Talk:Arabic phonology

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Languages (Rated C-class, Mid-importance)
WikiProject icon This article 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.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
WikiProject Linguistics / Phonetics  (Rated C-class)
WikiProject icon This article is within the scope of WikiProject Linguistics, a collaborative effort to improve the coverage of linguistics 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.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.
Taskforce icon
This article is supported by Phonetics Task Force.


I have really problems understanding this:

However, by the 8th century, the letter alif no longer represented a glottal stop only, but also a long /aː/. As a result, a diacritic symbol, hamza (ء),[citation needed] was introduced to represent this sound with alif, and a hamza can be used, separately, now without the letter alif, to indicate the sound.

It should definitely be more specific about "which sound". — Preceding unsigned comment added by (talk) 23:04, 29 August 2014 (UTC)


If anyone could add IPA to this, it would be helpful. Thanks!

Uhh, I somehow doubt Arabic phonology is quite as simple as a 1:1 correspondance with the writing system. This article seems too simplistic to be really useful.

One-dimensional Tangent 03:35, 14 Mar 2005 (UTC)

I've been studying Arabic for a few months now, and I've definitely noticed a difference in the vowels /i/ and /u/ around emphatics as well. The seem to have a schwa-glide towards the emphatic. I don't have any sources, but if anyone else does that would be a good addition. Makerowner 04:04, 17 November 2006 (UTC)

One-Dimentional Tangent, you are right about the absence of 1:1 correspondance when taking into account dialects and regional pronunciations. However, the nature of Fus7a is to keep the language standard, so there is relatively little variance (in theory).

Also, could someone with the technical know-how change the placement of (ظ) on the chart? It should be a pharyngealized voiced INTERDENTAL fricative, not a DENTAL. It does swing to a DENTAL in Egyptian and other regional pronunciations, but it is clearly the former in proper Fus7a. Thanks!

The short vowel /a/ does not differ in quality from the long vowel /ā/. If we are talking about either Modern Standard Arabic or Fusha, it is incorrect to state that the short vowel should be rendered as "[ɛ̈] (open e as in English bed, but centralised)". Doubtless this is true of some dialect somewhere. Maybe the author is trying to approximate the Tunisian /a/, which falls precisely on the dividing line between American English /a/ as in "bad" and /e/ as in "bed". Recommendation: change to "[æ]" as in "bad", except after /r/, where it is pronounced [ɑ], as the "o" in American English "hot". --Cbdorsett 05:59, 14 January 2007 (UTC)

In American English "bed" is pronounced as [bεd] but in British English as [bed]. As the Arab vowel is mostly [æ], yet sometimes [ε], but never [e], I do think that "bad" would be a better example. —Preceding unsigned comment added by (talk) 12:41, 1 September 2007 (UTC)

I reckon the article should mention the distinction between those consonants which get doubled after al-, when the leem becomes silent, (called "sun letters" because of the example "ash-shams") and the other consonants (called "moon letters" after the example "al-qamar").

Velar or uvular?[edit]

Isn't Arabic famous for having uvular fricatives ([χ ʁ]) rather than velar ones ([x ɣ]? Is that a difference between dialects or something? David Marjanović 20:55, 24 June 2007 (UTC)

Audio files: [χ] [1], [ʁ] [2], [x] and [ɣ] [3]. David Marjanović 21:28, 24 June 2007 (UTC)

The standard pronunciation of MSA features velar fricatives, but plenty of dialects have uvular realizations.The Dropper 21:42, 6 July 2007 (UTC)

Being German, I'm a bit of an expert in uvular and velar sounds and I'd say the difference is minimal.—Preceding unsigned comment added by (talkcontribs)
Being a native speaker of German (from Austria) myself, I'd say the difference is very easy to hear! Compare your speech to that of a Swiss or someone from Vorarlberg, and you'll find the difference between [x] and [χ]. David Marjanović (talk) 01:30, 6 December 2007 (UTC)
In all my time speaking MSA with Egyptians, Moroccans, Syrians and Palestinians, I've never heard a velar fricative once. Could someone point me in the direction of a recording of an Arab actually using these sounds? I'm not doubting that they exist. But I am skeptical of their being standard. If they were standard, I would expect to have heard them at least a few times from cultured Arabs. Szfski (talk) 19:47, 14 May 2008 (UTC)
I'm not sure if the standard makes the distinction between velar/post-velar/uvular. It could be one of those things that's influenced by the speaker's dialect. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:08, 15 May 2008 (UTC)
None of the dialects I am familiar with make the velar/uvular distinction. For that matter, none of the dialects I know actually have a velar fricative in their phonetic inventory. I've checked with every native Arabic speaker I know, testing the uvular fricatives on them, asking whether or not this sounds like something an Arabic-speaker might produce. The response was universally in the negative. I realize this qualifies as utterly useless and inadmissible OR. However, my theory is that, since a great many textbooks for Arabic learners describe the back fricatives in terms of velars (i.e. "Pronounce it like Russian X, German Ch, or like the 'ch' of Scottish 'loch'") this comparison has caused the sounds to be described as if they were velar instead of merely similar to velars. I'll look for some good source material on this later. Sorry if my English is a bit crummy. It's late. Szfski (talk) 06:51, 25 May 2008 (UTC)

IPA transcription questions or improvements?[edit]

Shouldnt [ɛ̈] be transcribed [ɜ] ?

Shouldn't [sˁ] be transcribed [sˤ] ?[4]

--Sonjaaa 03:24, 29 July 2007 (UTC)

As far as I understand, [ɛ̈] is not as central as [ɜ] is. What are the advantages to using <ˤ> over <ˁ>? —Preceding unsigned comment added by Aeusoes1 (talkcontribs)

Changes to IPA chart[edit]

Why was the chart changed? [g] is simply not a phoneme of standard Arabic. It is the realization of /dʒ/ in Egypt, but the standard pronounciation is still /dʒ/.The Dropper 07:21, 6 October 2007 (UTC)

If I understand correctly, speakers of Egyptian and other dialects that realize this as /g/ pronounce it as [g] even when speaking Standard Arabic. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 08:28, 6 October 2007 (UTC)

But that's not the standard pronounciation, which is what I thought the chart was intended to describe. MSA has a standard pronounciation which is taught to foreign learners and very important in reciting the Qur'an.

Also, notes are necessary to describe that the lateral approximant is only velarized in certain situations pertaining to the word for God, الله. It's definitely not a phoneme and I'm not sure if it should be included on the chart even in brackets.The Dropper 03:07, 8 October 2007 (UTC)

If you feel more comfortable with the table only showing the /dʒ/ that's fine, but we should still have the note. You're right about the velarized lateral being only in Allah and its derivatives but we shouldn't remove it from the table (that would be a bit of OR). What about the velar/uvular fricatives? Speakers of most dialects pronounce them as velar but I don't know if its different when they're speaking Standard Arabic. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 04:20, 8 October 2007 (UTC)

I see changes have been made. Looks good to me.:)

So the only question, I think, is with the back fricatives. I think it'd be sufficient to denote them as velar, with a note that they may be uvular depending on dialect or whatever. What do you think?The Dropper 20:24, 8 October 2007 (UTC)

I don't know what the standard is supposed to be. The article right now still has the dʒ ~ g with the tilde and if that's acceptable (with a note) then the velar-uvular one should be also. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:46, 8 October 2007 (UTC)

The standard pronunciation of Arabic is based on what's used for reciting the Qur'an. It's taught to foreign learners, as well.

