Talk:Varieties of Arabic/Archive 2

Archive 1

Archive 2

Wiki voice

@A455bcd9: Regarding this edit:

Overall, native Arabic speakers believe that Levantine and Egyptian are the most widely understood dialects, and that Levantine sounds the most beautiful. The primary source (a web based survey involving 140 participants) doesn't make such a claim.

Levantine Arabic, and especially Palestinian Arabic, is the closest colloquial variety to Modern Standard Arabic. The primary source (experiments involving only 5 dialects) doesn't make such a claim.

Please read WP:VOICE and adjust the content accordingly. Thanks. M.Bitton (talk) 13:19, 5 July 2021 (UTC)

Hi,

Thanks for your message. So, I based this edit on the following quotes and sources:

• "Given the preference of both groups for teaching dialects, the study also looked at beliefs about which dialects were best for learners to study, which were most widely understood, and which were most beautiful. Both NS and NNS felt that Levantine dialects were most beautiful. In terms of the dialects most widely understood, NS tended to choose one dialect (usually either Egyptian or Levantine), while NNS chose the multidialectal combination of Egyptian and Levantine, and there was a significant difference between the two groups on this question." (source, a web survey with 444 participants, 140 final participants after various exclusions)
• "The results are informative and indicate that Levantine dialects are very similar to each other and furthermore, that Palestinian appears to be the closest to MSA." (source, they compare MSA to Algerian, Tunisian, Palestinian, Syrian, Jordanian, and Egyptian)
• "Particularly, PAL is closest to MSA than other dialects are (Table 3)." (source, a study between MSA, two Algerian dialects, Tunisian, and two Levantine dialects)

I'm not sure I totally understood what you meant by "adjust the content accordingly", please let me know if this seems better:

• "According to Trentman & Shiri (2020), and based on a web survey involving 140 participants, native Arabic speakers believed that—among Levantine, Egyptian, North African, and Gulf Arabic—Levantine and Egyptian are the most widely understood dialects and that Levantine sounds the most beautiful."
• "According to two scientific papers, Palestinian Arabic is—among Algerian, Tunisian, Palestinian, Syrian, Jordanian, and Egyptian dialects—the closest colloquial variety to Modern Standard Arabic."

Best, A455bcd9 (talk) 15:18, 5 July 2021 (UTC)

Another useful source including Gulf Arabic:

• "It is not quite clear which dialect is closest to MSA but from the table we can observe that LAV is most likely to be miss-classified as MSA more than the other dialects in this corpus." (LAV = Levantine, "The dataset used in our research covers four major Arabic dialect groups: Egyptian (EGY), Levant (LAV), Gulf (GLF), and North African (NOR).", source)

A455bcd9 (talk) 15:34, 5 July 2021 (UTC)

Since the two sources are primary and the claims they make are not facts that we can state in Wikipedia's voice, we have to attribute the opinions to their authors (e.g., according to so and so) and put them into context (e.g., based on a web based survey) to make sure the readers understand what they are. Best, M.Bitton (talk) 15:42, 5 July 2021 (UTC)

Thanks for the explanation. So what do you think about the first sentence I suggested? Regarding the second one, here's a second version:

• "Three scientific papers concluded that Levantine dialects (and especially Palestinian) zere the closest colloquial varieties to Modern Standard Arabic: Harrat et al. (2015, comparing MSA to two Algerian dialects, Tunisian, Palestinian, and Syrian), El-Ha et al. (2018, comparing MSA to Egyptian, Levantine Gulf, and North African Arabic), and Abu Kwaik et al. (2018, comparing MSA to Algerian, Tunisian, Palestinian, Syrian, Jordanian, and Egyptian)."

I'll implement the suggested modifications if it's okay. A455bcd9 (talk) 16:53, 5 July 2021 (UTC)

The first sentence seems okay (I only glanced through the source). For the second and unless I'm mistaken, the sources (again, I only glanced through them) seem to be about natural language processing. To be honest, as it stands, I'm not sure it belongs in this article (regardless of how it's presented). M.Bitton (talk) 17:25, 5 July 2021 (UTC)

From my understanding, the authors used various natural language processing techniques to measure the lexical distance between dialects. This appears in the summary of A Lexical Distance Study of Arabic Dialects: "We conduct a computational cross dialectal lexical distance study to measure the similarities and differences between dialects and the MSA. We exploit several methods from Natural Language Processing (NLP) and Information Retrieval (IR) like Vector Space Model (VSM), Latent Semantic Indexing (LSI) and Hellinger Distance (HD), and apply them on different Arabic dialectal corpora." Or in the conclusion of Cross-Dialectal Arabic Processing: "We perform different experimentations in order to study the relationship between MSA and dialects on one hand and cross-dialects on the other hand. For this, we calculated the overlapping between each pair of vocabularies. We then calculated the distance between the distributions of each pair of languages in order to measure which language is closer to which one. The carried out results are consistent with the fact that Middle-East Arabic dialects are closer to MSA than those of the Maghreb."

So we could add:

• "Three scientific papers concluded, using various natural language processing techniques, that Levantine dialects (and especially Palestinian) were the closest colloquial varieties, in terms of lexical similarity, to Modern Standard Arabic: Harrat et al. (2015, comparing MSA to two Algerian dialects, Tunisian, Palestinian, and Syrian), El-Ha et al. (2018, comparing MSA to Egyptian, Levantine Gulf, and North African Arabic), and Abu Kwaik et al. (2018, comparing MSA to Algerian, Tunisian, Palestinian, Syrian, Jordanian, and Egyptian)."

A455bcd9 (talk) 17:43, 5 July 2021 (UTC)

Looking back at the first sentence and it's supporting primary source, it appears that its conclusion and the most important aspect of it "the question of teaching" have been ignored, while a small part of what of it is being analysed. Is there a reason for that? M.Bitton (talk) 18:33, 5 July 2021 (UTC)

There's no reason for that. The article is quite broad, as its title ("The Mutual Intelligibility of Arabic Dialects") and subtitle ("Implications for the Language Classroom") suggest, it studies first the mutual intelligibility of Arabic dialects (hence the questions about which one is the most widely understood) and then "Pedagogical implications are discussed" (last sentence of the abstract). The question in the survey was "Which  Arabic  dialect(s)  do  you  think  are  the  most  widely understood among Arabic speakers?" (page 18/31), so there was no mention of "the question of teaching" in the question asked. That's why I thought it was not necessary to mention the rest of the article. What do you think? A455bcd9 (talk) 18:58, 5 July 2021 (UTC)

I think there is a reason why we're told to be careful when using primary sources. Here for instance, the conclusion (the essence of what the source is about) is ignored.

For the second sentence, I think something like this should do: "Some analytical studies, using various Natural language processing techniques, suggest that Levantine dialects (and especially Palestinian) are the closest colloquial varieties, in terms of lexical similarity, to Modern Standard Arabic". See what you think. M.Bitton (talk) 19:05, 5 July 2021 (UTC)

I don't understand your first point about "the conclusion". What do you mean?

For the second sentence, you want to replace the list of papers by "Some analytical studies"? I'm fine with that but I was just afraid of MOS:WEASEL. And if we want to be accurate it's interesting to mention that not all dialects were compared in each papers. But the compared dialects can be added in the reference(s) or in a note. A455bcd9 (talk) 19:12, 5 July 2021 (UTC)

I suggest we lose the first sentence (a web based survey involving 140 participants is hardly worth a mention). I agree with your suggestion for the second. M.Bitton (talk) 19:36, 5 July 2021 (UTC)

I implemented the changes: removed the first sentence about the "most widely understood dialects" and edited the second one about the closest dialect to MSA. I also moved the second sentence from the introduction to a dedicated subsection because it was long for the introduction and not worth it given that it was based on primary sources only.

