Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Wiki Education Foundation-supported course assignment[edit]

Sciences humaines.svg This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Sanakareem20. Peer reviewers: Shalineem.

Above undated message substituted from assignment by PrimeBOT (talk) 14:37, 16 January 2022 (UTC)Reply[reply]

Semi-protected edit request on 17 November 2021[edit]

At Arabic#Structure there is a sentence "The suras, also known as chapters of the Quran, are not placed in chronological order". The link in that sentence points to Suras which redirects to Sudra Kingdom. Since the sentence is about the Quran, the link should probably point to Surah which is the actual article on the chapters of the Quran. Alternatively the redirect at Suras could be changed to achieve the same result. 2406:5A00:EC14:B800:A9A4:3E19:E094:B9 (talk) 01:35, 17 November 2021 (UTC)Reply[reply]

 Done You're correct, I fixed the link here. I'll check into the Suras redirect, which a priori seems strange. Largoplazo (talk) 01:49, 17 November 2021 (UTC)Reply[reply]

Semi-protected edit request on 29 December 2021[edit]

Boutarfa Nafia (talk) 20:17, 29 December 2021 (UTC) "change third most to third most spoken(even though it is false because mandarin is more spoken than arabic) "Reply[reply]

 Not done You've misunderstood the sentence: "Arabic, in its Modern Standard Arabic form, is the official language of 26 states and 1 disputed territory, the third most after English and French." Not the most spoken, the most states in which the respective languages are official. English and French are each official in more than 26 countries. Mandarin is not. Largoplazo (talk) 20:25, 29 December 2021 (UTC)Reply[reply]

Arabic and Eritrea, Djibouti & Somalia[edit]

These two countries do not have any spoken dialect of Arabic of their own. Only Standard Arabic is taught and is most widespread in these areas. Only a fraction of the Arabic speakers in Djibouti and Somalia have a genuine Yemeni Arabic dialect from prior habitation in Yemen, more people only know Standard Arabic than speak Yemeni Arabic there. The Yemeni dialect of Arabic is not predominant in Djibouti and Somalia. Afar and Somali are the local vernaculars, while Standard Arabic is the main form of Arabic learned for religious and commercial reasons. I suggest changing the map and somehow reflecting this reality. AzanianPearl (talk) 14:07, 20 January 2022 (UTC)Reply[reply]

@AzanianPearl: The map says "Yemeni and Somali" Arabic - it makes not pronouncement on the relative proportion of speakers in the different areas. The way the map is currently presented, it suggests that the dialects are probably related but not the same. Based on what sources are you seeking to establish the complete separation (as opposed to interrelatedness) between Yemeni and Somali Arabic? And do you have any sources that clearly establish Afar Arabic as a further distinct vernacular, separate again from both the Yemeni and Somali variants? Iskandar323 (talk) 14:37, 20 January 2022 (UTC)Reply[reply]
@Iskandar323: There is no Somali dialect of Arabic. Afars and Somalis all learn it as a second language, pretty much exclusively Modern Standard Arabic, albeit at low level due to the badly run education system. Currently the map suggests that a variant of Yemeni Arabic is spoken by most in the Eastern Horn, this isn't true. I suggest that somebody changes it and creates an additional category for areas with mostly only second language Arabic speakers. For the Amazigh, the Toubou, Nubians, Beja, Eritreans, Djiboutians, and Somalis - the first few mentioned speak Arabic dialects, but the last three only mostly know MSA and a mismatch of dialects from media and not local speech. AzanianPearl (talk) 18:48, 20 January 2022 (UTC)Reply[reply]
People speaking Arabic in Djibouti or Somalia must be speaking a dialect of some kind. The map says "Yemeni and Somali", which implies that the Horn of Africa's Arabic is somehow related to Yemeni Arabic while subtlety different from it. The article is just about Arabic, so it is largely irrelevant whether it is a first or second language. The map just shows different dialects. You have noted that, where people do speak Arabic, it is Yemeni. Standard Arabic is not a dialect, so what you are saying seems to suggest Yemeni Arabic is the dominant dialect. Iskandar323 (talk) 19:26, 20 January 2022 (UTC)Reply[reply]

@Iskandar323: Dialect suggests there is a pattern to it and unwritten rules. There is none for the Arabic of this region. It is mostly heavily accented or badly spoken Standard Arabic, not a real L1 colloquial dialect like the case is with Yemen or other areas with L1 Arabic. These maps often cause confusion because people often skim through text and look at maps. This may lead people to believe that the Yemeni Arabic dialect is predominant in Djibouti and Somalia, while when you get there you mostly only hear L2 or L3 Fusha. Perhaps it should be completely split from the Yemen association then and just call it the Somali dialect. I hope somebody reading this can make this change. AzanianPearl (talk) 02:05, 21 January 2022 (UTC)Reply[reply]