This article needs some improvements. Unfortunately, some areas of Arabic phonology are a little hazy, because any language with so many speakers covering such a wide area will inevitably have some variation. For example, the "emphatic" consonants of the language are most often described as pharyngealized but I've heard they are velarized as well. —Preceding unsigned comment added by (talkcontribs)

I understand how the Standard is used but I don't know, if it's specified, whether the dorsal fricatives are uvular or velar. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 09:37, 10 October 2007 (UTC)

Yeah, it's not specified. I think it varies according to dialect and maybe other factors. However, I think that velar is the dominant pronunciation ... maybe post-velar. Hard to say.The Dropper 17:42, 19 October 2007 (UTC)

More about the IPA chart[edit]

The article on the voiced epiglottal fricative cites the great and mighty Ladefoged as finding that the pharyngeal consonants are epiglottal, rather than pharyngeal, in some dialects. Should that be mentioned?

Also, the chart says [ðˤ], while the frequency table says [zˤ]. That's another dialect difference, or rather a regional difference in the standard, right? If so, this should be explained, too. David Marjanović (talk) 01:41, 6 December 2007 (UTC)

I'm going to change the frequency chart (/ðˤ/ is transcribed with a z) but we could indeed mention the dialectal variation. We could also mention the pharyngeal/epiglottal thing though if we can be specific about which dialects or regions that would be even better. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 06:12, 6 December 2007 (UTC)

Classical and modern standard pronunciation of ج?[edit]

The portion about /dʒ/ is very confusing. Many Arabs insist letter ج should be pronounced /ʒ/, this being standard, not colloquial pronunciation. I don't understand the section " In classical Arabic, this was either [ɟ] or [gʲ]." Where is this from - what's the source? Is it correct? What are these sounds? Isn't the classical pronunciation /dʒ/ of the letter? It seems it has shifted to /ʒ/ (Eastern Arab countries) and /g/ (Egypt and Yemen(?)) but otherwise, isn't the other pronunciation are part of spoken dialects? MSA in Western Arab countries still use /dʒ/, if I am not mistaken. Also, please confirm that [j] is used in formal pronunciation of the letter in some countries. (We are discussing standard, not dialectal phonology?). In Egypt, pronouncing ج as [g] is considered standard. --Atitarev (talk) 23:29, 21 January 2008 (UTC)

The source is Watson (2002). Look at Classical Arabic. Phonemically, there is no difference between Standard and Classical Arabic, but phonetically there are a few differences. Unfortunately, I'm not sure if the different pronunciations apply to standard speech in those areas as well as dialectal speech or if it's just the latter. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 01:48, 22 January 2008 (UTC)
Thanks. Isn't there just one sound [ɟ], not 2? Quote: "The palatals /ɟ/ /ç/ (<ج> <ش‎>) became postalveolar: /dʒ/ /ʃ/" (EDIT: the Arabic letters have become reversed in my post - I meant /ɟ/ only applies to letter "ج".) --Atitarev (talk) 05:38, 22 January 2008 (UTC)
I'm not sure what you're asking. [ç] is also a palatal sound and in classical arabic was represented by shinƵ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 06:49, 22 January 2008 (UTC)
I mean, the reference is for the sound /ɟ/ only but the article says "In classical Arabic, this was either [ɟ] or [gʲ]." Is there a reference that it was also pronounced as [gʲ]?--Atitarev (talk) 21:45, 22 January 2008 (UTC)
The reference claims that it was either [ɟ] or [gʲ]. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 23:15, 22 January 2008 (UTC)

I always thought /ʃ/ in Arabic developed from Proto-Semitic /ɬ/. At least, that's what happened in Hebrew. (talk) 20:57, 29 January 2008 (UTC)

Emphatic L?[edit]

The article states that /lˤ/ occurs only in a handful of loanwords and /ʔalˤˈlˤaːh/, the name of God, i.e. Allah, except when following long or short /i/ when it is not emphatic: bismi l-lāh [bismilˈlaːh] ('in the name of God').

However, work by a number of phoneticians has demonstrated, fairly conclusively, that the dark /lˤ/ occurs far more often than this. Charles A. Ferguson, for example, has shown that modern speakers of MSA tend to import their phonemic Dark L from dialects, such as Iraqi, where dark L is far more widespread. —Preceding unsigned comment added by Szfski (talkcontribs) 03:56, 5 May 2008 (UTC)

Oh man, so that's something that varies amongst speakers from different regions. It's strange, then, that MSA is given as a uniform thing when there's variance in a number of ways. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 04:12, 5 May 2008 (UTC)
True. The problem is that the different ways in which various Arabs speak MSA is poorly studied. Usually, when the phonology unique to a particular Arabic-speaking region is examined, it is almost always in the context of that region's colloquial dialect, and not the regionally influenced form of MSA that has sprung up in that area. Give me a few years to get my Ph.D. and I'll see what I can do about that. Until then, I'll do what I can with references like the one I just added.Szfski (talk) 07:03, 5 May 2008 (UTC)
Your English is quite clear. Certainly better than my Arabic. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 06:54, 25 May 2008 (UTC)

Foreign or new sounds in standard Arabic[edit]

Should the sounds missing in the classical Arabic but commonly used for foreign names be included, e.g. consonants: v, p, g, ʒ, tʃ and vowels: o, e, o:, e:? Atitarev (talk) 23:39, 20 August 2008 (UTC)