Regarding the first sentence, I think it could and should be mentioned in the article because it's the only paper I found studying mutual intelligibility (and perception of it) of Arabic dialects that well. I created a dedicated section about mutual intelligibility that I'll try to expand in the future... A455bcd9 (talk) 07:23, 6 July 2021 (UTC)

Proposal to Remove Maps

I hope you're all well. A week ago, I proposed the removal of the two maps here found here on the article Arabic for reasons of accuracy & poverty of sourcing. That discussion is in Talk:Arabic. I would like to remove the maps from here, as well. For anyone who's following this page, I would like to invite you to join the conversation at Talk:Arabic#Proposal_to_Remove_Two_Maps, which at this point is an echo chamber of mostly me. So far, there is no substantive argument against removing the maps, but I think it's possible that will change today. There has also been a conversation at Noticeboard#Probable_WP:CIRCULAR about whether or not some of the metadata on one of the maps constitutes circular citation. That is possibly also of interest to people following this page, but is a narrower question. Pathawi (talk) 17:10, 25 August 2022 (UTC)

This didn't provoke any disagreement in a week, so I removed the maps. Pathawi (talk) 10:53, 1 September 2022 (UTC)

Regarding "standard" MSA pronunciations

Mahmudmasri regarding your recent change adding /d͡ʒ, ʒ, ɟ, ɡ/ to MSA pronunciation of ج:

Phonology of MSA can be a tricky thing to pinpoint, because speakers of MSA tend to be influenced by their dialect to some extent. Pronunciations such as ظ as /zˤ/ or ذ as /z/ in Egypt and the Levant, or ض as /ðˤ/ in much of the Arabian Peninsula, are usually considered nonstandard. These pronunciations are not included in the chart under MSA. The ج as /g/ pronunciation is likewise not the standard pronunciation. High surv (talk) 13:51, 30 August 2022 (UTC)

This is a misleading discussion and presumptuous. It was from your misunderstanding what constitutes standard and informal. Quoting you "ذ as /z/ in Egypt", actually you mixed between the informal pronunciation for this consonant with the standard /ð/. ج has indeed many standard pronunciations. The consonant chart of MSA has the consonant in four blocks and a note underneath. Furthermore, if westerns prefer the fricative pronunciation, that doesn't make it the only standard. --Mahmudmasri (talk) 21:21, 31 August 2022 (UTC)

I did not "mix" between the informal pronunciation. They are both colloquial pronunciations that some speakers use when speaking MSA. ض as /ðˤ/ is very common in the Gulf region when speaking MSA.

The burden of proof is on you to provide a source, which you have not yet. High surv (talk) 09:22, 1 September 2022 (UTC)

This section has some formatting issues. My understanding from being in Cairo is that most educated people do understand the realisation [g] to be the appropriate Fuṣḥā realisation of جيم for Egypt. It is the news broadcasting standard. Many educated people will tell you that the variation between [g] & [ʤ] realisations is a product of which tribe settled where. I don't think this is a confusion of diglossic levels. (On the other hand, in Qur'anic recitation, Egyptians realise جيم as [ʤ], right?) For an appropriately reliable reference, one might take Karin C. Ryding's A Reference Grammar of Modern Standard Arabic (Cambridge University Press, 2005). On page 14, in a table of Arabic consonants, she writes the following for جيم:

There are three standard regional variants:
(a) voiced alveopalatal affricate; /j/ as in "jump";
(b) voiced alveopalatal fricative (zh): as the /z/ in "azure" or the medial sound in "pleasure";
(c) voiced velar stop; /g/ as in "goat";

She then adds in a footnote:

The variations are essentially as follows: the first is more characteristic of the Arabian Peninsula and Iraq, the second more Levantine and North African, and the third specifically Egyptian and Sudanese pronunciation. Occasionally, a mixed pronunciation of jiim is found, with one variant alternating with another, especially /j/ and /zh/.

I rather hate this for multiple reasons, but it's a reputable MSA descriptive grammar which makes a claim that's partially true. (The part I'm balking at: [g] is not a standard realisation of جيم in most of the Sudan—not in Khartoum, not in the East, not in Dar Fur. If it's standard anywhere, I don't know about it. In the Sudan, جيم is generally realised [ʤ], tho [g] or [dʲ] happens in some clusters [eg, مجتمع as [mudʲtamaʕ] is not uncommon].) Faruk Abu-Chacra's Arabic: An Essential Grammar (2d ed, Routledge, 2017) makes a similar claim (6–7):

This letter has three distinct pronunciations depending on the dialectal background of the speaker:
a) In Classical Arabic and the Gulf area, as well as in many other places in the Arab world, it is pronounced as a voiced palato-alveolar affricate as the j in 'judge', 'journey', or g in Italian 'giorno'.
b) In Lower Egypt (Cairo, Alexandria) it is pronounced as a voiced velar stop g, as in English 'great'.
c) In North Africa and the Levant it is pronounced as a voiced palato-alveolar fricative /ğ [this probably should be ‹ǧ›], ž/ in the English 'pleasure', and as j in the French 'jour'. This one is the most common and will be used in this book.

Abu-Chacra is explicitly describing "modern literary Arabic". Pathawi (talk) 13:08, 1 September 2022 (UTC)

The ommission of the Badawi-Carter-Gully Modern Written Arabic: A Comprehensive Grammar is not meant to conceal anything: They explicitly (11) choose not to deal with the issues of pronunciation. Pathawi (talk) 13:13, 1 September 2022 (UTC)

The Qur'anic recitation is [ʤ] as far as I know.

It does seem that [g] is common in Egyptian news. But I will reiterate that if [g] is to be included in the MSA chart, then the same thing applies to realization of ض as [ðˤ] in the Arabian Peninsula, which is very common in news, media, and religious sermons as well.

In both of those cases, many speakers outside of that region would consider the pronunciation incorrect. The contrary is not true, as ج as [ʤ] and ض as [dˤ] are still considered correct in both Egypt and the Peninsula, i.e. there is a canonical pronunciation that MSA speakers everywhere can agree on.

So I believe it's a question of determining the chart's scope. High surv (talk) 15:44, 1 September 2022 (UTC)

I believe that the Qur'anic recitation form is [ʤ] in Cairo, as well (altho I have to say: I'm really not sure if it's [ʤ] everywhere… in places where the usual formal realisation is [ʒ], what's the standard for Qur'anic recitation? I have no idea), but Qur'anic recitation isn't MSA. But the real issue for us as Wikipedia editors is reliable sourcing, right: Verifiability, not WP:TRUTH. I think it can safely be said that an MSA realisation of جيم with three different values is attested in at least two major grammars, neither of which says 'All Arabic-speakers agree on canonical [ʤ], but there's regional variation.' Ryding doesn't give the same account for ضاد—for her it's only a 'voiced velarized alveolar stop' (14). Here's what Abu-Chacra has: 'This is also an emphatic consonant, classified as a pharyngealized voiced alveolar stop. Arab phoneticians and reciters of the Koran recommend its pronunciation as a counterpart to د /d/ (8). In current use in many dialects it is, however, also pronounced the counterpart of ذ /ḏ/ (9), somewhat similar to the sound of th in the English 'thus…' (8–9). So, I don't know how best we should represent that in the article, but I do think there's a solid case for recognising [g] as a realisation of جيم in MSA drawing on multiple reliable sources. Fricative ضاد perhaps we'd have to list as an alternative, given Abu-Chacra's wording. Maybe another reliable source says something different? Pathawi (talk) 16:04, 1 September 2022 (UTC)

&, also, I suppose: Whatever we do should use the sources for what they say, rather than what we want to take from them or what we're interested in. So to give a fuller story: Ryder notes two standard variants of ظاء: /ðˤ/ & /zˤ/. For her description, that would put it on the same level as جيم as having different standard MSA (redundant, I know) variants. She does not, however, mention variation for ضاد. Abu-Chacra, however, describes ظاء as /ðˤ/, with dialect variation as /dˤ/ & /zˤ/. This makes the variation dialect-influenced—like his account for ضاد—& not basic—like his account for جيم. Neither grammar lists variation for any other consonantal phoneme. Pathawi (talk) 16:32, 1 September 2022 (UTC)

I'll include a couple more sources:

Alhawary's Modern Standard Arabic Grammar lists one pronunciation for each consonantal phoneme, with no variations (though he notes regional variation in consonant influence on vowels). The following pronunciations are listed:

• ج as a voiced alveolo-palatal fricative "produced with the tip of the tongue against the alveolar ridge and the middle of the tongue raised against the palate in this sequence: as "j" in judge".
• ض as a voiced alveolar stop emphatic.

