly admitted I do not understand spoken Arabic. But it does not concern me, I never wanted to speak Arabic, but I'm always very eager to learn everything I'm interested in. My status of a non-native is even a plus: I'm not preloaded with any agenda or dialect and I can see things from a neutral angle. I rely on sources not my personal experience or perception.
  • I do not know your background and do not want to underestimate anybody, but you seem not to clearly differentiate phonetics and phonology. Your concerns about whether we should define and transcribe stops /t, d/ as dental or alveolar is out of scope of this practical transcription. If you look around other "Help:IPA" pages you will see that they are not overloaded with minute allophonic details as this page.
  • William Thweatt above proposed a way out and I agree with him: this page must concisely and economically transcribe MSA per given sources, but it does not mean we ignore other geographic variants, we create help pages for them. Now the page does not fulfil either of its goals: it does not represent concisely Standard Arabic and does not cover the varieties in full.
  • As for names: we can freely write {{IPA-ar|/maˈruːq/}} and {{IPA-ar|/ħaˈmaida/}} in locally indefinite situations, but {{IPA-ar|/maˈruːq/}}, {{IPA-afb|[maˈruːɡ]}} and {{IPA-ar|/ħaˈmaida/}}, {{IPA-arz|[ħeˈmeːdæ]}} in articles where the locale is known.

--Lüboslóv Yęzýkin (talk) 11:20, 30 November 2015 (UTC)

All of a sudden William Thweatt came only to agree with you and all of a sudden disappeared. Well, OK, I'll not have presumptions, but you just did when you assumed that I can't differ phones from phonemes. Maybe you didn't understand that the pronunciation differences were not noting allophones alone which are the fewest, but the regional variants. Your current edit has errors, and also has some unnecessary style.
The templates of IPA-(lang) don't work the way you wanted us to use them, see Arabic pronunciation: [/ħaˈmaida/] Egyptian Arabic pronunciation: [[Help:IPA for Egyptian Arabic|ħeˈmeːdæ]]? And a note on the name: the Egyptian name has the origin of Classical Arabic /ħumajda/. As if transcribing Michael [ˈmaɪ̯kəɫ] as /mixaʔel/ well because that's the origin of the name!
Your suggestion to use too many templates is anything but simplification. One would not be sure whether to use this or that or another template.
Finally, I plead to you to help me access the sources you provided because there are many mixed up and falsely generalized claims you are currently citing! --Mahmudmasri (talk) 03:35, 1 December 2015 (UTC)
I didn't "disappear" (all you'd have to do is check my contribution history and/or ping me if you want further input). I still am in complete agreement with Luboslov, but I have other things to do and can't monitor this particular talk page all day every day. Also, much like Luboslov, I tire of your repeating the same non-arguments and not understanding the problem here.--William Thweatt TalkContribs 04:23, 8 December 2015 (UTC)
I cannot say more than to use internet search engines. Gairdner is freely downloadable from However, I wonder a little, how you edited this page before if you did not know and obviously did not use those quite basic well-known sources. The personal experience of an editor cannot be a reliable source at all, it is Wikipedia Rules 101.--Lüboslóv Yęzýkin (talk) 00:58, 9 December 2015 (UTC)
And your example with Michael hits the mark. We here do not transcribe the name the way you did, but /ˈmaɪkəl/. Your transcription [ˈmaɪ̯kəɫ] is a good approximation how this page looked before. It was a narrow transcription, which cannot be properly sourced or attributed. Now this page follows a "/ˈmaɪkəl/" way, not a "[maɪ̯kəɫ]" one.--Lüboslóv Yęzýkin (talk) 01:15, 9 December 2015 (UTC)
Any other language, other that English, is transcribed here using a narrow transcription and therefore for analogy I made my example, but it doesn't make a difference, /ˈmaɪkəl/ is very different from /mixaʔel/.
  • It's such a pity that you couldn't simply see which Arabic letter corresponds to which IPA symbol, because the Arabic script doesn't spell in a phonemic way. Half of the vowels are not spelled. So, yes, you should have some knowledge of Arabic to use the help page to transcribe words. I even refer not only to help pages but also phonology pages of given languages to transcribe words in non-familiar languages.
  • Citing pronunciations for Arabic is problematically impractical. For English, we can find easily accessible dictionaries citing how to pronounce words, but nothing is similar for Arabic. Why do you require me to cite how every word is pronounced? You demand the impossible!
  • I asked for accessing sources because I never find complete versions on the internet; if I changed obviously wrong claims in the misguiding help page, you'd claim it's uncited. Well, the Gairdner's source is interesting but from 1925.
  • It appears that he described Arabic pronunciation as used in reading Koran, which is a very special and a different case. He even admitted that [ðˤ] is the pronunciation in Koran only and otherwise it's [zˤ] (in Egypt; page 21).
  • He had his own method of using the IPA symbols, and IPA itself wasn't as developed as it is today. He was prescribing to foreigners (English speakers) how to pronounce Arabic, not how it is taught or used in the Middle East, clearly as he claimed the [ɡ] is only colloquial (in page 14) but then wrote it's generally speaking in the literary language (page 23).
    • Did you know that in gemination, users of /dʒ/ will geminate it as [ʒʒ], rather than [dd͡ʒ] or [d͡ʒd͡ʒ]. And did you know that geminates always belong to two different syllables? So they can't simply be transcribed with ː. Example: [ʔæxˈxæːð], is definitely not the same as [ʔæˈxːæːð] or [ʔæxːˈæːð], defying all the syllable structures of all Arabic dialects! Lüboslóv Yęzýkin, you currently advised that transcription in the help.
  • He also used the pharyngealized-or-velarized   ̴   diacritic rather than the pharyngealization ˤ or the velarization ˠ mark, then described the emphatics as velarized.
  • He based the previous claim on old Arab phoneticians, which also described the pronunciation of Literary Arabic to what all contemporary Arabic speakers don't pronounce as such in modern times, even the most hardliners.
  • He even prescribed using two symbols for glottal stop!
  • Classical Arabic, even as used at his time, has /v, p/ in loanwords. He claimed it lacked them, implying that they are never used at all. They are marginal, not non-existent.
  • I didn't find where did he base his pronunciations on? The person or persons, where did they come from? It is normal to mention these, like in describing English phonology, the scholars would mention these.
  • It's even more apparent in his old descriptions for Hebrew distinguishing between /k/ and /q/ (page 26), but is non-existent in Standard Hebrew.
  • In page 31, he seemed to mix between Egyptian Arabic glosses of Classical Arabic origin, with Literary Arabic loanwords in Egyptian Arabic.
  • Page 34 is unbelievable! Claiming that the open front vs open back vowels don't have any significance to Arabs, is plain wrong. Only careful speakers or highly educated Egyptians are able to pronounce, for example [sˤ], the primary distinguishing factor between /s/ and /sˤ/ in such words, whether they are spelled by س or ص is by turning æ to ɑ and for the same people if the words lack any open vowel, they might be confused with similar words if not understood from context. Backing the open vowel exists even when pronouncing Literary Arabic and the Koran! He went further and claimed that they are merely colloquial /aw, aj/, but /aw, aj/ are monophthongized to [o, e], not his claim [ɒ, ɑ, ʌ, æ].
    • His description of how words with emphatics are pronounced reminds me of a jocular video, resembling how would rural Egyptians try to make artificial purification for their pronunciation.
That's enough.
  • It appears that no single source can be the basis of the help page.