That's a possibility, though I'd rather put it in prose than stick it in the chart as a number of them are rare or dialectal. Here's what Watson (2002) says of Cairene:
  • /v/ and /p/ are found in the speech of educated speakers (she also points to /v/ as being present in educated speakers of other dialects but it's not clear how widespread this is from her comment)
  • /rˤ/ is found in European loanwords (thought it also appears in "native words with guttural vowels." )
  • /bˤ/ contrasts with /b/; though Watson doesn't explicitely say it comes from loanwords, the one example she gives is /bˤaːbˤa/ ('pope/pontiff/patriarch') which seems like a loanword.
  • /ʒ/ is found in a few loanwords.
In San'ani, [tʃ] appears from the concatenation of the perfect aspect subjective suffixes (/maːlibistʃ/ 'I/you[m.s.] didn't wear') in addition to loanwords such as /tilivizijuːn/ ('television') but in Cairene an epinthetic vowel breaks up the stop and fricative. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 03:36, 21 August 2008 (UTC)
Thank you for that. I would appreciate the references to the names you mentioned as I am not with Watson (2002). Also, do you have a resource to check standard pronunciation of foreign words in Arabic? --Atitarev (talk) 03:53, 21 August 2008 (UTC)
Aeusoes1, Rather than recklessly removing my edits like you did today, you could edit and make the information more user-friendly and provide sources. It's not always what YOU prefer to be in the article, it's for others to read and learn. The information on pronunciation of foreign names in Arabic is scarce as it is, please don't make it worse. Also, you need to respect other editors. --Atitarev (talk) 05:04, 1 September 2008 (UTC)
Pardon me for not being clearer, most of the information you've included is already in the article, especially after this edit where I did indeed "edit and make the information more user-friendly and provide sources." Keep in mind two things:
  1. Just because it's not worded and placed exactly the way you've edited does not mean that it's not present
  2. The generalities regarding Arabic pronunciation that you've included may be wrong or oversimplistic.
That said, here is a quick breakdown of your inclusion versus the article
  • You included: "Vowels /o/, /e/, /oː/ and /eː/, which are missing in the Classical Arabic but exist in many spoken dialects, are becoming more common in Modern Standard Arabic". The article states of Cairene Arabic "Classical Arabic diphthongs /aj/ and /aw/ became realized as /eː/ and /oː/ respectively; loanwords from Standard Arabic reintroduced the diphthongs, sometimes with minimal pairs..." and says in the list of common trends "Monophthongizing diphthongs such as /aj/ and /aw/ to /eː/ and /oː/, respectively"
  • You included: "especially in geographical and personal foreign names." Not sure about this. It strikes me as unsourced speculation.
  • You included: "Some consider this non-standard but this pronunciation is becoming common in the media" who considers this non-standard? Arabic-speaking media is far from uniform and this seems too much of a blanket statement.
  • You included: "Consonants /v/, /g/, /p/, /tʃ/ and /ʒ/, which are missing in the Classical Arabic but exist in many spoken dialects, are becoming more common in Modern Standard Arabic," the article states "A number of dialects have the marginal phonemes /v/ (for educated speakers) and /p/, largely from loanwords." as well as "Cairene also has /ʒ/ as a marginal phoneme from loanwords." You may notice that there still is no mention of /tʃ/ and that's because I only have information on Ṣan‘ā’ni Arabic having it and Cairene specifically not having it.
  • You list the Arabic words for Volvo, Vienna, and Seven-up, Moscow, Melbourne as examples. I don't really doubt the truth of the existence of these words but, again, overgeneralizing the whole of Arabic-speaking areas as having these pronunciations may be a stretch. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 06:10, 1 September 2008 (UTC)
A common well-known example of /tʃ/ is سندوتش /sandawitʃ/ (sandwich) where /tʃ/ is rendered with تش (t + sh). is the reference, Hans Wehr will have an entry (there is also a variant with a 'yaa' to make /i/ longer - /sandawi:tʃ/). If you ask me how do I know the pronunciation? From common textbooks, for example, "Teach Yourself Arabic" (J. Smart & F. Altorfer), which used the same spelling, romanisation and the sound recording. Apart from "Seven-Up", a curious and common example of /p/ is "jeep", which can be written in about 3 ways but especially for Egyptians can be written as چيپ /dʒi:p/ (using Persian letters to make sure it's /dʒi:p/, not /gi:p/, which may be pronounced by uneducated as /dʒi:b/, of course.
The other examples above I received from educated Arabic speakers and confirmed by listening to Arabic radio (in MSA).
I can get more examples of a long /o:/ used in standard Arabic, for which I can provide references. Another reference is again a textbook: "Elementary Modern Standard Arabic" (Cambridge University Press). "Doctor" (دكتور) is pronounced and romanised as /dukto:r/ (the textbook uses duktoor but oo stands for the long /o/, not /u/, "Georgetown" as /dʒordʒta:un/, Texas (تكساس) as /teksas/, Robert as /robert/ (both /o/ and /e/), New York as /nju: jork/.
Following recent political events, a number of Russian surnames were uttered in Arabic media. You can clearly hear "Sergey Lavrov" /sergej lavrov/, and Dmitry Medvedev /dmitri: medvedev/, where /v/ sounds like, well /v/. Is Al-Jazeera radio a good source for a standard pronunciation?
I am going to be busy these days, if you think some of the paragraphs are redundant, please suggest how to merge with some examples, I don't insist on the wording but I would like to include some examples. I haven't provided the Arabic spelling of all the examples, let me know if you need them.
As for how general these examples are, the trouble is, can anyone say for sure like in English can you say that letter H is called "aitch", not "haitch"? This all may look like an original research but I am doing my best to provide the scarce resources. --Atitarev (talk) 12:52, 1 September 2008 (UTC)


  • I found the best source - the same old Hans Wehr Arabic-English dictionary. Examples of /v/ and /g/ - vīrus, vīsa, Ingilīzi (English).--Atitarev (talk) 13:59, 1 September 2008 (UTC)
Yes, you are right that examples are helpful. Thank you for providing information on where you get your information (while some of it is OR, it's still helpful). As for H, it is called "aitch" overwhelmingly in standard media. I'll incorporate the examples in the pre-existing paragraphs. Perhaps native speakers can come along and approve or disapprove of such examples. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:50, 1 September 2008 (UTC)
Thanks. There might be some formatting issues (brackets, quites, etc. My observation is and from what I was told and read about, Arabs don't have issues with any of these "new" sounds as many of them are bilingual or trilingual (the 2nd and the 3rd is usually English/French), the sounds exist in dialects. /tʃ/ is common in Levantine, Iraqi and Gulf Arabic /ismitʃ/ - e.g. your name (female singular). The most common confusion is between /b/ and /p/, /f/ and /v/ and /dʒ/ and /ʒ/ (the latter was already mentioned). Yes, there are alternative spellings and pronunciations. Arabs are generally reluctant to use non-Arabic letters, so the usage is not too high but spelling may not match the pronunciation.
My point is, these sounds are not only part of Arabic dialects but have been introduced into MSA.
I'd like to change and add some more common examples. "Vietnam" or "Vienna" is maybe a better example than "Seven-Up". We need some /g/ examples, where it stands for foreign /g/ (not dialectal Egyptian /g/ for letter ج or Gulf Arabic /g/ for letter ق.
I prefer "aitch" but there are too many people in Australia, NZ, Ireland and other places who insist on 'haitch'. This was a rhetorical question.
Here's caramel "Caprice" with a پ :) :

--Atitarev (talk) 23:12, 1 September 2008 (UTC)

Isn't there a bit of selection bias with the native speakers giving you such information? Maybe it's the American in me that doubts that enough Arabic speakers speak English that foreign sounds aren't an issue for most speakers.
We don't need too many examples as they ought to be illustrative, not exhaustive. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 23:46, 1 September 2008 (UTC)
I aimed to get this information. The trouble is that even if the textbooks mention slightly or use in the recording, they don't dedicate paragraphs to foreign words transliterations. It's my interest, I admit, so I searched and found what I need.
By adding, I mean something that's missing - e.g /g/ in foreign words, which can be actually rendered in at least 4 ways!:
* ك k (alternates with /k/)
* غ gh (alternates with gh)
* ج j common for Egypt where it is usually pronounced that way but transferred to other countries, which Hans Wehr mentions,
* and lastly but seldom with the Persian گ.
The word for "English" has 2 common spellings إنجليزي and إنكليزي, it's pronounced /g/ in both cases (ingilīzi) and not just in Egypt (there are many sources for this, e.g. Hans Wehr). City names ending in -burg usually use غ, which cause the alternative pronunciation.
It may look like exhaustive, perhaps, it calls for a new article on foreign name transliteration in Arabic but there are too many nuances in spelling, pronunciation and trends, which may not be written down in one place. Atitarev (talk) 00:36, 2 September 2008 (UTC)

Maybe it's the American in me that doubts that enough Arabic speakers speak English that foreign sounds aren't an issue for most speakers.

We are talking about standard Arabic, which in fact, is known well and mainly used by educated Arabs, especially if we are talking about mass-media Arabic. You may check that university (secular) education is often carried out in languages other than Arabic in the Arab World. The foreign language influence is too great on the spoken Arabic dialects, it's not hard to see ("service room" is used verbatim in Egyptian hotels), in speech Arabs use a lot of foreign words but in writing and in media they are trying to observe some language purity, which is not followed for foreign names. Some mix-up of sounds (b/p) happens because the foreign words not always transliterated in Arabic to disambiguate and some because of the traditional pronunciation, so Paris is باريس Bārīs, even if they know it's /pari/ in French. Atitarev (talk) 00:36, 2 September 2008 (UTC)
  • In brief: I added the new source: Elementary Modern Standard Arabic, by Peter F. Abboud (Editor), Ernest N. McCarus (Editor). It provides a few words of foreign origin romanised to show the pronunciation in Arabic, Georgetown, doctor, etc.
  • Hans Wehr's Arabic dictionary is also an excellent source of the pronunciation of many foreign words and place names - especially consonants /v/, /g/ and /p/ and vowels /e/ and /o/ (both long and short.
  • Teach Yourself Arabic, by Jack Smart (Author), Frances Altorfer (Author) also provides a number of words containing sounds, "foreign" to classical Arabic: chocolate, Coca-Cola, November, lemon, secretary (it's already agreed on) (giving here in English). Anatoli (talk) 02:36, 7 January 2009 (UTC)

confusion in article about different levels[edit]