I agree with Azanian Pearl. People speaking Arabic in Chicago & Berlin must be speaking some dialect of Chicago, but they're not included on this map for a reason. The map is split into two shades, the darker of which is meant to indicate that Arabic speakers are the majority. The map suggests that a significant portion of Somalia/Somaliland has a majority Arabic-speaking population. We really need to know what the source of the data behind this map & related maps are. At present, that's not at all clear. Pathawi (talk) 03:20, 27 July 2022 (UTC)Reply[reply]
This is all already clarified in the infobox, which contains two images, one showing that Arabic is a minority in Somalia, and another that it is a co-official language due to significant Arabophobe minorities, or for historical and cultural reasons - so the assessment of this single image lower down the page in isolation is misleading. You have suggested that a bastardized version of Arabic is spoken - well, so be it, but if that is the case, there is still nothing inherently wrong with calling that a 'Somali dialect'. Moroccan Arabic is a long way from Levantine Arabic because it is mixed with Berber language and French. Normal. Iskandar323 (talk) 06:46, 27 July 2022 (UTC)Reply[reply]
I think there are two gaps in the communication here: First, a key issue is that the second map shows that there is a significant portion of Somaliland & Bari in which Arabic is the majority language, as well as a minority language thruout Somaliland & in most of the rest of Somalia. If it is a majority language, more than half of the population speaks Arabic as a first language in these regions. That is surprising, and has got to be able to be substantiated somehow if we're going to promote the claim. It's not clear what source substantiates that claim. The issue there is not Arabic as a minority language. Second, I think that AzanianPearl is saying that the Arabic spoken in Somalia is a second language, learned (poorly, they claim) in schools. In this sense, it is, as they say, 'not a real L1 colloquial dialect' like Levantine & Moroccan Arabic are, as it is not these speakers' L1 ('L1' here meaning 'first language'): It is an L2 (second language) learned to less than complete fluency. (I don't mean to belabour the point of how well or poorly the average Somali who learns Arabic learns Arabic: I just want to be clear that the issue isn't a "bastardized version of Arabic" comparable to native speaker regional dialects.) An additional point beyond the miscommunications: Regardless of any of this, material on Wikipedia needs to be sourced. It's very unclear what the source is for this map & the others in its series. That might be the most important issue. Hope that cleared some things up. Pathawi (talk) 02:59, 28 July 2022 (UTC)Reply[reply]

Pathawi I think the best solution is to split Somali from the Yemeni and Somali label on that map and give the Horn of Africa region new subcategories: Eritrean, Djiboutian, Somali. The problem however is that these are not real L1 dialects, but L1.5 or L2 dialects / accents (one may argue it could be L1 for some in Eritrea). There are various Muslim groups in Eritrea who use Arabic as a lingua franca. I personally know an Agaw (Bilen) Eritrean who speaks more Arabic with his family than he does his ethnic language (Agaw/Bilen) and he does not live in the Arab world. Some Tigre Eritreans may also be using Arabic as something in between L1 and L2, so an Eritrean 'dialect' of Arabic may be argued for. As for Djiboutian, it is an official language there on par with French. In order to avoid tribal/ethnic conflicts between Afars and Somalis the two 'neutral' languages of Arabic and French were made official instead. The trained ear can likely detect a Djiboutian dialect in that region. As for Somali, no Somalis use Arabic as an L1 or anywhere close to it, nevertheless it is an official language in both Somaliland and Somalia and people are actively encouraged to learn Arabic there for religious and commercial reasons. Given the linguistic influence of Somali on their Arabic speech one can pick up on it as well and it definitely doesn't sound like the Yemeni dialect (more like Somali phonetically influenced MSA). In sum, I suggest somebody removes the Yemeni and Somali category and creates three HOA Arabic variants (Eritrean, Djiboutian, Somali) for this map. Thank you in advance whoever is able to do it. AzanianPearl (talk) 13:10, 29 July 2022 (UTC)Reply[reply]

AzanianPearl: Whatever we do should be based in reliable, noteworthy sources. Arabic is an official language in Eritrea because of the Rasha'idah, so it's definitely an L1 for a portion of the population. (Oddly, I think that that means that the dialect of Arabic which is an actual L1 in Eritrea is a bit different from the Arabic that many Eritreans can speak as an L2, which isn't that different from what you hear in the Sudan.) But I don't think it's on us to determine which dialects exist: We need to represent the sources. I think the map should only show Arabic as a majority language where we have reliable, noteworthy sources that show that it's a majority language. Whatever is meant by minority language should be consistent, & also supported by sources. I don't think that's the situation we have. Pathawi (talk) 23:56, 29 July 2022 (UTC)Reply[reply]
Pathawi I haven't looked up the statistics for this, but I think there are significantly more non-Rashaida Arabic speakers in Eritrea than just the Rashaida in Eritrea. The Tigre use it a lot due to living close to the Sudan border, also the Muslim Bilens, Beja, Kunama and Afars also adopted it as an L2. Usage of Arabic in Eritrea is about as common as it is in Djibouti or Somalia, and unlike those two countries Eritrea has some groups who use it as an L1 (or in some cases what I would call L1.5 - in between L1 and L2, very advanced speakers of it) (not just Rashaida - many Bejas, Tigres, and Bilens as mentioned before), but these are my personal observations and given that Eritrea is such a closed off country (often called the North Korea of Africa) I am not sure if reliable info on this can be obtained. AzanianPearl (talk) 02:06, 9 August 2022 (UTC)Reply[reply]
AzanianPearl I'm also working from personal observation rather than good data. Knowledge of Arabic is clearly widespread in Eritrea, but the fact that many Jabartis speak Arabic at near-native competence isn't the reason that the country has adopted Arabic as one of its official languages, right? The reason that Arabic is an official language is the same reason that Kunama is a national language: Some Eritreans speak it as their home & heritage language. Again, this is just very limited experience, but impressionistically the Jabartis and Tigre people I have known who speak Arabic fluently essentially speak Sudanese Arabic. The Rasha'idah people I've met speak a markedly different Arabic. None of this is relevant to the question at hand, though: We really can't use a map produced for Wikipedia from unknown & unidentifiable sources, especially when it seems not to be correct (whatever the situation with Eritrea, Somalia really seems to be wrong). I think we should remove this map. Pathawi (talk) 02:20, 9 August 2022 (UTC)Reply[reply]

Requested move 10 February 2022[edit]

The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review after discussing it on the closer's talk page. No further edits should be made to this discussion.