--Mahmudmasri (talk) 08:02, 9 December 2015 (UTC)

Your bashing of Gairdner, whose work is still, after 90 years, regarded well by Arabist is not impressive. Yes, this page is not based on a single source. And it would not certantly be based on a single unreliable source called "a native speaker who knows best than those scientists". That you personally cannot or do not want to acess them is your personal problem. As you might notice I could acces them, and Arabic is not even my main area of interest. And you, a boastful "native speaker", cannot access sources known by every student Arabist! --Lüboslóv Yęzýkin (talk) 23:15, 13 December 2015 (UTC)


I think the discussion is becoming pointless, we repeat the same things again and again. Better to see, than hear another thousand words. I've already created a revised version. It is not based on the current, but rather fully reworked. I do not pretend it is the best version, but it sticks to one variant of transcription, while not ignoring entirely allophones. Nearly everything is sourced. The wiki markup is also different and much more neat. I hope it would not be reverted right away.--Lüboslóv Yęzýkin (talk) 11:44, 30 November 2015 (UTC)

I repeated nothing to you. You are just loaded with prejudice. Please read my latest reply above. --Mahmudmasri (talk) 03:35, 1 December 2015 (UTC)
Please see No Personal Attacks. Calling another editor "prejudiced" isn't going to help the situation. The problem is this "help" page is not helpful at all and is needlessly complicated by local variants. Stick to the issue and work toward a solution.--William Thweatt TalkContribs 04:23, 8 December 2015 (UTC)
I have no more prejudices (against what?) or errors, than the cited Arabists. And all small mistakes could be easily corrected, that's not much of a problem. --Lüboslóv Yęzýkin (talk) 00:29, 9 December 2015 (UTC)

It's "Help:IPA ..." not "Help:Arabic script"[edit]

This is the first "Help:IPA ..." page I've seen that is not focused on IPA. Instead, it is utterly hard to find help for any given symbol because of the following reasons:

  1. IPA is not in the first column
  2. The table is not sorted so that one can find IPA symbols easily
  3. The IPA column is much less legible than that of other such pages since it uses slashes "//" around the symbols and ...
  4. ... adds pronunciation notes in the same column and ...
  5. ... uses letters for those notes, which confusingly are the same as those used in IPA.
  6. It does not contain some IPA symbols that have by others been found important enough to be included in the topmost IPA help page, Help:IPA#Main_symbols, such as [G], or relegates them to footnotes, such as [χ].