IMO this article, and the discussion among you two, are currently rather messed-up and are going to remain that way until you clarify some issues. Koranic Arabic, CA, MSA and dialects are all different and need to be clearly distinguished when talking about Arabic phonology. Consider the following different cases, and note that NONE OF THEM are a single system with a single "phonology":

  • The language of the Koran, as recorded in the Koran and with a phonology reconstructible by linguistic means. Has 28 consonant phonemes and 6 vowels. This is the closest you get to a single system, and even then, linguists differ on exactly how the sounds were pronounced -- and of course, there were different dialects, some with radical changes (e.g. Meccan Arabic had a separate /e:/ phoneme, preserved in writing with the alif maqsuura, and also had no glottal stop, again preserved in writing with all the hamza complications).
  • The modern pronunciation of the Koran. Has the same 34 phonemes but with quite different pronunciations in many cases as compared to ancient speech. Also has some differences depending on the speaker's native dialect, especially concerning vowels -- hence the current section on vowel pronunciation is basically garbage. For example, an Iraqi speaker's /a/ differs radically in its allophones from an Egyptian speaker's /a/, and an Egyptian's /a/ may vary depending on whether he's speaking dialect, MSA, or Koranic (so claims one of my Egyptian Arabic books, which says that a Cairene /a/ in the vicinity of /x/ is front in dialect but back in "Classical" -- whether MSA or Koranic is not specified). Iraqi and Egyptian ayin are also quite different (Iraqi ayin is pronounced with a simultaneous glottal stop), and I'm 100% certain this carries over into MSA or Koranic pronunciation. Most sources -- including highly technical ones written by linguists -- do not seem to clearly understand these facts, and hence their assertions about general tendencies are highly suspect.
  • MSA.
    • This is strongly influenced both by dialect, by register (i.e. level of formality of the speech context), and by social class. A typical speaker of MSA has a continuous variation of lects between most "Classical" and most "dialect", depending on how formal or informal the situation is. This applies equally to upper-class educated and lower-class non-educated speakers, provided they both know MSA; however, for any given register, the more upper-class and educated you are, the more "Classical" your speech is. (This latter tendency is not specific to Arabic; see sociolinguistics. Linguists use the terms "acrolect", "mesolect" and "basilect" to refer to high, medium, and low lects, with more or less deviation from a formal standard.)
    • Furthermore, a distinction needs to be made between "core" and "peripheral" phonemes. Core phonemes for MSA are approximately the same 34 phonemes as above; however, the lower the speaker's lect is (i.e. the more basilectal), the more it will diverge in the direction of the local dialect phonology, eliminating some phonemes (e.g. the distinction between emphatic /D/ and /DH/) and adding others (e.g. different /a/ phonemes).
    • On top of this, there isn't even a totally well-defined "formal standard" at the top, unlike e.g. RP. For example, the "formal standard" for MSA has two possible pronunciations for ج , either affricate or fricative, depending on which part of the Arab world you're in (Levant vs. North Africa vs. Iraq vs. peninsula), and has no pronunciation specified for ayin or for /a/ -- see above comments for Iraq vs. Egypt, for example.
    • Also, the peripheral phonemes such as /v/, /p/ or /e:/ need to be clearly distinguished from the core phonemes. Peripheral phonemes occur mostly in proper names or loanwords. If these sounds aren't present in the local dialect, they will tend to disappear in basilectal speech (e.g. /v/ -> /f/, /p/ -> /b/). Words like /duktoor/ and /sikriteer/ are quite stable because /o:/ and /e:/ are present in most dialects. On top of this, some words may "legitimately" have substituted phonemes, e.g. your example of /baariis/ "Paris" above, and will appear this way even in acrolectal speech.
  • Finally, the dialects.
    • Every dialect has a different phonemic inventory, and a different set of allophonic rules, AND a different set of suprasegmental rules, from every other. For example, as noted above, Sanaa speech often allows clusters of three consonants (at the end of words and elsehwere), whereas Egyptian doesn't. Egyptian is probably at one extreme of the continuum, in universally prohibiting three-consonant clusters, whereas Levant dialects occasionally permit them, peninsula dialects often allow them, and Moroccan is the other extreme, with clusters of six or seven consonants (or more) in some cases. Most Gulf and Iraqi dialects have a /tʃ/ phoneme, whereas neither Sanaa nor Egyptian nor Moroccan do. Gulf dialects typically replace /q/ with /g/, or even /tʃ/ or /dʒ/ (witness the city of Sharjah from CA shaariqah); Iraqi dialects vary (so-called "qultu" dialects vs. so-called "gilit" dialects); in Cairene and urban Levantine /q/ becomes /'/, but some Classical borrowings retain /q/; in Moroccan /q/ splits lexically into /q/ and /g/. There is also a dialect continuum regarding pharyngeal spreading, again with Egyptian and Moroccan as extremes: Pharyngeal spreading was apparently minimal in Koranic Arabic but in Egyptian Arabic may spread across the whole word, including prefixes and suffixes; in Levantine it's somewhat similar but tends not to spread as far, and is blocked by /i/ and /sh/; in Moroccan it doesn't (usually) go past the nearest full vowel, but its effects are more pronounced in that all vowels are affected, not just /a/ (/u/ and /i/ are lowered to /o/ and /e/). Many (most?) dialects allow initial geminate consonants, but not Egyptian; most reduce final geminates, and geminates before consonants, to single sounds, but Moroccan keeps geminates in all positions (and according to some analyses, even has phonemic triple consonants). Eastern dialects maintain Classical stress patterns, but Egyptian often moves the stress to the penultimate syllable (Eastern MAD-ra-sa, Egyptian mad-RA-sa); Moroccan has no consistent stress pattern at all. Etc. etc. etc.

Benwing (talk) 05:18, 2 September 2008 (UTC)

Thank you for your input, Benwing. Since we have an article on Classical Arabic, which covers 8th century pronunciations, it's wisest that the focus of this article (as this article states) should be on MSA. I don't know what sort of research has been done on generalizations for all or most dialects, so I don't know what's correct. Since, presumably, all local variations of MSA are mutually intelligible while local dialects are not, there must be many more similarities than differences. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:49, 2 September 2008 (UTC)
Thank you, Benwing. Although, you said our conversation is messed up, what you said is agreeing with what I suggested (doctor, secretary). Some sounds, which you call unstable are not that unstable, in my opinion, again depends on the speaker's knowledge, location. I also tend to trust Hans Wehr dictionary, check just a few words: visa, virus, Vienna, Vietnam and jeep to get my point. I am not insisting that those sounds have no alternative but the fact is they penetrated the standard high level Arabic, are used in the media, are used to taught foreigners, therefore are legitimate and have the right to be included. The Arabic phonology is not only the sounds of Qur'ān (Classical Arabic) but what constitutes the Arabic phonology as a whole. If dialectal sounds may be included why MSA sounds can not? Even if they don't make the core of the language and only appear in some (not all) foreign words and uttered by many (not all) educated Arabic speakers. What you said is true, there are no strict consistent rules that govern this pronunciation over the large Arabic speaking world. It is of interest to linguists. Why? If the `ayn (ع) sound doesn't exist in English, then it can't be included. If /p/ or /v/ exist in Arabic, they are worth to be included and explained when and why.
The rigidness of Arabic linguistics is well-known. They still talk about one and the only one Arabic language - the language of Qur'ān. Western linguistics about Arabic shouldn't be that rigid and clearly explain the facts and show the real situation. The real situation, among other things, is that the standard Arabic language has evolved (no surprise)and, (which is more recent), and absorbed some foreign or dialectal sounds, which were not originally in Qur'ān.