The result of the move request was: Not moved per WP:SNOW (non-admin closure) (t · c) buidhe 09:14, 11 February 2022 (UTC)Reply[reply]

ArabicArabic languageWP:CONSISTENT, every other language has the word "language" after the language name CR-1-AB (talk) 21:18, 10 February 2022 (UTC)Reply[reply]

Oppose Every other language doesn't (examples: Yiddish, Urdu, Tok Pisin, Esperanto). As a free-standing word, the language is the primary topic. For your background, this has been discussed before. See Talk:Arabic/Archive 6#Requested move 8 May 2020. Largoplazo (talk) 21:34, 10 February 2022 (UTC)Reply[reply]
  • Oppose. The word language is optional. Only when a related ethnic group uses the same name is it added to the title. Just to add examples: Latin, Chavacano, Haitian Creole, Interlingue. Showiecz (talk) 23:50, 10 February 2022 (UTC)Reply[reply]
  • Oppose (I want to support because I've gotten confused myself in the past because the vast majority include "language", but guidelines are what they are). The naming conventions for languages says to use "language" when the word for the language is not the primary topic. Basically, it's meant to be used as a disambiguator. Arabic is even used as an example for when not to include "language" on the page. As a note outside of this immediate discussion, if you feel strongly about this, then you should probably go to the guideline talk page to propose a change to the convention, or workshop it with the village pump or relevant wikiprojects. --Xurizuri (talk) 00:53, 11 February 2022 (UTC)Reply[reply]
  • Oppose The language is a primary topic, similar to Hindi, Urdu, etc. (talk) 04:18, 11 February 2022 (UTC)Reply[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Spread of the arabic language[edit]

In section “spread”:

“Arabic spread with the spread of Islam. Following the early Muslim conquests, Arabic gained vocabulary from Middle Persian and Turkish.”

Editor states that arabic gained vocabulary from middle persian and turkish, yet article cited states differently.

From the article:

“The professor explains: “In pre-Islamic times you find borrowings from Akkadian, ­Aramaic, Ethiopic, South ­Arabian, Greek, Latin; after the conquests, when the Arabs came into contact with other people, there is, for example, a lot of Middle Persian and ­Turkish, and in early Abbasid times, when you had the Bayt Al Hikma in Bagdad, where all the translations were made, there is a heavy influx of classical Greek…”

It sounds to me like the professor was underlining the influence many other languages. Furthermore, the article doesn’t state that the arabic language gained vocabulary from middle persian and turkish, more like they just “came into contact with” them.

If I’m wrong, at least put the other languages up there along with middle persian and turkish, which would better utilize the cited article.

Thank you! Asalhathlool (talk) 21:22, 19 July 2022 (UTC)Reply[reply]

What the source is talking about is the influence that contact had on the language. It says specifically that "there is, for example, a lot of Middle Persian and Turkish". "There is" where? Clearly, the "where" is the Arabic language. Further, the emphasis in the source was on the Persian and Turkish, as it singled them out and used the phrase "a lot of". There's nothing wrong with the article summarizing the info in this source by reflecting the source's own emphasis. Largoplazo (talk) 21:58, 19 July 2022 (UTC)Reply[reply]

Semi-protected edit request on 7 August 2022[edit]

The article says that Arabic first emerged in the 1th century to the 4th century, but this is wrong, 1th century to 4th CE century is when Arabic script was born, but the Arabic language is much older than, Arabic language first emerged in 10th to 5th centuries BCE, even if we don’t know a lot about the language in that period, and don’t forget the Nabateans who spoke Arabic in 4th century BCE to 2th century CE (Aramaic was their official language, but their every day language was Arabic),so pleas edit it 2A02:ED0:528D:EB00:59E4:3E80:B8E5:B59F (talk) 22:44, 7 August 2022 (UTC)Reply[reply]

 Not done: For an edit request we'll need reliable sources that assert that the Arabic language is significantly older than it's script (this can either already be in the article, in which case you'll point them out for easy review, or be newly introduced) alongside with what changes you want exactly, with an example being something akin to please change X to Y. —Sirdog (talk) 04:57, 8 August 2022 (UTC)Reply[reply]

Semi-protected edit request on 12 August 2022[edit]

There are some mistakes in the article: Firstly, Qariat al faw inscription is dated back to the first century BCE not AD. Secondly Arabic is much older than than the first century AD, I mentioned that in a former edit request but I didn’t include sources, Arabic as a spoken language is at least 3000 years old but we don’t know a lot about the language in that period because nearly all survived fragments consist personal names and single proper nouns , and even the Assyrians (who are the first to mention the Arabs) mentioned Arab kings and queens of Qedar and their names were pure Arabic in the 9th century BCE , which means that the Arabic language was spoken back then, but started to be written later, Arabic started to be a written language in the second half of the first millennium BCE in different scripts than the Arabic script( like safaitic, Hismaic etc) we know today, so its wrong to say that Arabic first emerged in the first century AD, the earliest Arabic texts written in Safaitic script are dated to the 3rd century BCE while the latest to the 3rd century is more accurate to say that Arabic first emerged in the first half of the first millennium BCE but also mention that it wasn’t written back then, I hope you edit the article

Sources: 1.