David Cowan's Introduction to Modern Literary Arabic notes three variations in consonantal phonemes:

"The following letters are pronounced more or less as in English: ب = b, ت = t, ث = th as in 'think', ج = j (in Lower Egypt as g in 'get')..."

"ض is an emphatic 'd', or in certain countries (e.g.. Iraq) 'th' as in 'that', pronounced with the tongue pressing hard against the edge of the upper teeth with the tip protruding."

"ق is a guttural 'k' pronounced from the back of the throat. Colloquially, with the exception of Lower Egypt and some parts of Syria and Palestine where it generally becomes a glottal stop, it is almost universally pronounced as 'g' in go." (this is of course false)

Schulz's A student grammar of modern standard Arabic does not note any variations. The following pronunciations are noted:

• ج like g in gentle
• ض like d articulated with emphasis

High surv (talk) 00:25, 2 September 2022 (UTC)

It was inappropriate from High surv to revert without reaching a consensus, that's apparently against his claims, but I'll come back when I have time. If anyone claimed the plosive gim pronunciation is nonstandard, he knows nothing about Arabic, standard, and how it's used. And actually, this plosive pronunciation is also the standard by coastal Yemenis all the way to the neighboring region of coastal Oman. The Sudanese variant is mostly [ɟ] or [d ͡͡ʒ]. Mahmudmasri (talk) 18:18, 1 September 2022 (UTC)

The problem with /d͡ʒ, ʒ, ɟ, ɡ/ in slashes is that no regional (standard or basilectal) variety has four different segments for ج in its inventory, and that is what this notation says and why it is misleading. In every variety, there is only phoneme that is variously realized as [d͡ʒ ~ ʒ ~ ɟ ~ ɡ] depending on location. The choice of the phoneme symbol is arbitrary (it could even be /☎/ as suggested by Mark Hale in a different context); if you feel that /d͡ʒ/ gives undue weight to a reading that is commonly taught to non-Arabs as "Standard" but actually shouldn't have a monopoly on that label, we could simple change all the sounds in the table in the MSA column to phones in brackets, and naturally have [d͡ʒ ~ ʒ ~ ɟ ~ ɡ] in the ج-row. The key information in the table isn't the MSA pronunciation anyway, but the regional pronunciations; MSA is just a reference point. –Austronesier (talk) 19:30, 1 September 2022 (UTC)

What if we just do away with MSA altogether? It's a pedagogical term that only has limited utility when talking about the actual Arabophone world anyhow. You're absolutely right that the MSA isn't the point of the table. Why not just use the Wikipedia strict Romanization? This could get rid of what's ultimately a distraction. Pathawi (talk) 00:12, 2 September 2022 (UTC)

(Outdenting because of the table.) This could look like:

		ق q	ج j	ث ṯ	ذ ḏ	ظ ṭ	ض ḍ
Levantine	Jordanian (Western Amman)	[g], [ʔ]	[d͡ʒ]	[t], [s]	[d], [z]	[dˁ], [zˁ]	[dˁ]
	Syrian (Damascus)	[ʔ]	[ʒ]
	Lebanese (Beirut)
	Palestinian (Jerusalem)

& so on. (Flipped the axes for readability.) The dialect values, after all, don't derive from modern media varieties. If someone really wanted, we could throw in historical dialects as well, drawing on the work over the past couple decades on Northern Old Arabic, Old Hijazi, &c. If there's a desire, we could include a row on "MSA", as it is a set of varieties.

In general, this section would benefit from sourcing. There's almost none.Pathawi (talk) 16:52, 2 September 2022 (UTC)

References

Map

I added a new map to show the geographical distribution of Arabic varieties. Special:Contributions/41.232.227.205 removed it and commented:

• "removal of a non-academic unsourced and unofficial map, it has many factual errors. Unsourced."
• "No it's not accurate or sourced for many countries. Egypt map is incredibly misleadingly, you highlighted literal "dots" as huge populations, real and actual numbers are absloutely nothing like what this map shows. Incredibly manipulative map, and a similar map got removed by an Egyptian linguistic here before."

In the specific case of Egypt mentioned by the IP, the source is Ethnologue's map of Egypt here. Ethnologue's map seems to agree with other sources I could fine (although defining strict dialect borders is always a challenge).

So I suggest we add the map back. (poke @Pathawi for feedback). A455bcd9 (talk) 10:52, 22 November 2022 (UTC)

The map is definitely sourced. I don't mean to split hairs, but it seems that what the IP editor wants to say is that it's not reliably sourced. I don't love Ethnologue, but if you want to have that conversation article edit comments aren't the place for it: WP:RSN would be a place to start. In the meantime, the map should probably remain while this is handled in Talk. WP:PRESERVE: 'Rather than remove imperfect content outright, fix problems if you can, tag or excise them if you can't.'

Minor thing that seems to be a misunderstanding: The dots aren't population centers—the map does not indicate population in any way. The legend says 'Dotted area fill: speakers of this variety are mixed with other varieties in the area'. Pathawi (talk) 13:39, 22 November 2022 (UTC)

Ah you're right @Pathawi, the IP thought that the dots were population centers. (I wonder whether we should open a topic in WP:RSN regarding Ethnologue to hear the various opinions from the community and hopefully settle the issue...) A455bcd9 (talk) 13:43, 22 November 2022 (UTC)

To be clear, opening a conversation on RSN won't settle the issue: It will just initiate a conversation. Because Ethnologue is cited so many times in Wikipedia (9,121 pages), & because it provides a kind of data that people want to exist (whether it's reasonable or not), I think that having such a conversation is a huge project. Pathawi (talk) 13:51, 22 November 2022 (UTC)

I don't know if it's such as huge project, for instance Uanfala mentioned here that Ethnologue has already been reviewed by WikiProject_Languages. A455bcd9 (talk) 13:55, 22 November 2022 (UTC)

The above links are just to conversations—suggestions from particular interested groups of Wikipedia editors. I like the gist of this conversation (because it says what I want it to, obviously), but there's nothing binding about it. A conversation that settled the issue would be a bigger deal. Pathawi (talk) 14:03, 22 November 2022 (UTC)

I like the gist of this conversation as well. I'm also trying to find what other reliable sources say about Ethnologue and I'm adding these opinions here. A455bcd9 (talk) 14:37, 22 November 2022 (UTC)

@41.232.227.205 reverted and commented: "Again, the EGYPT MAP IS FALSE. @Mahmudmasri removed that map before, an actual native speaker. Kindly don't put this FALSE Egypt map again, not only the distribution is false but also missing many Dialects like Fallahi, Menoufi and many others, not to mention covering EMPTY areas with false colors."

I don't think Mahmudmasri removed this map before by the way. But it would be great for the IP to engage in the discussion here... A455bcd9 (talk) 14:55, 22 November 2022 (UTC)

Why

An example of a realistic map on native Egyptian Arabic speakers, even though I'd say it is a little bit more spread westwards further along the northern coast.

I haven't seen the map before to have removed it as claimed by an IP editor. However, having just seen it, I can confirm that it's at least false about Egyptian dialects, for the following reasons:

Egyptian Arabic (the traditionally northern dialect) is much more spread. It ought to extend to Middle Egypt, fading gradually in the south and around the Delta. The Southern Sinai part should be mixed Egyptian Arabic and the traditional Bedawi dialect. I'd say it's an Egyptian Arabic majority today.

Where are the Nubian languages in Egypt and Sudan? Is the map ideologically Arab nationalist, denying any other minority or majority languages spoken across Northern Africa and Western Asia?

Whoever made the map never went to Egypt, never consumed its media, and doesn't comprehend any Arabic.