Atitarev (talk) 06:13, 2 September 2008 (UTC)

OK, Let me be clearer about what I mean.
  • The article's title is simply "Arabic phonology". It has only one sentence explaining the complex situation w.r.t. Koranic vs. MSA vs. dialect. It doesn't explain at all what "colloquial Arabic" is (this term is introduced without definition at the beginning of section 3). MSA is claimed to be "the standard variety shared by educated speakers throughout Arabic-speaking regions". If I didn't know better, I'd assume that MSA is the native language of educated speakers across the Arab world, just like General American is the native speech of most educated Americans. The term "Colloquial Arabic" brings to mind a slangy way of speaking English, which is NOT AT ALL the actual situation. ("Arabic dialects" would be a much better word.) The fact that MSA and "Arabic dialects" are entirely different languages, and that MSA is a semi-artificial construct and the native language of NO ONE, and MSA and the various different dialects are actually different *languages* (or at least, dialects well along towards becoming languages, with different vocabulary including of very basic terms, with quite distinct syntax, morphology, and phonology), would not at all be apparent, and hence much of the rest of the article would be misleading and difficult to understand.
  • Furthermore, the article (despite its assertions) doesn't just cover MSA. Section 3 contains a random, disordered collection of facts on local dialects. It should be deleted entirely and instead put a link somewhere near the top to Varieties of Arabic.
  • Section 1: As I pointed out previously, there is no single standard for how to pronounce /a/, and it varies depending on underlying dialect. Section 1 doesn't acknowledge this at all. Some or all of the sources used for section 1 described MSA phonology based on a single speaker or the speakers of a single dialect, without understanding that other dialects do MSA differently. Hence, section 1 in general is a mishmash of statements, some applicable to Iraqi MSA, some to Egyptian MSA, some to Moroccan MSA, etc. Only way to do justice would be to list the common trends, and then put comments about variations with particular MSA dialects.
  • Section 2:
    • Note 2: Note: There is more than one way to say "ayin". The source from Thelwall demonstrates a classic mistake made by researchers who are unable to accept the concept of a "pluricentric" standard. Thelwall interviewed a bunch of speakers from Iraq and Kuwait, and they all happened to pronounce a pharyngealized glottal stop (as normal in their dialects), so Thelwall concluded that previous linguists were simply "wrong" in describing it as as a voiced pharyngeal fricative. In fact, Thelwall was wrong, and this note needs removing.
    • Note 3: This section mixes up dialect and 6th-century Koranic pronunciaton into supposedly an MSA-focused article. It would be much more correct to say that, in MSA in specific, /dʒ/ is pronounced as /ʒ/ (NOT /g/) by some speakers, esp. in the Levant. /g/ is a dialect feature of Cairene Arabic and some Caireners *MAY* use /g/ when speaking pure MSA, but I suspect not.
    • Meanwhile, no note is made of the /q/ vs. /g/ difference, which DOES appear in the MSA of people where all /q/ is replaced in dialect speech with /g/.
    • Statement near the end that "between a geminate and a pause, an epinthetic [sic -- use "epenthetic"] [ə] occurs." This is sourced to Thelwall, which as I pointed above, doesn't understand (or at least, explain) the differences among dialects, and simply presents a fact from his informer as universal behavior. I have a copy of this essay, and it is full of mistakes of this sort -- statements referring to one particular behavior along the various continuums of dialect variations I discussed, and asserting that this is how Standard Arabic works. In reality, he's describing the MSA of one speaker, born in N. Palestine, educated in Beirut from 8 to 15, afterwards lived in Damascus, then Scotland, then Kuwait.

Benwing (talk) 07:45, 2 September 2008 (UTC)

Let's see if I can't take you point by point.
  1. Because of the complex sociolinguistic situation, there is no single phonological system but there is also not a whole lot of study on regional pronunciations of Standard Arabic. That alone indicates that we have to choose what we're going to focus on and the artificial standard variety is the best choice.
  2. While I didn't see the article as giving the impression you outline, I can see how it might, so I'd suggest we expand the lead a little and link to another place on Wikipedia that gives (or should give) great depth to the sociolinguistic situation.
  3. We could be more explicit by what we mean by "colloquial varieties" but the article's primary focus on the pronunciation of MSA does not mean that we can't discuss regional variations in regard to MSA pronunciations when there are regional pronunciations of MSA. No matter how you look at it, the section is woefully underdeveloped and consists only of my contributions from Watson (2002).
  4. On the vowels section and your discrediting of Thelwall, I ask that you provide sourced criticism of the scholarship rather than original research assertions. This is another woefully underdeveloped part, based primarily on a two page summary from JIPA (which, I agree, is pretty oversimplistic). As we expand on it, we may find sources that do indeed articulate commonalities amongst dialects' allophony.
  5. AFAIK, the jury's still out on the correct phonetic attributes of Ayin. We've got Thelwall's account as well as Ladefoged & Maddieson's. There may be more to include, but I haven't come across them.
So if you've got some good sources and wish to contribute to the article, you're welcome to. I'm hesitent to take away Thelwall, as I'd like the article to have sourcing, but perhaps as you expand (if you do so), the information Thelwall provides will fall by the wayside. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:40, 2 September 2008 (UTC)

Where did you get that vowel "ɐ"?[edit]

  • Where did you get that vowel "/ɐ/" for the Cairene Arabic dialect?
  • We pronounce it as short /æ/ or sometimes, as short /α/
    • In Cairo (& most of our famous dialects) we pronounce /dʒ/ of the official Arabic (Literary Standard Arabic, that is connected the Classical[Quranic Arabic]), [...] as /g/, ALWAYS!
  • In Cairo (& most of our famous dialects), we never pronounce /dʒ/ except when we recite the Quran (& in that case /ʒ/ is pronounced). Other than that, we never pronounce /dʒ/ [...] & if we used it from loan-words from foreign languages or Quran, we pronounce it as /ʒ/ . --Mahmudmasri (talk) 21:29, 24 October 2008 (UTC)
/ɐ/ is not explicitely stated for Cairene. We need to get more information on dialectal pronunciations.
  • Look user:Aeusoes1! If it were true that /ɐ/ is pronounced somewhere in Egypt, that would be rare, because I'm Egyptian & I'm sure that I've never heard that vowel before. If /ɐ/ was pointing at Cairene dialect, then it's absolutely wrong! I don't know if Cairenes used to pronounce it like that a very long time ago, but all our classic movies, songs or any media content DOES NOT HAVE THAT VOWEL AT ALL. --Mahmudmasri (talk) 00:38, 26 October 2008 (UTC)
Yea I tend to err on the side of the Masris here. j (dʒ or ʒ) is pronounced in parts of Egypt, but not in Cairene Arabic (it certainly is uncommonly heard in Cairo by way of immigration from the S3iid and other (probably primarily) Upper Egyptian areas. I am assuming (possibly incorrectly) that the ɐ confusion is from the example gibnɐ versus gebnɐ. I again agree, this is a pretty laughable example. Pronunciation-wise (and I would actually really love to be corrected by an Egyptian here as my normal source is in Dahab for Eid), "gibnɐ" is basically pronounced gibn-eh (i.e. the taa marbuta is not pronounced explicitly, or necessarily, as ɐ. And as for the geb-naa, that pronunciation is not really all too confusable, it is likely a contrived example. (i forgot the tildes, i posted this before the stamp... sorry) Msheflin (talk) 01:50, 24 September 2009 (UTC)


« Thelwall (1990) argues[12] that Arabic descriptions of a voiced pharyngeal fricative are incorrect and that Arabic varieties instead possesses a pharyngealized glottal stop ([ʔˤ]. Epiglottal realizations for /ħ/ and /ʕ/ have also been reported.[13]»

  • The reason is, maybe linguists mix between the official Arabic (Literary Standard Arabic, that is connected the Classical[Quranic Arabic]) that is different from the Arabic dialects spoken across the Arab League states.
  • In some Gulf states, their spoken dialect pronounces the Epiglottal /ħ/ and /ʕ/, While this isn't the standard pronunciation. However, the local dialect affects too much the pronunciation of the official Arabic (Literary Standard Arabic, that is connected the Classical[Quranic Arabic]). --Mahmudmasri (talk) 14:28, 25 October 2008 (UTC)
That may be true, but my understanding is that so far no one has said "it may be epiglottal or a pharyngealized glottal stop in other dialects but it's a true pharyngeal fricative in this one." — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:11, 25 October 2008 (UTC)