Some of these would arguably make sense if the title of this page were "Help:Arabic script and pronunciation" instead of "Help:IPA for Arabic". But that's what such pages are needed for.

Since I'm not sure if the above discussion has reached a conclusion (sorry, it is TLDR), I'll wait a bit before making any major change. — Sebastian 04:08, 8 December 2015 (UTC)

I afraid the above discussion, at the rate it's progressing, will never reach a conclusion, just a stalemate. This is supposed to be the Help:IPA page for Modern Standard Arabic, not multiple local variants (see my proposed solution above). Also, all of the problems you point out need to be addressed as well. If you have a way to fix this and feel inclined to do so, I encourage you to be WP:BOLD. It will, at the very least, give us a place from which to start.--William Thweatt TalkContribs 04:31, 8 December 2015 (UTC)
I've stopped to participate in the discussion because I did not see what else to say. Moreover, I'm somewhat tired from Arabic, I spent fairly enough time for it and I decided to make a break from it.--Lüboslóv Yęzýkin (talk) 00:22, 9 December 2015 (UTC)
Done. By #2 do you mean an alphabetic "abcde..." order? I tried to follow the logical order of the IPA. As for #6, we have spoken about that already, see the above discussion.--Lüboslóv Yęzýkin (talk) 00:22, 9 December 2015 (UTC)
Thanks a lot, Lüboslóv! Sorry, I missed your comment here; if I had seen it earlier I would have answered that what I meant was the same sorting applied in the Help:IPA page. — Sebastian 02:01, 10 December 2015 (UTC)

What to do about hard to pronounce consonants?[edit]

I wonder how helpful the words "None, emphatic ..." really are for readers trying to approximate the pronunciation of emphatic consonants. "None" seems utterly useless and misleading, since it means Ø. Everybody will at least try to pronounce something there, at least the non-emphatic version of the consonant.

Now, I see this as a special case of a larger problem: This list doesn't follow a clear concept on how much detail we want to offer. Our parent article, Help:IPA has a separate column for explanations, as in

   Symbol     Examples Description
 (i) listen [ q ]   q   Arabic Qur’ān Like [k], but further back, in the throat.

Since our Examples column are English examples, they already serve as a kind of description, which makes the description column not necessary. However, we could use it to provide two different levels of explanation for hard to pronounce consonants, such as the following:

Symbol English
Description Arabic
twinkle t pronounced while raising the back of the
tongue as in w, but without its lip rounding.
ط ...

For the first, most simplistic level, I'm using "twinkle" as a sample because the English <w> at least contains some velarization, which is a step in the right direction, and because I think C[w]V is rare enough in Arabic to not lead to confusion. But that's only a side point here; I just wanted to provide some example for a simplistic explanation that's better than "None". The column would work th same way if it contained "tin".

The Description can then clarify it for those readers who want to be more exact.

Would that work? Or does anyone have a better idea? — Sebastian 02:01, 10 December 2015 (UTC)