I was, anyway, impressed by the map that seemed to give a very quick glimpse for Arabic dialects. --Mahmudmasri (talk) 15:14, 22 November 2022 (UTC)

The map is purely a map of Arabic dialects, throughout the Arabophone world—not a map of all the languages of the region. That's why the Nubian languages, Beja, & Siwi don't appear in Egypt, as well as hundreds of other languages thruout the region. It's not Arab nationalist: It's about Arabic.

I have spent a lot of time in Egypt & do speak Arabic. I roughly agree with your assessment of the what an ideal map should look like. The issue for Wikipedia, of course, is that it has to be based on reliable sources—not your or my (unpublished) personal knowledge. The map as it stands for Egypt is based on a reliable source. But a reliable source can be wrong (the good old 'verifiability, not truth' adage). Egypt's dialects are among the best documented in the Arabic-speaking world. Unfortunately, I have not seen a dialect map for Egypt that could be used in a broad regional overview like this: The multi-volume atlas by Woidich & Behnstedt is wonderful, & represents the complexity of dialect boundaries, but it gives no map that would allow a reader to say 'this is Cairene, this is Saʕīdi, &c'. Do you know of good sources that provide the kind of map for Egypt you'd like to see? Pathawi (talk) 15:24, 22 November 2022 (UTC)

Then, call it the map of historical or traditional dialect areas, because by no means is it representative of the modern days. I also remember that Ethnologue mentioned that Egyptian Arabic is spread outside the confined Delta region. A note on what the IP mentioned about rural or other subdialects that they are either included in Egyptian Arabic or Saidi Arabic. --Mahmudmasri (talk) 15:30, 22 November 2022 (UTC)

We've got to follow reliable sources. Not personal knowledge. There's the rub. There's plenty on Wikipedia that each of us who has some kind of area expertise knows to be wrong, but working from personal knowledge leads to travesties like the previous Arabic map. At present, I don't know of a better dialect map of Egypt than Ethnologue's to direct the map's creators to. The source describes this as the dialect situation in Egypt. (Ethnologue doesn't verbally describe the geographical extent of "Egyptian Spoken Arabic" in the current version—it just says "widespread". It provides the map that is the source for this one as a representation of the language situation in Egypt.) Pathawi (talk) 15:38, 22 November 2022 (UTC)

User:A455bcd9, if you seemed to be the one who created the map, please make sure not to confuse traditional native dialect regions with the actual wnd current state. What is traditionally known as Said in Egypt isn't anymore all Saidi speaking, particularly not far from the Delta region which includes Cairo, the capital. I know many natives from Middle Egypt who speak no Saidi Arabic at all! The expanded northern coast west of Alexandria also speaks barely any local dialects, only in hinterland villages. --Mahmudmasri (talk) 15:21, 22 November 2022 (UTC)

@Mahmudmasri: thanks for your feedback. User:Goran tek-en made the map based on sources I provided (mostly Ethnologue's maps), see the request here.

I'm not confusing anything, this is what the source map says. According to Ethnologue's guidelines:

• "we typically use the convention of showing a language in an area if at least 25% of the population in that area speak the language fluently as an L1. However, a minority language may be shown on a map even if the speakers do not comprise more than 25% of the population in any given area. This exception is made when the minority language is indigenous to that area only or, when there is historical significance for the language in that region."

So we could indeed change the legend of the map on Wikipedia as "Historical distribution of the Arabic varieties". What do you think? A455bcd9 (talk) 15:40, 22 November 2022 (UTC)

@Mahmudmasri, the map you mentioned is unfortunately pure WP:OR, no source is mentioned at all. A455bcd9 (talk) 15:42, 22 November 2022 (UTC)

That's really going to be a problem. Historical distribution as of when? What about the dialects of Arabic no longer in existence that no longer appear on this map? Pathawi (talk) 15:52, 22 November 2022 (UTC)

I've quickly checked the sources provided under the map and what was available for me to see had nothing about Egypt or barely anything about dialects rather than "regions". On the other hand, captioning the map as a historical or traditional should be a wise wording, since it actually deals with traditional areas. But, are you really sure, Suez is Saidi? It has never been part of the cultural or historical Saidi. The map indeed has errors. --Mahmudmasri (talk) 15:54, 22 November 2022 (UTC)

The map is based pretty directly on Ethnologue's map for Egypt. At the end of the day, the core issue is reliable sourcing. I can agree with you on everything you say, but a dispute over content has to be based on reliable sources—not editors' personal knowledge. Reliable sources can be wrong, & when multiple reliable sources conflict, we can begin to reason about which to prefer. At present, only one reliable source has been presented. Pathawi (talk) 16:08, 22 November 2022 (UTC)

@Mahmudmasri, there's the source about Egypt. Do you have access to it? (I have an Ethnologue account so I don't know if you can see this map, if not please let me know). A455bcd9 (talk) 16:21, 22 November 2022 (UTC)

Wow! According to this map, Cairo is part of the Said! So, Cairene/Egyptian Arabic is falsely claimed Saidi Arabic? Is Saidi Arabic the dialect of the capital?! It seems, Ethnologue contradict themselves saying Egyptian Arabic is in Cairo, but claiming it is not in the map. For the context: Upper Egypt, Middle Egypt, Cairo, Suez Governorate, Egyptian Arabic --Mahmudmasri (talk) 16:28, 22 November 2022 (UTC)

@Mahmudmasri: no, look more closely (I agree it's not clear...) they say that Cairo is an "area with multiple languages". (native Cairene speakers + Saidi migrants + Greek minorities I guess?). But I agree the map isn't clear there and they'll fix it in the next edition. A455bcd9 (talk) 16:36, 22 November 2022 (UTC)

Well, as Cairo is the cultural pole, it has people from all over the country, but that's not a valid argument for considering Cairo part of Upper/Southern Egypt and definitely Suez region belongs neither to Middle Egypt (and by extension Southern Egypt), nor to Lower Egypt.

People inhabited the Nile Valley where the traditional cultural divisions were made up, lower, middle, and upper Egypt. With time, those belonging to modern Lower Egypt expanded past the traditional area, namely Delta+Giza/Cairo and around.

I'm leaning to giving up on this discussion, because it's not the first time on Wikipedia, regarding Arabic language and dialects. Honorable(!) mentions: syllable structures, phonology, and "standard" Arabic. They are very poorly studied/documented, and where there are studies, they cite highly generalizing, opinionated, often faulty (idelological) sources.

By the way, the older map was more correct regarding Arabic dialects in Egypt. We haven't asked the guy who made the Egyptian Arabic map about his source(s). Just saying... --Mahmudmasri (talk) 18:09, 22 November 2022 (UTC)

@Mahmudmasri: we did contact the guy who made the previous map and he's never answered... So we decided to do a new one, using the sources we could find. And I agree they're not perfect. But the previous map showed the Nubi language as spoken in Egypt (vs Kenya and Uganda in reality...) and the Somali language as an Arabic dialect. That's a different level in terms of mistake. There's a difference between an issue of borders (defining a border cannot be done with certainty) and these huge mistakes. That's why I consider the new map to be way better overall (and sourced...). A455bcd9 (talk) 18:23, 22 November 2022 (UTC)

Many Somalis speak it due to centuries-old ties with the Arab world, the far-reaching influence of the Arabic media, and religious education. Cited by three sources, quoting Demographics_of_Somalia#Languages. There is a form of Somali Arabic. It's different from Somali language. --Mahmudmasri (talk) 08:15, 23 November 2022 (UTC)

1. I don't think there are sources on "Somali Arabic" as a language.

2. I don't think there are reliable sources showing where "Somali Arabic" is spoken either.

3. The map's legend says "Somali". A455bcd9 (talk) 10:41, 23 November 2022 (UTC)

I did in fact contact the maker of the prior map regarding errors in that map. They never responded—nor did they respond to most other questions about errors in that map over a period of years. (They did respond to a couple early critiques, from which it's clear that the map is the result of a great deal of speculation & sometimes understandable-yet-faulty inference.)