A Neutral Pronunciation?[edit]

In English for example the American mid-land accent is considered to the neutral English accent for all of North America. Does modern standard Arabic have an area that represents a neutral pronunciation? Egypt has media influence, Saudi Arabia has the heirtage, and Jordan a centering location, to any of them reflect a neutral pronunciation? —Preceding unsigned comment added by (talk) 00:56, 14 November 2008 (UTC)

  • The answer is: No. For Literary Arabic Standard: In Egypt; we pronounce ج /g/, in most of the Arab League countries it is pronounced /ʒ/ & in the Gulf, it is /d͡ʒ/. In Egypt, we don't pronounce short /i u/ at the beginning of a word nor in the middle, instead; it is /e o/ rarely if ever /ɪ ɵ/. In Egypt we never pronounce interdental sounds; such as /θ ð ðˤ/, instead /s z zˤ/. So, almost each country/part of the Arab league has its own accent in pronouncing the "Literary Arabic Standard".
  • For spoken languages/dialects of Arab League countries, there is no standard, because they are not standardized not recognized officially by governments. However, there are some unoffical efforts to standardize those dialects & there are prestigious dialects of major cities, such as Cairene accent as a standard for Egyptian Arabic for Egypt.... & so on.
The Saudi Arabian Arabic variants aren't the closest to the Literary Arabic Standard. The Syrian is. --Mahmudmasri (talk) 16:52, 7 January 2009 (UTC)
Despite the diglossia, where it is difficult to define what pronunciation is right or wrong in Arabic, there is still the Modern Standard Arabic, which has more or less clear pronunciation rules with some local variations. Anyway, we can stick to sourced materials and mention the local variations where needed. In formal situations, the accents of Al-Jazeera anchors may be of high prestige, IMHO, and may be used as a guide, everything else is dialects, discounting some minor differences. --Anatoli (talk) 21:57, 7 January 2009 (UTC)
The differences in the possible pronunciations of MSA are enough that defining the prestige form is still problematic. How, for example, is one to know whether the final vowel of a word like اقتباس is [a] or [æ]? On Al-Jazeera alone, I have heard this word pronounced with [a] by Uthman Uthman and [æ] by Faisal al-Qassem. Likewise, the short /u/ (Damma) when following ع is problematic as well. I have heard Yusuf Al-Qaradawi pronounce this as a simple [o], whereas Faisal al-Qassem pronounces it as something resembling a close mid front rounded [ø] (a pronunciation I seem to hear from the occasional Syrian.) Others on Al-Jazeera pronounce it as a Near-close near-back [ʊ]. Then there is the whole mess of the emphatic L which almost always deviates from what textbooks tell you. My point is that even the forms looked upon as exemplary prestige MSA display great variation. When I have the time, I'm going to see if I can make this article reflect that. Szfski (talk) 12:37, 23 February 2009 (UTC)
You wrote:

Would you please mention that this is (probably) Gulf-Arab phonology?

  • In Egypt, we pronounce those differntly; /kokæˈkoːlæ/, /læˈmuːm/, /dokˈtoːɾ/, /ˈʒoːn ~ ˈʒon/, /belˈʒiːkæ/, /sekɾeˈteːɾ/. (There is no /d͡ʒ/) Mahmudmasri (talk) 16:28, 7 January 2009 (UTC)

I will edit some, but as I mentioned before, it's the dialectal Egyptian pronunciation (although, it is considered prestigious it can't be called standard, if you know what I mean? /d͡ʒ/ and /ʒ/ variations are already described. As for vowels /a/, /æ/, the article groups them together as allophones, strictly speaking it's not correct but the article needs sources to change that. --Anatoli (talk) 21:57, 7 January 2009 (UTC)

No, the accent of Aljazeera isn't of high prestige. The supposedly standard MSA pronunciation isn't prestigious.  The Modern Standard Arabic spoken by Egyptians has [ɡ] not [ʒ] or [dʒ]. We were taught Arabic like that. You would never hear any Egyptian presenter pronounces [ʒ] or [dʒ]. Only [ɡ]. It is not a matter of what is prestigious in Egyptian Arabic, it's a matter of how MSA is taught & used in Egypt. Generally [ɡ, ʒ, dʒ] are all considered standard & correct pronunciation for ج‎. The [ɡ] pronunciation is even closer to Classical Arabic pronunciation ([ɡʲ]~[ɟ]), of which MSA was constructed from. So, speaking about what is correct & what is standard, [ɡ] must be the standard & consequently more correct. Some words in Arabic have gemination, with [dʒ] pronunciation, it is impossible to clearly audibly geminate the consonant. I've never heard [dʒdʒ], so as it is impossible by Arabic phonology rules, Classical or MSA (only, [ɡɡ] or [ʒʒ]). --Mahmudmasri (talk) 09:22, 15 September 2010 (UTC)
Yes, the Egyptian pronunciation of MSA is obviously prestigious (and as such unmarked) in Egypt, but the original question was where there was a regional variation of the pronunciation of MSA that is (explicitly or effectively) considered neutral in the Arabic world in general. As for how to pronounce a geminated affricate, you simply geminate the first part only, namely the stop: [ddʒ]. --Florian Blaschke (talk) 01:22, 26 March 2013 (UTC)

Ratraction of [a] to [ɑ], and vowel-emphasis in general[edit]

The article states that velar/uvular fricatives cause [a] to be retracted to [ɑ]. I must admit to skepticism on this matter. While this was probably true in classical Arabic (and is still the proscribed/preferred practice for Qur'anic recitation) one rarely hears Arabs do this when speaking MSA naturally unless their native dialect does it (and even then not always.) It is quite common, for example, to hear the word خائف (afraid) pronounced as [xaːʔif] or even [xæːʔif] when speaking MSA. This is even true for many southern Palestinians whose dialect has [xɑːjef]. Moreover, I have yet to find a single MSA user who pronounces a word like غداً (tomorrow) as [ɣɑdɑn]. It's a good bet that the source given for this particular statement (which, admittedly, I have not read) describes a theoretical prescription, rather than a practical description. Other wiki articles on Arabic (such as ʾIʿrab) give both prescriptive rules and a descriptive summary of what occurs in practice. It would seem logical that this article do likewise. My instinct is to say that the statement in question should be removed, or at least qualified.

In addition, the rules governing whether or not a given vowel such as [a] is to be retracted are (even allowing for the above complexity) incredibly intricate. Watson (1999) and Davis (1995) have demonstrated that vowel retraction is contingent on several factors, including the place and manner of articulation of other consonants in the stem and how those consonants are distributed. If I use such sources to add a section that gives a fuller description on the complexities of Arabic vowel-retraction, does that sound like a good idea?Szfski (talk) 10:41, 22 February 2009 (UTC)