3. 2A02:ED0:528D:EB00:B0B9:BB51:1A9E:87C2 (talk) 02:01, 12 August 2022 (UTC)Reply[reply]

I've requested the attention of a linguistics expert to this page. There are certainly some very strange statements about the chronology of Arabic's origination, not least these two, apparently mutually somewhat unintelligible sentences: "Linguists generally believe that "Old Arabic" (a collection of related dialects that constitute the precursor of Arabic) first emerged around the 1st century CE. Previously, the earliest attestation of Old Arabic was thought to be a single 1st century CE inscription in Sabaic script at Qaryat Al-Faw, in southern present-day Saudi Arabia." Linguists believe ... the same thing as they previously believed. Something is definitely off here. Iskandar323 (talk) 17:21, 15 August 2022 (UTC)Reply[reply]
 Not done for now: please establish a consensus for this alteration before using the {{edit semi-protected}} template. Aaron Liu (talk) 15:11, 16 August 2022 (UTC)Reply[reply]

Semi-protected edit request on 14 August 2022[edit]

Arabic first appeared in the mid-ninth century BCE, Qariat al faw inscription is dated back to the 1 century BCE Zedtased (talk) 03:32, 14 August 2022 (UTC)Reply[reply]

 Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format and provide a reliable source if appropriate. Aaron Liu (talk) 15:12, 16 August 2022 (UTC)Reply[reply]

Proposal to Remove Two Maps[edit]

Hope you all are well. This is a continuation of the discussion under Arabic_and_Eritrea,_Djibouti_&_Somalia above. Briefly: There are three maps on this page:

  1. a dispersion map, which purports to identify the locations in which Arabic is a majority language & those in which it is a minority language;
  2. a map of countries in which Arabic has some official status, or a significant body of Arabic-speakers;
  3. a map of Arabic dialects.

I believe that we should remove the first & third of these. In theory, such maps should be very useful to Wikipedia's readers. However, it has not been possible to identify the source of the information in these maps, & it appears that some of it is incorrect. We shouldn't spread misinformation, & we should only work with reliable sources.

Wikimedia Commons user Fobos92 has made a great many very attractive maps that are used throughout Wikipedia & affiliated projects. Unfortunately, over the past years this user has ignored most questions regarding the source of information on language maps. They appear not to have had a verbal interaction on a Talk page at Wikimedia Commons since 2018. For relevant unanswered questions about the sources of linguistic maps: 2017, 2018 +, 2019 +. These dates aren't meant to criticise that user, but to emphasise that it is unlikely that answers about the source data will be forthcoming. In 2015, the user was unable to defend a map of the Balkan Romance languages because they couldn't recall the source. Similar deletion proposals based on implausible map contents & the lack of any sourcing in 2016, 2017 (another), 2018 (another), 2021 (another) saw no response from this user. (In some cases the maps were deleted, while in some they were kept: Error in a map is not a reason for it to be deleted from Wikimedia Commons.) The user did participate briefly in a 2018 discussion concerning a Czech map, but did not respond to questions about source. Similarly, in a 2019 discussion of a map of Francophonie, they defended the existence of the map based on its use by sister Wiki projects, but did not respond directly to concerns about the lack of sourcing. Inaccuracy led to the removal of this user's Welsh language map from Wikipedia's Welsh language article in 2018.

Again: My point is not to criticise an editor on a sister Wiki, but to emphasise that it is unlikely that we will learn the source of these maps. Wikimedia Commons user High surv asked about the source of the Arabic dialects map in 2021. I reiterated the question about a month ago. AzanianPearl asked a question about the same map in Wikimedia Commons in January. We have not received a response in either of these locations, & I don't think we will.

On the grounds that these maps do no reflect information from a reliable source, I propose that we should delete them. In fact, I think that the argument is solid enough that I would have just done that & then explained myself on the Talk page. However, the semi-protected status of this page suggests that there's been some contentious editing here in the past, & I thought it might be worthwhile to open a conversation before deleting. An alternative approach would be to edit the files in Wikimedia Commons: It is generally better to improve than to delete. Unfortunately, I think that this is a substantive task that won't happen quickly. I think that the appropriate course of action would still be to remove the maps from this page now because they contain unsourced probable misinformation, then reintroduce them once they've been revised in accordance with reliable secondary sources. Pathawi (talk) 12:10, 18 August 2022 (UTC)Reply[reply]