The fundamental issue is sourcing. Our personal knowledge isn't a reliable source for Wikipedia purposes. I know I keep saying this, but it is really the core issue, & the criticisms that are being levelled against this map are not providing other reliable sources. (I want to add that I am not wedded to this map as it is. I am arguing against its being featured elsewhere. What matters to me is that if we're going to have dialect/language maps on Wikipedia, that they be reliably sourced. While I wouldn't be surprised if there were some sourcing errors in a project of this scale, I believe that A455bcd9 and Goran tek-en have done an incredible amount of work to make sure that this map is sourced. That's why I'm defending its inclusion here.)

I don't think there's a legitimate issue here until someone presents a better source. Pathawi (talk) 18:30, 22 November 2022 (UTC)

We are not here to praise each other. I didn't cite my personal knowledge. I provided you with the definitions of what constitutes Upper, Middle, and Lower Egypt, which contradicted the latest faulty map. If you need sources, then you should look for sources of the aforementioned terms. I think we have brains and can read and understand why the Ethnologue map on Arabic dialects is just plain wrong by all defintions. Ethnologue contradics itself in its geographical borders on Arabic dialects in Egypt which makes it an unreliable source (map) for this particular topic. And who made that map? Was he consulted in the process? --Mahmudmasri (talk) 18:51, 22 November 2022 (UTC)

I don't understand what you think praise has to do with anything. WP:DON'T PRESERVE is policy that explains when material should possibly be removed: unsourced material, own research, things that fall into WP:NOT, material that gives undue weight to one viewpoint in a debate, libel, nonsense, vandalism, copyright violation… Unless we're dealing with personal biographies, that's it. This isn't covered by any of those. You've got to bring a better source, or there's no issue.

I don't agree with Ethnologue, but I don't think you've succeeded in making the case that it contradicts itself. I can't see how it contradicts itself.

& yes, I contacted the maker of the map you're referring to several months ago. They never responded. They have also not responded to many other questions about their numerous maps. The maps seem pretty clearly to be WP:OR. No sources at all are listed for the map in question. Pathawi (talk) 19:14, 22 November 2022 (UTC)

Pathawi, don't be selective. I referred to basic definitions which make me feel you simply don't know them and sticking to a source that obviously did a mistake. Ethnonogue's job isn't to produce maps, I think. There are no sources for the Arabic dialects map. All you'll have to do is get the definition of each dialect, where its spoken natively as L1 and apply that to the map.

• Egyptian Arabic, locally known as Colloquial Egyptian, or simply Masri is the most widely spoken vernacular Arabic dialect in Egypt. It is part of the Afro-Asiatic language family, and originated in the Nile Delta in Lower Egypt. ... The term Egyptian Arabic is usually used synonymously with "Cairene Arabic", which is technically a dialect of Egyptian Arabic. The country's native name Maṣr, is often used locally to refer to Cairo itself. As is the case with Parisian French, Cairene Arabic is by far the most prevalent dialect in the country.
• Lower Egypt is the northernmost region of Egypt, which consists of the fertile Nile Delta between Upper Egypt and the Mediterranean Sea, from El Aiyat, south of modern-day Cairo, and Dahshur.
• Upper Egypt is the southern portion of Egypt and is composed of the lands on both sides of the Nile that extend upriver from Lower Egypt in the north to Nubia in the south. --Mahmudmasri (talk) 02:23, 23 November 2022 (UTC)

  From praise to selectivity, I still don't know what you want to say… I don't know what you think I'm being selective about or what you're replying to here. I also don't know where you're getting those definitions from, & I don't understand how you think they demonstrate a point. We shouldn't imagine that Ṣaʕīdi Arabic necessarily matches up with the geographical limits of Upper Egypt, but, in fact, if that's what you want to do, it matches the map here: If you want to describe everything south of Dahshur and al-'ayyat as the southern limits of Lower Egypt, then what you're saying matches the map.

  I think that this conversation would be more productive if we only worked from reliable sources. The most comprehensive source I know of for Ṣaʕīdi Arabic is Abdelghany A. Khalafallah's A Descriptive Grammar of Sāɛiːdi Egyptian Colloquial Arabic, published by Mouton in 1969. He describes the geographical region in which Ṣaʕīdi is spoken stretching from Cairo to Aswan. You might object that that's old, which would be fair! But to change this we'd need to depend on some other reliable source that said something different. Do you know of one? Pathawi (talk) 02:49, 23 November 2022 (UTC)

@A455bcd9 Seeing all the different opinions, thoughts etc (which to some degree are personal) I'm so impressed that all of you editors actually get something done, it's not easy. Really impressed. --always ping me-- Goran tek-en (talk) 16:32, 22 November 2022 (UTC)

Yes @Goran tek-en, it's hard 😅 A455bcd9 (talk) 16:37, 22 November 2022 (UTC)

@Pathawi: Sir/Ma'am , I have checked who removed the old similar-map, and you were the one who correctly removed it, so my deepest apologies to Mahmudmasri. I would greatly appreciate it if the current map of Egypt were removed and replaced with something that accurately represents the truth. First of all, here is the World Atlas[1] for the Major languages and dialects spoken in Egypt [with percentages]: Egyptian Arabic is 68%, Upper Egyptian Arabic [called Saidi] is 29%, The Bedawi Arabic is 0.15%.

Now, back to the map; the green coloring on the map is incorrect, both linguistically and visually. A massive amount of empty regions have been colored green, making it appear that "Libyan Arabic" dominates Egypt, which is far from the truth and very misleading to any foreign reader who is trying to learn the varieties of the country.

Not to mention, what some foreigners incorrectly call "Libyan Arabic" on the Western tribal border, is in fact NOT the variety of Tripoli Arabic, it's a variety shared ONLY by Eastern Libyan Bedouins AND Western Egyptian Bedouins, that is too close to other Bedawi Arabic and shares separate vocabulary, and because Wikipedia lacks any page for Bedawi: Bedawi/Western Egyptian Bedawi Arabic, they have just called it "Libyan Arabic" many years ago, which is also factually incorrect.

Thank you. 41.238.82.251 (talk) 22:36, 22 November 2022 (UTC)

@41.238.82.251: yes in the ISO standard, "Libyan Arabic" and "Western Egyptian Bedawi Arabic" share the same code (ayl). Hence, the map. A455bcd9 (talk) 22:40, 22 November 2022 (UTC)

I know, they have incorrectly shared the similar code years ago due to lack of any Bedawi page, so someone who is likely not even an Arabic speaker just linked the two. The other false thing is the green color distribution which is literally coloring massive empty areas. Kindly my brother/sister, at least put percentages on the map, and also the Western Egyptian Bedawi only exists in Western area. Thank you. 41.238.82.251 (talk) 22:52, 22 November 2022 (UTC)

Unfortunately, we can only rely on reliable sources, not on opinions expressed in the talk page. A455bcd9 (talk) 22:59, 22 November 2022 (UTC)

This is not an opinion, this is well-sourced. The World Atlas on the languages and dialects of Egypt literally has no mention of it [2]. Actually, there is no any reliable source for the false green coloring on that map. I gave you the actual percentages of the dialects, so yes the map is false and isn't even supported by any local Egyptian source. 41.238.82.251 (talk) 23:08, 22 November 2022 (UTC)

I think that this conversation might go better if we were all on the same page about sourcing in Wikipedia. The core guideline is WP:SOURCES: Reliable sources. (If anyone in this conversation hasn't read that, I really think they should.) I think there are two issues on the table:

1. Some unclarity around what this map is of.
2. Two sources disagree. We can have a reasonable conversation about this.

First: This is a dialect map. If we look at Egypt, most of the areas where Arabic is spoken at all are agricultural or desert. Roughly 20% of Egyptians live in the Cairo metropolitan area. The Sinai Peninsula is forty times the size of the Cairo metropolitan area, but only has a little under 3% of its population. What the map represents isn't the number of speakers, but the areas in which a reliable source has identified this version of Arabic as a home dialect. It's simply not a population map. &, it's worth noting, most political maps are also not population maps. If they were, you'd be used to seeing Saudi Arabia one third the size of Egypt. If you wanted to create a diagram that depicted relative population sizes, there are some other ways you could do it. If you want to take on that work, I'm happy to give you a couple of suggestions for how it could be done.