That sounds like a great idea. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:35, 22 February 2009 (UTC)
Done. I tried to squeeze as much as I could into one paragraph (near the beginning of the section on vowels) without getting too technical for an encyclopedia article. Let me know if it needs fixing. Szfski (talk) 07:16, 24 February 2009 (UTC)
Good job! If you could include the page numbers in the citations, that would help them be complete and consistent. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 08:11, 24 February 2009 (UTC)
Also done. There's another issue that I forgot about, but which might be worth including in the article. There are some Arabic dialects wherein [æ] and [a] (or [æ] and [ɑ] depending on the dialect) have clearly split into two separate phonemes which occur in positions which cannot be predicted based on surrounding consonants. Such dialects include certain parts of the Cairene dialect bundle, as well as much of North Lebanese. (For example, North Lebanon has a minimal pair [ktɑ:b] 'write!' and [ktæ:b] 'book.' I can think of similar minimal pairs for Cairene and Palestinian Arabic as well. I'm pretty sure I can find a source for the former at the very least.) The article mentions these sounds as allophones (which they are in most dialects) but I would think mention should be made of the fact that this is not always so. What would be the best way to include this information? Should it even be included, seeing as how relatively marginal it is? Szfski (talk) 10:53, 25 February 2009 (UTC)
I think it's still relevent if it manifests itself in their MSA speech. If your sourcing doesn't say that, then we probably should keep it out of here for now.
Not to be a bugger about this, but could you also put the page numbers on the inline citations as well? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 14:41, 25 February 2009 (UTC)
Also, done. I'm not aware of any source describing this affecting their MSA speech. This touches on what I've come to see as a major problem hindering this article's usefulness. It seems a bit misleading to describe Arabic phonology primarily in terms of MSA. The actual situation in the Arab world isn't actually one of simple diglossia, so much as a kind of "polyglossia" where, outside printed and broadcast media, there is no real way to tell where "Dialect" ends and "Standard" begins. Using aspects of one's native phonology in one's MSA production is quite common precisely because Arabic speakers, more often than not, perceive the differing varieties of Arabic as variations on a theme, rather than separate languages/dialects/(insert neutral word here) with their own distinct phonological systems. Moreover, though, since MSA is not (and never has been) anybody's native language anywhere, I wonder how useful it will be if sources describing "Arabic phonology" are required to give specific reference to MSA. You must know that the way data on a language's phonology is gathered is by observing its native speakers and discerning patterns. One can do this for the spoken varieties of Arabic in a relatively straightforward manner. But in the case of MSA, the reverse occurs whereby the learner (whether it be an American university student or a Syrian school child) is presented with a set of arbitrary rules to internalize often by rote, in such a way that the person's MSA speech will be a secondary addition to their native phonology. Moreover, the hypothetical Syrian schoolchild is not taught that the two systems are entirely separate, and therefore a certain amount of blending occurs. It seems to me that focusing on MSA would make this article focus on a tiny (relatively idiosyncratic) subset of what actually comes out of Arabic speakers' mouths. The true phonological complexity and underlying patterns of Arabic speakers' speech, I believe, can best be accessed in reference to their native dialects. Imagine, for example, taking a group of Portuguese, Andorrans, Languedoquians, Frenchmen, Spaniards, Italians, Romanians and Sardinians, teaching them all a reconstructed version of late vulgar Latin, and attempting to develop a theory of "Romance phonology" and the local variations thereof by listening to them speak it. One would certainly be able to glean a large amount of interesting data. But it would hardly be the best heuristic for learning about their native phonetic inventory. Szfski (talk) 16:21, 25 February 2009 (UTC)
Unfortunately, this may be one of those situations where verifiability trumps truth. The best we can do is generalize MSA speech but incorporate the caveats according to sources. What you've included on /a/ retraction is a good example of what this might look like. What we can't do, however, is assume that speakers of dialect X with phonological feature Y will transfer this to their MSA speech; even if it seems obvious, this is an example of synthesis (this has been an issue at Talk:Non-native pronunciations of English). In addition, if we set this article up as the phonologies of standard Arabic, then that can really help distinguish this article from Classical Arabic. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:21, 25 February 2009 (UTC)
Oh that's not what I was advocating. I'm well aware that the way in which dialects affect one's MSA production is very poorly studied and, correspondingly, poorly documented. I'm not trying to remedy this with a Wiki article. My point (which I think got lost in the rambling) was that the there is a reason why Arabists who work on phonology are primarily concerned with dialects. It's because that's how people's native phonetic inventories are best accessed. I guess what I would do is enlarge this article so that it not only includes a general sketch of what MSA phonology looks like or is supposed to look like, but also the phonology of major dialect bundles (Levantine, Gulf, Maghreb etc.) with some space dedicated to the variations of each (supplemented by a link to an article on that dialect,) rather like what's going on with the article on English phonology. Szfski (talk) 03:46, 26 February 2009 (UTC)
I see what you mean. Since MSA pronunciation is not only an artificiality but also a bit of a myth, you think we should instead focus on the phonologies of the various dialects. However, I think something we could do instead (or in addition) is detail this phonetic information on the articles of the dialects themselves and, if we've got enough information and sourcing, have pages like Levantine Arabic phonology and Iraqi Arabic phonology. Maybe other people can weigh in on this. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:06, 26 February 2009 (UTC)
I should like to point out that the section on "Local Variations" describes features that are very much not part of the phonology of even the most adulterated forms of MSA. For example, no one, not even the most uncouth Cairene plebian, will pronounce ق as a glottal stop in MSA, since they are all quite capable of the uvular pronunciation, as evidenced by the fact that not even habitual users of the lowest registers will describe the Qur'an as a [ʔurʔaːn] rather than as a [qurʔɑːn]. The section is very clearly describing dialects as such, and not in terms of their influence on MSA. So, if this article's purpose is to focus on MSA, then why even have this section? If this section is the place to discuss dialect phonolog(y/ies), then why not go all the way and give actual, complete descriptions thereof? If this is too hefty, perhaps we could include a table with a list of Arabic words, and an IPA transcription of each word's realization in the various dialects. Szfski (talk) 10:18, 26 February 2009 (UTC)
Yes, I'm not sure why I added that section other than that maybe I wanted a place to put phonological information that I had sourcing for. We may have to move it somewhere else or incorporate it differently if we're to take my idea above seriously. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:15, 27 February 2009 (UTC)
If the article is going to go the rout of more narrow focus on MSA, then perhaps we might consider renaming it Phonology of Modern Standard Arabic and removing any and all statements that are extraneous to MSA with a re-direct from "Arabic Phonology." Though I'm thinking that this will lead to a massive Wiki-uprising on the part of Arab nationalists. Perhaps we should also then create a more macro-level article for Phonology of Colloquial Arabic. I could furnish much of the material for the latter, complete with a map of isoglosses. Szfski (talk) 18:37, 27 February 2009 (UTC)

(unindent) Let's put off the rename for now, though you're right that that may be a better title if we are to confine the scope at the same time that we talk about Colloquial Arabic. I like the Colloquial Arabic article idea and look forward to seeing it; my only suggestion is that, since it will be an article discussing multiple dialects that it should be titled Phonologies of Colloquial Arabic. Eventually, it may even become a hub for phonology articles of various colloquial dialects. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:13, 27 February 2009 (UTC)

For Szfski
قرآن‎ is pronounced by Egyptian Arabic speakers as: [korˈʔɑːn]~[qɔrˈʔɑːn] no [qurʔɑːn]. You would notice that the native vernacular of speakers affects the pronunciation of Modern Standard Arabic.
For Ƶ§œš¹
Yes, you are very true about the myth of the Standard Arabic. There is no standard. I wonder why was it called standard, since no one pronounces it with one standard, not even fanatic clerics!

For anyone[edit]

That is how Arabic letters are mostly pronounced by Egyptian Arabic speakers & are pronounced as that while reading in MSA. Notice also the Arabic diphthongs (/aj, aw/) are often pronounced [eː, oː] so as, Egyptian vowels are pronounced instead of the supposedly Arabic vowels even in MSA:

  • Bold transcriptions are the supposed standard taught in Egypt.