The dialect map has a source that it is attributed to on Commons: Relevance of Arabic Dialects: A Brief Discussion (2019) Iskandar323 (talk) 17:56, 18 August 2022 (UTC)Reply[reply]
I believe that's backwards: Schmitt uses the map from Wikimedia Commons—her chapter isn't the source of it. Pathawi (talk) 18:26, 18 August 2022 (UTC)Reply[reply]
I've read the Schmitt chapter, looked at the map, & checked the source history to be clear. Schmitt states clearly that she is employing an open source map, tho she does not identify what the source is. There are two other maps in the chapter for which she explicitly states that she is the source. We don't use these. The map in Schmitt's book is visually identical to the 18 February 2013 version of this map. The book that that chapter appeared in was published in 2020. Furthermore, the description in Wikimedia Commons does not say that the map is drawn from Schmitt. It says: 'Used 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)"' That is: Schmitt used the Wikimedia Commons map: The Wikimedia Commons map was not derived from Schmitt.
However, I've also realised that I made a mistake. The dispersion map is by Fobos92, but the dialect map is not. It has a much longer history, & has been edited by many hands. It looks extremely few of them used any sources other than personal knowledge. It's a bit complex & the version history doesn't all make sense to me yet. In any case, it is definitely in very large part not from secondary sources. I am not 100% certain, but I believe that the original version is also not drawn from a secondary source. I'll try to work out the version history, but I think that for any interested parties, a quick glance at the last ten years of the file's history will show that nearly all changes have been unsourced. My apologies, however, for the incorrect characterisation. Pathawi (talk) 21:41, 18 August 2022 (UTC)Reply[reply]
Thank you for bringing this interesting case to our attention. I too believe that this is a case of WP:CIRCULAR, but I'm not a 100% certain either. So just to be on the safe side, I will start a discussion on WP:RSN and will ping you from there. M.Bitton (talk) 13:03, 19 August 2022 (UTC)Reply[reply]
(The following content is moved from the above-mentioned WP:RSN discussion, modified to remove reference to prior comments in that thread.) A thing to keep in mind is that Wikimedia projects have different guidelines. Most of what's happened at Wikimedia Commons is normal & acceptable by their project guidelines. The sourcing criteria for English Wikipedia are different. We require reliable sources, ideally secondary. Only one of the retained edits over at Wikimedia Commons has a sourced, as best I can tell: The 3 January 2013 edit. (This has been removed from Wikimedia Commons, & transferred to a Wikipedia file discussion.) The sources are multiple, but principally Wikipedia (!) & Ethnologue. The former is clearly problematic. Where Ethnologue is cited, the map used here does not match Ethnologue maps. I'm going to give one example—the one which led to my digging into this in the first place—to show why this is such a problem.
The map we have here portrays a "Yemeni and Somali" dialect of Arabic in use throughout Southern Red Sea Region in Eritrea, Djibouti, & all of Somalia except the southern Ethiopian & Kenyan border region. The Wikimedia Commons cartographer admits in the discussion that they do not have a specific reason to believe that Somalia speaks a subdialect of Yemeni Arabic, but they use a map prepared by Fobos92 as evidence that Arabic is spoken throughout most of Somalia. Fobos92 lists as their source another map from Wikimedia Commons. This map was uploaded by Vispec as their own work. We don't have to posit that Wikimedia Commons editors did anything wrong by that projects criteria to recognise that from a Wikipedia sourcing perspective there are problems every step of the way. Vispec's map is supposed to be a map of countries in which Arabic is an official language. As best I can tell, it seems accurate. I have no argument with it. Fobos92's map cites Vispec's map, but: 1) purports to tell us something different: the dispersion of Arabic as a majority & minority language; &, 2) draws intricate borders that cannot be derived from Vispec's map. I have written above about the problems with sourcing & verification of Fobos92's maps. User Moester101 then uses Fobos92's map as justification for arguing that Arabic is spoken thruout most of Somalia. But note that even if Fobos92's map were correct, that wouldn't justify considering the Arabic in Somalia to be Yemeni Arabic. Now, initially Moester101 cited Ethnologue as a source for Somali Arabic (same discussion page). Ethnologue recognises that Arabic is one of the "Principal Languages" (their category) of Somalia because it's one of the two languages recognised by the Transitional Federal Charter. 2.05 million Somalis use Arabic but not as their native language. (For what it's worth, Ethnologue is behind the times: There is a subsequent Provisional Constitution that demotes Arabic to the status of "second language". But I'm okay for now with us accepting Ethnologue as a reliable source.) My best guess is that the reason for thinking that the Arabic of Somalia is Yemeni is that Ethnologue notes 17,000 Ta'izzi-Adeni Arabic-speakers in Somalia. This is a very small number (less than 1% of the Somali population), & they are explicitly noted as non-native—that is, it's a minor immigrant language. Nothing in Ethnologue on either the Somalia page or the Ta'izzi-Adeni Arabic page allows us to localise speech of that Arabic dialect in Somalia. The Ethnologue map of languages in Somalia shows only Somali languages, & then a very small pocket of Mushungulu speakers & another coastal pocket of Swahili speakers. Arabic does not appear at all on the Ethnologue map. There is no Somali Arabic dialect. There are of course speakers of Arabic in Somalia just as there are in Toronto & Tokyo. In fact, Dearborn, Michigan, USA is far more Arabic-speaking in term of percentage of L1 speakers than Somalia seems to be & in terms of total Arabic-speaking population.
I've checked a few others of the Ethnologue citations & they just don't match. This has been long because I want to show how this mapping practice has led to the fabrication of a dialect of Arabic spoken over a wide area that does not exist. I want to reiterate the core point, however: This is what's going on in the sole case where sources are listed. There are numerous other edits for which no source is listed. This is a huge problem. Pathawi (talk) 01:10, 20 August 2022 (UTC)Reply[reply]
I agree though I believe it might be best to try to make a more accurate map, which would be best if done by a group of researchers and linguists rather than one.
I'll also point out that Fobos92 has made the dialect maps for most of the Arabic dialects, and they reflect portions of the map #3 in question. For example, see the maps in Maghrebi Arabic, Moroccan Arabic, Najdi Arabic, Gulf Arabic, Egyptian Arabic, etc. High surv (talk) 21:18, 27 August 2022 (UTC)Reply[reply]

It's been a week, so I've removed the two maps.