Second: Okay, this is substantive. One source, WorldAtlas.com, has been introduced to counter another, Ethnologue. I think there are three stages to how to discuss this: First, are the two sources actually in conflict? Second, if they are in conflict, on what grounds can we compare them? Looking at reliable source criteria, is one clearly a preferable source? Third, how do we put the result into the article? Is this a case where only one source is going to take the day, or is it a case where we can integrate multiple sources?

The IP editor has suggested that WorldAtlas.com counters Ethnologue because it gives the following speaker percentages for Egypt: Egyptian Arabic 68%; Ṣaʕīdi Arabic 29%; Bedawi Arabic 0.15%. The map here, however, doesn't deal with population size at all. It deals with areas—some rural, some urban—where particular dialects are majority or significant home dialects. This isn't a point of conflict. As far as regions in which the dialects are spoken, WorldAtlas.com seems to be in agreement with this map. Of Ṣaʕīdi Arabic: "The language is mainly spoken by the Sa'idis, who are concentrated in the South of Cairo all the way to the Sudanese border." There's no conflict. It matches the map.

I'll address the Western Desert Arabic in a separate comment. Pathawi (talk) 01:36, 23 November 2022 (UTC)

I think the main reason that there's no page on Western Egyptian Bedawi Arabic is that editors have not been able to find good sources. Glottolog is generally a pretty good starting bibliography for scholarly linguistic sources. Here's their page for Western Egyptian Bedawi Arabic: nothing. That of course does not mean that the dialect doesn't exist! What it means, tho, is that a Wikipedia editor who wanted to start such a page based on reliable sources might find that task impossible. Unfortunately, Ethnologue has listed it as another name of Libyan Spoken Arabic. Glottolog shows it as one of four dialects of Libyan Arabic. I can't find a single scholarly source to contradic them. That lands us in the situation we're in now: I'm not saying that it's true that Western Egyptian Bedawi Arabic is a subdialect of Libyan Arabic! What I'm saying is that without reliable sources that say otherwise, the appropriate path forward for Wikipedia editors is to go with the sources we've got. Pathawi (talk) 01:47, 23 November 2022 (UTC)

I think that this conversation has been extremely unproductive. I think it would be productive if participants took a look at Wikipedia's Verifiability policy (WP:VERIFY) & the guidelines on Reliable sources (WP:RS). I don't think that we can have a productive conversation otherwise. If we can't agree on founding the conversation on source reliability, then I think we need to go to dispute resolution (WP:DRN) or request comment (WP:RFC). Pathawi (talk) 02:55, 23 November 2022 (UTC)

Here's a geographic description, not a map, describing where Egyptian Arabic is spoken, cited from Aljazeera which is commonly used in Wikipedia in other articles:

وتتمدد اللهجة القاهرية جنوبا للنصف الشمالي من محافظة بني سويف (جنوب الجيزة) والفيوم (غرب بني سويف)،

...

وتبدأ اللهجة الصعيدية رحلتها من شمال محافظة المنيا

These were the anchoring texts to read around. Use Google Translate if can't understand Literary Arabic. --Mahmudmasri (talk) 04:24, 23 November 2022 (UTC)

It might be worthwhile to compare the dialect map we're discussing, the text of this article, & a map of Egypt's governorates. Adding the entirety of Fayyum Governorate & the northern half of Bani Suef would basically convert the area in the current map where the Ṣaʕīdi dialect expands out from the Nile into Cairene dialect. That seems plausible! (I've never spent any time there.) But this doesn't seem like a wild error given the scope of the map that Goran tek-en & A455bcd9 created: It seems like something that could get fixed rather easily. However, if you wanted to take this Al-Jazeera description seriously, we would have to read the whole section—not just the two partial sentences quoted above. We would then break off Damietta as a separate dialect, reducing Cairene Arabic's spread in the Delta, & remove Alexandria as well. We'd also lose al-Sharqiyyah in the east and Buhayrah in the west. In the end, the region of Cairene Arabic on the edited map would actually be much smaller.

But it's not what you or I want, right? It's what the reliable sources say. So is this source preferable to AA Khalafallah's book & Ethnologue? That's a conversation worth having. One major reason it might not be: WP:NEWSORG 'Scholarly sources and high-quality non-scholarly sources are generally better than news reports for academic topics.' This article might still be useful: It would be worthwhile to look into the work of Professor Ahmad Karim to see if any of his work is usable here! Pathawi (talk) 01:48, 24 November 2022 (UTC)

The discussion on the map's talk page, in case someone wants to follow. It's clear that since its creation, it created controversy. --Mahmudmasri (talk) 06:11, 23 November 2022 (UTC)

The old map was also sourced: Map of Varieties of Arabic. A previous version of this file was published in the following source (backwards copy): Schmitt, Genevieve A. (2019). "Relevance of Arabic Dialects: A Brief Discussion". In Brunn, Stanley D.; Kehrein, Roland (eds.). Handbook of the Changing World Language Map. Springer. p. 1385. doi:10.1007/978-3-030-02438-3_79. ISBN 978-3-030-02437-6. as "Fig. 1 Major dialects of Arabic, by region. (Open source)". --Mahmudmasri (talk) 06:33, 23 November 2022 (UTC)

@Mahmudmasri, I asked to add this mention in the description of this image. However, unfortunately, it is NOT a source. The image is original research BUT an OLD version of this old map (so not even the current one) was cited LATER (in 2019) by Schmitt, Genevieve A. As noted by @Pathawi back then (message from 3 September 2022 above): "If you've read Schmitt's chapter, the citation is, in fact, in passing. Schmitt is not a dialectologist, but a researcher in education who is looking into attitudes toward ʕāmmiyyah in Jordan & Egypt. She uses this map as a figure in the chapter, but references it once parenthetically, along with another map, & offers no explanation, & does not anywhere discuss the map. There's nothing in the chapter to suggest that there has been any further review of the map or consideration of its contents." A455bcd9 (talk) 07:20, 23 November 2022 (UTC)

Result

Don't add the map due to the failure to find enough sources. It just creates controversies for one obvious reason, it is very faulty, citing only one source! I'd suggest deleting it all together. --Mahmudmasri (talk) 06:18, 23 November 2022 (UTC)

The main source used, namely Ethnologue, is considered reliable in linguistics, and the only comprehensive continually maintained inventory of the languages of the world containing maps (Linguasphere isn't maintained anymore and Glottolog doesn't contain maps):

• Ethnologue#Reception,_reliability,_and_use
• It has generally been considered reliable on Wikipedia: Wikipedia:WikiProject Languages#Interpreting online sources of data

If you want to challenge the reliability of Ethnologue, please start a discussion here.

In the meantime, I'd suggest keeping the map and removing the above header. Other opinions are welcome. A455bcd9 (talk) 07:30, 23 November 2022 (UTC)

(Or let's start an Wikipedia:RFC as suggested by @Pathawi.) A455bcd9 (talk) 07:31, 23 November 2022 (UTC)

• Ethnologue was similarly criticized for not documenting its sources. How is that very different? Why are you pushing so much Ethnologue? This is very unreasonable. I've never witnessed that before in Wikipedia! To me, it looks somewhere along the lines between WP:ADVOCACY and WP:BOOSTER.
• If you are a contributor to a source, insisting to dubiously cite alone, then it is enough to raise questions about your neutrality!
• No, I am not interested in contributing to Ethnologue or taking its approval. It is a source available like other sources that you both insist on not citing.
• Seeing an older version of the older map, I only see color changes with the same dialect areas, so the map was still cited and unchanged. Claiming it has been altered is a misleading statement by you.