ا [(ʔ)æ, (ʔ)ɑ, ʔe-] name: [ˈælef] — ب [b] name: [be] — ت [t] name: [te] — ث [s] or [θ] name: [se] — ج [ɡ] only! no [ʒ] or [dʒ] name: [ɡiːm] — ح [ħ] name: [ħɑ] — خ [x] name: [xɑ] — د [d] name: [dæːl] — ذ [z] or [ð] name: [zæːl] — ر [r~ɾ] name: [re] — ز [z] name: [zeːn] not [zæːj] — س [s] name: [siːn] — ش [ʃ] name: [ʃiːn] — ص [s] & [sˤ] name: [sˤɑːd] or [sɑːd] — ض [d] or [dˤ] name: [dɑːd] or [dˤɑːd] — ط [tˤ] & [t] name: [tˤɑ] — ظ [zˤ] & [z] not [ðˤ] name: [zˤɑ] — ع [ʕ] name: [ʕeːn] (notice that Gulf Arabic /ʕ/ sounds more like [ʕˤ] — غ [ɣ] not [ʁ] name: [ɣeːn] — ف [f] name: [fe] — ق [k~q] name: [kɑːf~qɑːf] — ك [k] name: [kæːf] — ل [l] name: [læːm] — م [m] name: [miːm] — ن [n] name: [nuːn] — ه [h, -æ, -ɑ] name: [he] — و [w, o(ː), uː] name: [wɑːw] — ي/ى [j, e(ː), iː, -i, -æ, -ɑ] name: [je, ˈælef læjˈjenæ].

Modern Standard Arabic compared to Classical Arabic resembles Ecclesiastical Latin compared to Classical Latin. --Mahmudmasri (talk) 08:43, 15 September 2010 (UTC)


The famous drink Vimto does! And for Arabic marketing did and I guess does (I've never actually seen it, only really old-school ads). According to this ( it's used in Algerian and Tunisian... can't comment on that, and in the Jawi script for Bahasa Malaysia. I have seen it infrequently written in Arabic (I believe Classical because I think the examples I'm remembering are from al-Kitaab fii t3llum etc. the classic Tonsi book (love that guy)) for proper-noun loanwords (not like televison), but I can't cite that beyond the Ve_(Arabic) page. Msheflin (talk) 01:54, 24 September 2009 (UTC)

(Eastern) Arabian Peninsula & Iraqi Pronunciation[edit]

  • In the section Cairene: [ˈɡibnɐ] ('cheese') vs [ˈɡebnɐ] ('our pocket'). The last two are pronounced as [ˈɡebnæ], to the extent, there is a famous joke mocking this indistinguishable pronunciation [ˈgebnæ ˈgebnæ, ħɑtˤːeˈnɑˑhɑ fˈgebnæ, wæʔːˈæʕet ˈgebnæ] ‹gébnæ gébnæ, ħɑtˤtˤenɑ̄hɑ f-gébnæ, wæʔʔǽʕet gébnæ› may also be transcribed in Arabic script as: جيبنا جيبنه, حطّيناها فـ جيبنا, وقّعت جيبنه, meaning: we got cheese, we put it in our pocket, Cheese dropped. The study of Watson fails to precisely explain the Egyptian pronunciation, especially in vowels. In Egyptian Arabic (Cairene), you can never pronounce [i(ː)], [u(ː)] before two consonants, they are changed to [e], [o], or [ɪ], [ɵ]. The so called [ɐ] in word boundaries, is only in (East of) the Arabian Peninsula & Iraq. It is pronounced [æ] in Egypt, even when pronouncing Modern Standard Arabic. In Lebanon, it is [e].
  • In the section Consonants: In Arabic, they are called "mushaddadah" (strengthened), but they are not pronounced any stronger, just held longer. Between a geminate consonant and a pause, an epenthetic [ə] occurs. The pronunciation of the schwa [ə] in this case, which is as well pronounced when reciting Koran, is Arabian. I see that this study generalizing pronunciations specific to certain regions. --Mahmudmasri (talk) 05:14, 25 February 2010 (UTC)

ǧīm in table[edit]

Right now the cell that represents the voiced velar plosive pronunciation isn't included in the group of cells reflecting the variable pronunciation of ǧīm. How should we address this? We could maybe make the border between cells invisible, but I don't know how to do that. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 22:22, 24 December 2011 (UTC)

arabic phonetics[edit]

arabic phonetics are the three voices A o and E. they are not written, but deducted from reading. in English Basket is written in Arabic BSKT. BSKT could be read basket or beskuit, Baskat basakat, basakit, etc etc etc. to know the pronunciation (aka vowels of each letter ie the hidden vowel after the letter, you definitely need to know arabic grammar. there is no substitution for that. Even if the vowels were given to you you still need arabic grammer to understand what it means. a o e are very much exchangeable in a sec. don't rely on seeing a vowel and deducing that it will be repeated.

the basic grammar of ARABIC IS THAT ALL words are derived from three letter verb (root word). If their is no root word for a word then it is not arabic inorigin. this verb (root word) is in the past tense example: FSL a root word (means seperated ie past tense) the phonetics of the root word is always aaa Fsl reads Fasala

now lets come to a derivative word from FSL which is FSL but not pronounced Fasala but pronounced Fasl meaning the action of separating.


the subject ie the one who do FSL (the separator)

Fasel. FASL (pronounced as Fasel) is the standard subject derivative of the root word but is not the only one. fsal (pronounced Fassal) is another one (denote extra doer) Fsil (pronounced fasseel) is another one

it is the word place in the sentense that tells you it is a verb or a NOWN ETC HENCE ITS PRONUNCIATION THAT COMES INSTANTLY WITH THAT KNOWLEDGE.

If vowel letters are added to a sentense it will become impossible to read to the arabs. since arabic is the most vowelistic language (you cant put that in writing) there will be too many vowel letters. arabs hate consonants without being voweled. the too many vowels in arabic is a characteristic of arabic not shared by another language.

seeing a word like BSKT you have to expect 4 hidden vowels immediately. arabs like to say rather Basaketa than basket.

the letters V G Q P are not existant in arabic or semitic languaqges, even arabs can pronounce them they are rather piggy sounds.

to say these letters in ancient times means you are foreigner or a perv, very shameful.

chinese referred to the Russians in ancient times as Yezuhi. of course Russian word come from Rose (the red color) meaning the red people Rothschild means the red shield, the red color the color of the blood (hence the red flag was a testimonia that since blood are in humans and animals then humans are animals and should do away with religion, because blood is the only soul that comes out of dead hence the red flag of the communists) — Preceding unsigned comment added by Zahida2013 (talkcontribs) 19:07, 16 July 2013 (UTC)

Syllable structure[edit]

I notice that the section on syllable structure has no sources. I also notice that, if we are talking about MSA or CA, it is incorrect. In FusHa, semivowels are treated identically to consonants, and cannot come between either an initial consonant and the vowel or the vowel and a final consonant. The only exception is in pausal forms, where two consonants may come at the very end of a final syllable, but even then semivowels are treated no differently from other consonants.

2620:105:B00B:6801:FC96:A281:67EA:864A (talk) 21:45, 28 April 2014 (UTC)


Please add the orthographic version of the sample to the article. I can't do it, as I don't know the Arabic alphabet. Peter238 (talk) 13:29, 25 September 2014 (UTC)

Too many phonemes?[edit]

It seems to me like we should be using the Standard Arabic phonemes rather than trying to accommodate every single dialect. Our chart should be more similar to the International Phonetic Association's. --Monochrome_Monitor 18:28, 18 June 2015 (UTC)

Assessment comment[edit]

The comment(s) below were originally left at Talk:Arabic phonology/Comments, and are posted here for posterity. Following several discussions in past years, these subpages are now deprecated. The comments may be irrelevant or outdated; if so, please feel free to remove this section.

As far as it goes, this article seems to have a very good quality and is well-referenced. There may be minor issues, but these are not relevant before it is nominated for GA. However, any information on accent, suprasegmental characteristcs such as sentence intonation, are lacking. Thus, this article seems to be exactly C class. G Purevdorj (talk) 21:18, 19 July 2009 (UTC)

Last edited at 21:18, 19 July 2009 (UTC). Substituted at 08:08, 29 April 2016 (UTC)