Indeed, many non-Arabs have been pronounced the Arabic emphatics as labialized /tʷ sʷ zʷ/ (note, that and /ḏ̣ are merged in most pronunciations, both native and foreign). But here we meet some problems:
1) The labialization in the emphatics is rather a secondary or tertiary articulation. It is noticeable in many realizations but it is not the point of the emphatics. The main area of the articulation is at the back of the oral cavity.
2) Another point, there are Arabic words where the emphatics are followed by /w/. Not so many, though, but amongst them, for example, a wide-spread name رضوان‎ Riḍwān.
3) Many non-Arabs perceive the Arabic emphatics as plain consonants of a corresponding non-Arabic language. That means, from the one side, the emphatic quality is not so obvious and noticeable for an outsider (unlike, for example, ʿayn which is a very noticeable Arabic sound), and, from the other side, the difference between Arabic plain and emphatics consonants could be more properly compared to tea and too (or tip). Such an approximation is provided for Irish which have a full range of velarized consonants.
A conclusion: I think it would be better to show English /Ciː/ for Arabic plain consonants and /Cuː/ for emphatics.--Lüboslóv Yęzýkin (talk) 22:53, 13 December 2015 (UTC)
And your first point: we have no strict style guides how to write such "Help:IPA for..." pages. In most cases they do not follow the "Help:IPA" page, which is not the "parent page" for them as well, look to any of such page (the example of Irish has been already given). Your strict alphabetic order is a little overdo, in most cases at least consonants and vowels are listed in different (sub)tables, but I will not argue against, be it as it is now, it's not important.--Lüboslóv Yęzýkin (talk) 23:01, 13 December 2015 (UTC)
As Luboslov says, these are rather small details. The table as it is now is a vast improvement in readability for the casual reader, but for the record, I would prefer leaving the English Approximation column blank for emphatics and just describe (in layman's terms) the sound. I know it's just an "approximation" but English doesn't really have a sound that even approximates an emphatic. Also, the "t" in "twinkle" is clearly aspirated (at least in my dialect), which brings me to my next point. The English Approximations for /t/ and /k/ should be something like "stop" and "skip" to show lack of aspiration. Thirdly, I would also prefer that vowels and consonants be separated. That provides some logical organization and, is fairly typical of our Help:IPA pages -- especially in cases where the native script is an abjad or abugida (see Help:IPA for Hebrew, Help:IPA for Thai and Lao, etc.).--William Thweatt TalkContribs 23:52, 13 December 2015 (UTC)
Note, that the Arabic /t k/ are aspirated, so there is no need for English st-/sk- examples. The Arabic emphatic consonants are not principally different from the Irish velarized ones, so I do not see why we cannot aproximate the former if we can the latter.--Lüboslóv Yęzýkin (talk) 20:39, 18 December 2015 (UTC)
Ah, I see. The source I had on hand just used /t/ and /k/ and didn't explicitly say they were aspirated (neither does our article on Arabic phonology for that matter), but a quick check of a few other sources confirms you are correct about that. Our articles should use [tʰ] and [kʰ] or at least note that we're using /t/ and /k/ to represent aspirated stops in Arabic. And, while emphatics maybe not principally different from velarized, they are practically different. With the "tongue action" even further back (Uvularization) or Pharyngealization), emphatics are even further away from any possible English examples. But this is all a matter of degrees and a small issue to the non-linguist reader. I'd rather not be misleading, but if you and others think those approximations are helpful, I won't disagree.--William Thweatt TalkContribs 11:56, 19 December 2015 (UTC)
The note to the table already mentions it. The usage of /t, k/ is for the same reason as for English: there is no phonemic contrast between aspirated and unaspirated stop consonants in Arabic, no need to write aspiration explicitly in the transcription, it's redundant.--Lüboslóv Yęzýkin (talk) 11:29, 23 December 2015 (UTC)

Two-level explanation[edit]

I realize it was a mistake when I mixed two different ideas in the section "What to do about hard to pronounce consonants?". Please therefore allow me to single out the discussion about having two columns, as in the following:

Symbol Rough English
More detailed description Arabic
q scar pronounced further back than the c in scar. ق q g ...

Can we make this change? I believe it would invite healthy compromises between between people who want this page to be as easy as possible and people who want it as exact as possible. — Sebastian 01:55, 17 December 2015 (UTC)

Not strongly against, but it reminds me the descriptions from 18th-19th centuries books or from modern tourist phrasebooks. I'm not sure if such arbitrary descriptions do any good.--Lüboslóv Yęzýkin (talk) 20:39, 18 December 2015 (UTC)

Vowels separate?[edit]

Two respected editors have opposed my sorting vowels alphabetically together with the consonants. I grant that my reference to the default page (which I somewhat clumsily called "parent page") has at least to some degree been neutralized by pointing to other Help:IPA... pages. I agree with both of them that it's not a big issue, so if we just go by majority, I'm fine with such a change. I am, however, not convinced of the merits of that change, and would like to take the occasion to state my case:

This page is intended for non-linguists, so we should remember that even such an admittedly fundamental concept as the distinction between vowels and consonants does not come naturally to many readers. (Just as an aside, it isn't that easy for linguists, either, with phonetic vs. phonological definitions. But that's not an issue for Arabic.)

More importantly, on seeing a symbol for the first time one can't know if it represents a vowel or a consonant. Please, put yourself in the shoes of someone who doesn't know IPA. How would you know in which table to look for "ʃ", "ħ", and "ð"? You can tell that they look similar to "S", "h" and "d", but that doesn't mean that they represent consonants: "θ" resembles an "O" or an "e", but it's not a vowel. "ɣ" resembles "y" which one might also search under vowel.

That was my reason for sorting them alphabetical by similarity. I'd be curious to see any arguments for breaking that order. — Sebastian 01:55, 17 December 2015 (UTC)

I was the one who has raised some doubts, but nevertheless I understand your arguments. Many might know the IPA, but most editors/readers most likely do not. Even if this help page be the only one with such an "abc" order, it will not do any harm. Do not consider it as a vote "for", I do not think the order is very important.--Lüboslóv Yęzýkin (talk) 20:39, 18 December 2015 (UTC)