How many more sources you need? I provided one, the map had another, the anonymous user gave you a third. You have three sources disputing Ethnologue's map! --Mahmudmasri (talk) 08:06, 23 November 2022 (UTC)

Let's calm down please @Mahmudmasri. To answer your various points:

• Ethnologue is considered a reliable source by most of the community of linguists on Wikipedia. I'm not "pushing it", I just say that this map is using a reliable source.
• Anyone can send feedback to Ethnologue, as mentioned in Wikipedia:WikiProject_Languages#Interpreting_online_sources_of_data: "For years, Ethnologue had a reputation for being unresponsive, so many linguists do not bother to correct the errors they find, but since ca. 2012 they have been appreciative of feedback, and the quality of their coverage has improved markedly." So it's a recommended practice to contact Ethnologue to fix potential mistakes. Anyone can do so by joining their Contributor Program and then the Ethnologue team checks feedback looking at primary sources and asking their local informants and correspondants to reach a conclusion.
• I'm not insisting on "not citing" other sources. But this is a map, so we cannot "cite" different sources and do WP:OR on a map.
• I'm not misleading anyone. Schmitt's article was first published online on 23 October 2019. If you look at comments in the file history, not only colors were changed in 24 June 2021: "Corrections to peninsular dialects, and added areas for Anatolian Arabic and Maltese" Also, the author isn't a linguist. Her work aren't reliable sources in linguistics. Here's her bio: "Genevieve A. Schmitt earned her Master of Education in instructional leadership and academic curriculum with an emphasis in world languages from the University of Oklahoma."

Anyway, let's start an RFC to get more inputs! A455bcd9 (talk) 10:57, 23 November 2022 (UTC)

(I initially posted the text below in the "RfC on map of varieties" and moved it here as suggested by M.Bitton. A455bcd9 (talk) 16:26, 23 November 2022 (UTC)) Context: A map of Arabic varieties was previously included in this article. In August and September 2022, contributors decided to remove this old map because it was WP:OR, unsourced, inaccurate, and erroneous. I (a455bcd9) requested a new map to be made in the Graphics Lab/Map workshop. Goran tek-en accepted the request and created a new map, following my directions and mostly using one source: Ethnologue's maps. When Ethnologue didn't have a map for a specific country, other reliable maps were used. All sources used are listed on Commons. Once the new map was created, I added it, together with a clickable legend on this article and on Arabic. Two weeks later, an IP (41.238.82.251) removed the map, claiming: "non-academic unsourced and unofficial map, it has many factual errors. Unsourced". @Austronesier added the map back, commenting: "The map is sourced. It may have its flaws, but not to the extent that you claim.". A debate then started on this talk page. @M.Bitton added the {Disputed} template above the map and commented on FPC that the new map was "riddled with asinine errors" and created by "an evangelical Christian organization that doesn't submit itself to any kind of academic review" On this talk page, @Mahmudmasri suggested to delete the new map because "It just creates controversies for one obvious reason, it is very faulty, citing only one source!" @Pathawi and I suggested to keep it as it was sourced using a generally reliable, although not perfect, source per Ethnologue#Reception,_reliability,_and_use and Wikipedia:WikiProject Languages#Interpreting online sources of data. Helpful resource: Wikipedia:Using maps and similar sources in Wikipedia articles

Regarding "standard" MSA pronunciations

Mahmudmasri regarding your recent change adding /d͡ʒ, ʒ, ɟ, ɡ/ to MSA pronunciation of ج:

Phonology of MSA can be a tricky thing to pinpoint, because speakers of MSA tend to be influenced by their dialect to some extent. Pronunciations such as ظ as /zˤ/ or ذ as /z/ in Egypt and the Levant, or ض as /ðˤ/ in much of the Arabian Peninsula, are usually considered nonstandard. These pronunciations are not included in the chart under MSA. The ج as /g/ pronunciation is likewise not the standard pronunciation. High surv (talk) 13:51, 30 August 2022 (UTC)

This is a misleading discussion and presumptuous. It was from your misunderstanding what constitutes standard and informal. Quoting you "ذ as /z/ in Egypt", actually you mixed between the informal pronunciation for this consonant with the standard /ð/. ج has indeed many standard pronunciations. The consonant chart of MSA has the consonant in four blocks and a note underneath. Furthermore, if westerns prefer the fricative pronunciation, that doesn't make it the only standard. --Mahmudmasri (talk) 21:21, 31 August 2022 (UTC)

I did not "mix" between the informal pronunciation. They are both colloquial pronunciations that some speakers use when speaking MSA. ض as /ðˤ/ is very common in the Gulf region when speaking MSA.

The burden of proof is on you to provide a source, which you have not yet. High surv (talk) 09:22, 1 September 2022 (UTC)

This section has some formatting issues. My understanding from being in Cairo is that most educated people do understand the realisation [g] to be the appropriate Fuṣḥā realisation of جيم for Egypt. It is the news broadcasting standard. Many educated people will tell you that the variation between [g] & [ʤ] realisations is a product of which tribe settled where. I don't think this is a confusion of diglossic levels. (On the other hand, in Qur'anic recitation, Egyptians realise جيم as [ʤ], right?) For an appropriately reliable reference, one might take Karin C. Ryding's A Reference Grammar of Modern Standard Arabic (Cambridge University Press, 2005). On page 14, in a table of Arabic consonants, she writes the following for جيم:

There are three standard regional variants:
(a) voiced alveopalatal affricate; /j/ as in "jump";
(b) voiced alveopalatal fricative (zh): as the /z/ in "azure" or the medial sound in "pleasure";
(c) voiced velar stop; /g/ as in "goat";

She then adds in a footnote:

The variations are essentially as follows: the first is more characteristic of the Arabian Peninsula and Iraq, the second more Levantine and North African, and the third specifically Egyptian and Sudanese pronunciation. Occasionally, a mixed pronunciation of jiim is found, with one variant alternating with another, especially /j/ and /zh/.

I rather hate this for multiple reasons, but it's a reputable MSA descriptive grammar which makes a claim that's partially true. (The part I'm balking at: [g] is not a standard realisation of جيم in most of the Sudan—not in Khartoum, not in the East, not in Dar Fur. If it's standard anywhere, I don't know about it. In the Sudan, جيم is generally realised [ʤ], tho [g] or [dʲ] happens in some clusters [eg, مجتمع as [mudʲtamaʕ] is not uncommon].) Faruk Abu-Chacra's Arabic: An Essential Grammar (2d ed, Routledge, 2017) makes a similar claim (6–7):

This letter has three distinct pronunciations depending on the dialectal background of the speaker:
a) In Classical Arabic and the Gulf area, as well as in many other places in the Arab world, it is pronounced as a voiced palato-alveolar affricate as the j in 'judge', 'journey', or g in Italian 'giorno'.
b) In Lower Egypt (Cairo, Alexandria) it is pronounced as a voiced velar stop g, as in English 'great'.
c) In North Africa and the Levant it is pronounced as a voiced palato-alveolar fricative /ğ [this probably should be ‹ǧ›], ž/ in the English 'pleasure', and as j in the French 'jour'. This one is the most common and will be used in this book.

Abu-Chacra is explicitly describing "modern literary Arabic". Pathawi (talk) 13:08, 1 September 2022 (UTC)

The ommission of the Badawi-Carter-Gully Modern Written Arabic: A Comprehensive Grammar is not meant to conceal anything: They explicitly (11) choose not to deal with the issues of pronunciation. Pathawi (talk) 13:13, 1 September 2022 (UTC)

The Qur'anic recitation is [ʤ] as far as I know.

It does seem that [g] is common in Egyptian news. But I will reiterate that if [g] is to be included in the MSA chart, then the same thing applies to realization of ض as [ðˤ] in the Arabian Peninsula, which is very common in news, media, and religious sermons as well.

In both of those cases, many speakers outside of that region would consider the pronunciation incorrect. The contrary is not true, as ج as [ʤ] and ض as [dˤ] are still considered correct in both Egypt and the Peninsula, i.e. there is a canonical pronunciation that MSA speakers everywhere can agree on.