I am incapable to follow the ontological turn the WP:RSN discussion has taken, so I will comment here. Agree to remove a map that is obviously inaccurate in some parts and is eventually based on OR uploaded to Commons. This also applies to Varieties of Arabic. –Austronesier (talk) 13:49, 26 August 2022 (UTC)Reply[reply]
I see no reasons to remove File:Arabic Dialects.svg, as mentioned in Wikipedia:Featured article candidates/Levantine Arabic/archive1, the map is used in a reliable source, which is a validation by a scholar, a posteriori, that the information in the image is reliable. A455bcd9 (talk) 17:57, 2 September 2022 (UTC)Reply[reply]
I have unfortunately created a fork in this discussion by responding to your reversal of my edit at Talk:Levantine Arabic. I will create a comment directing people here.
The problem with that reasoning, A455bcd9, is that it creates a WP:CIRCULAR citation, which is expressly against our Verifiability policy: The map is full of errors. Someone—in passing—uses Wikipedia as a source, replicating those errors. Wikipedia then uses that citation as a source to justify the errors that we created in the first place. See the problem? Pathawi (talk) 19:20, 2 September 2022 (UTC)Reply[reply]
Falsehoods don't become truth because somebody, however reliable in general, inadvertently didn't recognize that they were false. You know, like when major newspapers issue retractions after discovering that something they'd written was incorrect. If things became the truth just because they'd been published, there'd be no need for retractions. Largoplazo (talk) 10:29, 3 September 2022 (UTC)Reply[reply]
Hi everyone,
It's not our task to determine what is true or false. We use reliable sources. Here we have a reliable source about Arabic dialects that, I think, essentially says "Here is an open source map that I found, and I agree with its content". It's not "Someone" who "in passing" "used Wikipedia as a source". Wikipedia is not the source: it's just an image that happened to have been uploaded on Wikimedia Commons. Genevieve Schmitt does not use any information from Wikipedia. Also, she could have found a similar map anywhere else (on Twitter, in another book, on a blog, etc.) and could have added it to her article and we would reuse it without any problem. Why does it make a difference that the image was uploaded on Wikimedia Commons? (Who knows, maybe Genevieve Schmitt originally uploaded the map on Commons :) ) So I consider that the map cited in this reliable source is sourced. Any posterior update made to that make is not though. WP:CIRCULAR is about Wikipedia, it says nothing about open source images uploaded on Wikimedia Commons. I ping @Buidhe as they agreed with that reasoning in Wikipedia:Featured article candidates/Levantine Arabic/archive1 and added the source to Commons back then.
Then, we can discuss whether this map is accurate and if it should be used or not. But that's a different debate I think. A455bcd9 (talk) 13:11, 3 September 2022 (UTC)Reply[reply]
(We can also argue whether the source [and Genevieve Schmitt in particular] is good enough or not, but that's also a different debate.) A455bcd9 (talk) 13:25, 3 September 2022 (UTC)Reply[reply]
One thing that you say which I agree with is that it's our task to rely on reliable sources, rather than our own knowledge. However, the point of WP:CIRCULAR is to not perpetuate our own errors. In fact, I think that many of the really striking errors that occur in this map come from editors depending on their own research & what they believe to be their own knowledge rather than on appropriate sourcing. The errors are an index of the sourcing failure. I urge you to look over the errors concerning Somali Arabic, Nubi Arabic, & Badawi Arabic that I mentioned in the summary over at Talk:Levantine Arabic to see what kind of errors we're perpetuating here. In each case, I'm noting the conflict between what appears on the map & what occurs in the source it's meant to be drawn from.
It's true that the version of the map that Schmitt uses is from Wikimedia Commons, but I think that's a loophole that isn't: If you go thru the edit history at Wikimedia Commons, you'll find that the original version of the map is from Wikipedia. The conversation concerning sourcing is on Wikipedia. It's just a larger circle. & there's another piece: The actual editors of the map at Wikimedia Commons do not claim Genevieve Schmitt as a source in any of their edits: You and User:buidhe decided in a conversation on Wikipedia to use Schmitt's use of the map as post hoc sourcing. Buidhe then added the information (with an explicit backwards copy notice) to the Wikimedia Commons metadata. It is very, very clear that no Wikimedia Commons editor who has added anything to the map itself has claimed to draw on Schmitt. Schmitt is not part of the chain of sourcing, & is only introduced by you in Wikipedia, in what then becomes a case of circularity. (That is: In Wikimedia Commons, it's just largely unsourced, & mostly badly sourced in the rare cases it's sourced; when we retroactively use Schmitt to justify usage of the map, it becomes circular.) You all had to review a lot of material to get Levantine Arabic into shape for Featured Article status. I think you did amazing work. But I disagree with this particular judgment.
One last thing: 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. Pathawi (talk) 13:59, 3 September 2022 (UTC)Reply[reply]
I think the premise here is that other reliable sources make it clear that the information in the map is false. It doesn't become true because one ordinarily reliable source blindly picked up a graphic from Wikipedia, the falsehood of which is made plain by previously existing sources. Largoplazo (talk) 12:31, 4 September 2022 (UTC)Reply[reply]
I agree with this premise, & I think it's an important argument against using this map. My point—I probably wasn't clear—was that there wasn't reason from Schmitt's research or background to assume additional review of the map (& we oughtn't make that assumption anyhow: it's got to be explicit). Pathawi (talk) 15:04, 4 September 2022 (UTC)Reply[reply]
  • Just adding the fact that there is another source (by a linguist) that uses a similar map. M.Bitton (talk) 14:05, 3 September 2022 (UTC)Reply[reply]
    & I'll add that I've contacted this author, & he confirmed that he drew the map from Wikimedia Commons, was upset by the errors, & wants to use a corrected map in the second edition of his book. I understand that my private communication isn't a reliable source for Wikipedia's purposes, but that doesn't matter: Prof al-Sharkawi's map isn't cited here. No editor of the actual map in use here claims to have drawn from al-Sharkawi. One could take this map, edit it to show that Omani Arabic is spoken in inland Saudi Arabia, Iraq, Jordan, & Syria, while Najdi Arabic is spoken exclusively in Oman, upload that modified map, & then claim to draw it from al-Sharkawi. But why WP:GAME the system like that? Pathawi (talk) 14:16, 3 September 2022 (UTC)Reply[reply]
    So I see three issues:
    1. The accuracy of this map. For sure there are mistakes, all maps are wrong: the map is not the territory. I contributed a lot to Ethnologue to try to improve maps but it's a complex topic. Even the definition of Arabic varieties are unclear (for instance, what are the distinctions between "South Levantine" and "North Levantine"? What's the border? No source clearly defines these points...). The current map, for all its mistakes, seems "okay" to me, as long as it is sourced.
    2. The general validity of a posteriori sourcing: if an unsourced map is uploaded to Wikimedia Commons and later used in a reliable centered source where the author says "Here's a valid map from Wikimedia Commons", then, can we consider that this specific version is sourced? I think so. Wikimedia Commons is just an image bank and this author would use an image from there to illustrate their point.
    3. If we agree on point 2, then, can this principle be applied to this particular instance? I think the source is "okay" but as Pathawi said: "Schmitt is not a dialectologist, but a researcher in education who is looking into attitudes toward ʕāmmiyyah in Jordan & Egypt." so it's definitely not a "high-quality reliable source" for a featured article such as Levantine Arabic (I removed it). On the other hand, I think Sharkawi's modified version comes from a "high-quality reliable source". It would be nice to upload it to Commons.
    What do you think? A455bcd9 (talk) 08:41, 4 September 2022 (UTC)Reply[reply]
All reliable sources are "generally reliable", but there are instances where editorial judgement is also needed to deal with apparent inaccuracies in a RS. Of course this has to happen with utmost care and consensus to avoid abuse for POV-pushing purposes, but in the case of @Pathawi's comments it should be obvious that their motivation is a genuine concern for accuracy and not some kind of hidden ethnochauvinist agenda. Unfortunately, a map is an all-or-nothing matter (in prose, we can just trim the text in the affected parts); if contains flaws (beyond the usual grey areas that come with artificial cut-off points, e.g. "South Levantine" vs. "North Levantine"), we shouldn't use it. –Austronesier (talk) 09:56, 4 September 2022 (UTC)Reply[reply]
I greatly appreciate A455bcd9's removing the map: Thank you. I don't think we agree on the enumerated issues, however. As I see it, the crux of things is their item number two: I do not believe that we can engage in this kind of post hoc sourcing—I think it's definitionally circular in a way prohibited by our verifiability policy. No editor of the map itself has cited Schmitt as a source, so as things stand the map is largely unsourced, & in the cases where it is sourced the sources are often misread & frequently creatively interpreted. Thus, the map itself runs afoul of verifiability in the most fundamental way. If we bring Schmitt in to retroactively justify what's already been done, we create an issue of circularity, & we again run afoul of verifiability—I think quite explicitly. This is the core issue, & the rest of the following is a less important response to the other points.
As for accuracy: It's true that our mandate is 'verifiability, not truth', but it seems to me that if one traces how things came to be wrong, all of it is traceable back to a failure to draw on reliable sources. Sure, no map is perfect; this map invents a dialect out of thin air & places multiple dialects in the wrong countries by hundreds of kilometres. These errors would not be in the map if it drew on reliable sources & if multiple editors were checking one another's work, as we do in good articles. I know I've already said this, but: The map's errors are an index of a looseness with verifiability. (Al-Sharkawi's book introduces additional problems—not the result of verifiability looseness—because of errors the printer made in swapping in black & white patterns for this map's colours.)
Finally, on the third point, I didn't mean to create a context for judging Schmitt or her work. I mentioned her background & the topic of her research & her chapter because I didn't think we could reasonably assume that she had reviewed the map's contents with an expert eye. However, even if Schmitt were a decorated Arabic dialectologist, her use of the map would be an explicit case of circularity: '[D]o not use… publications relying on material from Wikipedia as sources. Content from a Wikipedia article is not considered reliable unless it is backed up by citing reliable sources.' No additional reliable sources are added for the map. The same is true with al-Sharkawi's map.
I really think our best path forward is to start from scratch, using reliable sources every step of the way. Pathawi (talk) 10:08, 4 September 2022 (UTC)Reply[reply]
I agree that it would be amazing to have a better map. What would you suggest?
I strongly disagree with your interpretation of circularity. Circularity is about Wikipedia, not content uploaded on Wikimedia Commons. Let's imagine that a decorated Arabic dialectologist publishes a map of Arabic dialects in a centered reliable source (for instance the "Arabic Dialects" chapter of the "Encyclopedia of Arabic Language and Linguistics"), for me it doesn't matter whether:
1. They created the map from scratch (purely based on their own knowledge)
2. They found the map in another book (reliable or not, whether they cited this source or not)