So I believe it's a question of determining the chart's scope. High surv (talk) 15:44, 1 September 2022 (UTC)

I believe that the Qur'anic recitation form is [ʤ] in Cairo, as well (altho I have to say: I'm really not sure if it's [ʤ] everywhere… in places where the usual formal realisation is [ʒ], what's the standard for Qur'anic recitation? I have no idea), but Qur'anic recitation isn't MSA. But the real issue for us as Wikipedia editors is reliable sourcing, right: Verifiability, not WP:TRUTH. I think it can safely be said that an MSA realisation of جيم with three different values is attested in at least two major grammars, neither of which says 'All Arabic-speakers agree on canonical [ʤ], but there's regional variation.' Ryding doesn't give the same account for ضاد—for her it's only a 'voiced velarized alveolar stop' (14). Here's what Abu-Chacra has: 'This is also an emphatic consonant, classified as a pharyngealized voiced alveolar stop. Arab phoneticians and reciters of the Koran recommend its pronunciation as a counterpart to د /d/ (8). In current use in many dialects it is, however, also pronounced the counterpart of ذ /ḏ/ (9), somewhat similar to the sound of th in the English 'thus…' (8–9). So, I don't know how best we should represent that in the article, but I do think there's a solid case for recognising [g] as a realisation of جيم in MSA drawing on multiple reliable sources. Fricative ضاد perhaps we'd have to list as an alternative, given Abu-Chacra's wording. Maybe another reliable source says something different? Pathawi (talk) 16:04, 1 September 2022 (UTC)

&, also, I suppose: Whatever we do should use the sources for what they say, rather than what we want to take from them or what we're interested in. So to give a fuller story: Ryder notes two standard variants of ظاء: /ðˤ/ & /zˤ/. For her description, that would put it on the same level as جيم as having different standard MSA (redundant, I know) variants. She does not, however, mention variation for ضاد. Abu-Chacra, however, describes ظاء as /ðˤ/, with dialect variation as /dˤ/ & /zˤ/. This makes the variation dialect-influenced—like his account for ضاد—& not basic—like his account for جيم. Neither grammar lists variation for any other consonantal phoneme. Pathawi (talk) 16:32, 1 September 2022 (UTC)

I'll include a couple more sources:

Alhawary's Modern Standard Arabic Grammar lists one pronunciation for each consonantal phoneme, with no variations (though he notes regional variation in consonant influence on vowels). The following pronunciations are listed:

• ج as a voiced alveolo-palatal fricative "produced with the tip of the tongue against the alveolar ridge and the middle of the tongue raised against the palate in this sequence: as "j" in judge".
• ض as a voiced alveolar stop emphatic.

David Cowan's Introduction to Modern Literary Arabic notes three variations in consonantal phonemes:

"The following letters are pronounced more or less as in English: ب = b, ت = t, ث = th as in 'think', ج = j (in Lower Egypt as g in 'get')..."

"ض is an emphatic 'd', or in certain countries (e.g.. Iraq) 'th' as in 'that', pronounced with the tongue pressing hard against the edge of the upper teeth with the tip protruding."

"ق is a guttural 'k' pronounced from the back of the throat. Colloquially, with the exception of Lower Egypt and some parts of Syria and Palestine where it generally becomes a glottal stop, it is almost universally pronounced as 'g' in go." (this is of course false)

Schulz's A student grammar of modern standard Arabic does not note any variations. The following pronunciations are noted:

• ج like g in gentle
• ض like d articulated with emphasis

High surv (talk) 00:25, 2 September 2022 (UTC)

It was inappropriate from High surv to revert without reaching a consensus, that's apparently against his claims, but I'll come back when I have time. If anyone claimed the plosive gim pronunciation is nonstandard, he knows nothing about Arabic, standard, and how it's used. And actually, this plosive pronunciation is also the standard by coastal Yemenis all the way to the neighboring region of coastal Oman. The Sudanese variant is mostly [ɟ] or [d ͡͡ʒ]. Mahmudmasri (talk) 18:18, 1 September 2022 (UTC)

The problem with /d͡ʒ, ʒ, ɟ, ɡ/ in slashes is that no regional (standard or basilectal) variety has four different segments for ج in its inventory, and that is what this notation says and why it is misleading. In every variety, there is only phoneme that is variously realized as [d͡ʒ ~ ʒ ~ ɟ ~ ɡ] depending on location. The choice of the phoneme symbol is arbitrary (it could even be /☎/ as suggested by Mark Hale in a different context); if you feel that /d͡ʒ/ gives undue weight to a reading that is commonly taught to non-Arabs as "Standard" but actually shouldn't have a monopoly on that label, we could simple change all the sounds in the table in the MSA column to phones in brackets, and naturally have [d͡ʒ ~ ʒ ~ ɟ ~ ɡ] in the ج-row. The key information in the table isn't the MSA pronunciation anyway, but the regional pronunciations; MSA is just a reference point. –Austronesier (talk) 19:30, 1 September 2022 (UTC)

What if we just do away with MSA altogether? It's a pedagogical term that only has limited utility when talking about the actual Arabophone world anyhow. You're absolutely right that the MSA isn't the point of the table. Why not just use the Wikipedia strict Romanization? This could get rid of what's ultimately a distraction. Pathawi (talk) 00:12, 2 September 2022 (UTC)

(Outdenting because of the table.) This could look like:

		ق q	ج j	ث ṯ	ذ ḏ	ظ ṭ	ض ḍ
Levantine	Jordanian (Western Amman)	[g], [ʔ]	[d͡ʒ]	[t], [s]	[d], [z]	[dˁ], [zˁ]	[dˁ]
	Syrian (Damascus)	[ʔ]	[ʒ]
	Lebanese (Beirut)
	Palestinian (Jerusalem)

& so on. (Flipped the axes for readability.) The dialect values, after all, don't derive from modern media varieties. If someone really wanted, we could throw in historical dialects as well, drawing on the work over the past couple decades on Northern Old Arabic, Old Hijazi, &c. If there's a desire, we could include a row on "MSA", as it is a set of varieties.

In general, this section would benefit from sourcing. There's almost none.Pathawi (talk) 16:52, 2 September 2022 (UTC)

Everyone's mad about the map! Does no one want to weigh in on this? Pathawi (talk) 01:49, 23 November 2022 (UTC)

Including a set of MSA varieties might be a bit better than removing MSA entirely. I think MSA is still a useful reference point to compare colloquial dialects to, and perhaps even more useful when the comparison includes both the regional colloquial dialect and the regional MSA. High surv (talk) 18:09, 25 November 2022 (UTC)

References

Sudanic and Egyptian dialects

There's a disagreement with user_talk:24.190.69.161 regarding classification of dialect groups.

The classification of a single Egyptian-Sudanic group is noted in every relevant source:

• Oxford's Handbook of Arabic Linguistics
• Glottolog
• EISELE, JOHN C. (1987). "Arabic dialectology: A Review Of Recent Literature". Al-'Arabiyya. 20 (1/2): 199–269. JSTOR 43191695.
• "Arabic, a great language, has a low profile". The Economist. 2018-10-20. Retrieved 2020-06-24.
• Dana Hooshmand (2019-07-11). "Arabic Dialects Compared: Maghrebi, Egyptian, Levantine, Hejazi, Gulf, and MSA". Discover Discomfort. Retrieved 2020-06-24.
• Al-Wer, E. (2018). "Arabic Languages, Variation in". In Brown, Keith; Ogilve, Sarah (eds.). Concise Encyclopedia of Languages of the World. Elsevier Science. p. 53,54. ISBN 978-0080877747.
• Kamusella, Tomasz (2017). "The Arabic Language: A Latin of Modernity?" (PDF). Journal of Nationalism, Memory & Language Politics. 11 (2): 117–145. doi:10.1515/jnmlp-2017-0006. S2CID 158624482.

The dialect group is called the "Egypt" dialect group in some sources, but includes both Egyptian and Sudanic dialects. I've seen no source which describes six dialect groups with a separate "Sudan" group.

@24.190.69.161: Are you really insisting that Saidi Arabic and Khartoum Arabic are from different dialect groups? They are very clearly similar dialects.— Preceding unsigned comment added by High surv (talk • contribs) 01:36, 17 October 2020 (UTC)