3. They found the map published by a random dude on Twitter or on a blog.
4. They found the map on Wikimedia Commons and the map is Original Research.
5. They found the map published by a random dude on Twitter or on a blog BUT it happens that this random dude also uploaded the map (Original Research) on Wikimedia Commons.
On the other hand, if I understand your reasoning, for you 1, 2, and 3 are okay, but 4 and 5 are not? Just because the map was uploaded at some point on Wikimedia Commons it would make it invalid for subsequent use by a reliable source? A455bcd9 (talk) 12:04, 4 September 2022 (UTC)Reply[reply]
Well now we're going in circles! :) The map originates in Wikipedia. It was later moved to Wikimedia Commons. It's just a bigger circle: Wikipedia → Wikimedia Commons → published source → Wikipedia. It's not that the map was uploaded on Wikimedia Commons that makes it unreliable: It's that uploading Wikipedia content to Wikimedia Commons doesn't magically exempt Wikipedia content from verifiability standards. I think that the Wikipedia/Wikimedia Commons distinction here is a technicality here, as there's so much back & forth (the sole explanation of sources, for example, got moved from Wikimedia Commons to Wikipedia; all the editors I've discussed in the creation of the map have accounts on both), but even attending to that technicality, this is circular. So none of your examples 1–5 are at play here. Since you've asked about my reasoning, I'd say that the verifiability policy does not cover 1–3 in clear explicit ways; I would argue that 4 violates the spirit of circularity, but that it would be a debate that might suggest a need to clarify the circularity policy; I don't know what to think about 5, but I will say: It's absolutely possible to dream up difficult border cases for lots of policies (all policies?). I don't think that this is one.
I hope this won't come off as cross, but I do want to say: This is a lot of talk to keep a map which has some really, really big errors off of Wikipedia. Please note what I've said about Somali Arabic (doesn't exist), Nubi (absent from the right countries; in the wrong countries by multiple borders & hundreds of kilometres), Badawi (in the wrong country by hundreds of kilometres). We shouldn't be struggling to interpret policies in ways that allow us to keep this.
With regard to what I suggest for a better map, I've been sounding out Arabic faculty at my university for their opinions (I'm a graduate student in linguistics; I work in the Arab world, but not on Arabic). I can't find a single source on Arabic dialectology that gives a map similar in scope to this one. My opinion is that this reflects a gap between what some Wikipedia editors want, & what specialists feel is the current state of knowledge in the field. Kees Versteegh has a very general statement on large dialect groups that's repeated in numerous sources, & I think could be taken as a starting point. Working from there to more localised specialist sources, one might use coloured dots to indicate sites of dialect attestation. This would allow the overlap that really exists, & wouldn't require the construction of arbitrary borders. One Arabic prof suggested that the dots might have a diffuse glow around them to show that these are centres of a form of speech, rather than hard boundaries. I don't feel attached to this vision: I feel like it's a conversation to be had among Wikipedia editors who are willing to work on it. What is important is that any such map be reliable sourced & consistent. Pathawi (talk) 14:00, 4 September 2022 (UTC)Reply[reply]
The map originates in Wikipedia this is what we think, it's not a fact that we can establish beyond a shadow of a doubt. M.Bitton (talk) 14:08, 4 September 2022 (UTC)Reply[reply]
No. We know this. Look at the document history on Wikimedia Commons. It's explicit. Pathawi (talk) 14:12, 4 September 2022 (UTC)Reply[reply]
I was referring to the maps that are cited in the sources. M.Bitton (talk) 14:21, 4 September 2022 (UTC)Reply[reply]
You could of course spend time contemplating the Infinite Monkey Theorem, but it's generally not very persuasive in real world application. It's also not relevant to the case at hand, in which the editors of the actual map in question do not cite Schmitt or al-Sharkawi as a source. Pathawi (talk) 14:30, 4 September 2022 (UTC)Reply[reply]
Now, you're making sure that we keep going round in circles. Ad I said previously, there is nothing stopping anyone from uploading a map that is based on the two sources; and contrary to what you seem to think, doing so has nothing whatsoever to do with gaming the system M.Bitton (talk) 14:36, 4 September 2022 (UTC)Reply[reply]
You tried to have this conversation over at Noticeboard#Probable_WP:CIRCULAR, but there was no real uptake. What you want to do is conduct some sort of social experiment on Wikipedia editors to find out what we would do if this were the case. You suggested the possibility of uploading al-Sharkawi's map knowing—from our conversation at the Noticeboard—that I had contacted him, that he stated that he copied the map from Wikimedia Commons, & that he wanted to correct the errors in the next edition of his book. I don't believe you disbelieve me. I also don't believe you believe that a map with the same arbitrary borders & the same errors was independently produced thru a path that we can trace in the Wiki projects & by an academic. The proposal entails placing information on Wikipedia that you know to be false in order to experiment with a way of pressing a possible loophole in the verifiability policy. I have already elsewhere suggested that you could have this conversation as a hypothetical at the Village Pump without this experiment of perpetuating bad information on Wikipedia. It does not relate to the map at hand, which is the topic in question. It relates to a prospective similar future map. Pathawi (talk) 15:01, 4 September 2022 (UTC)Reply[reply]
It's not a social experiment. The way I see it, there is no difference between assuming, a some editors do, that the map is sourced (since it's used in RS) and uploading a new map that is based on the cited sources. M.Bitton (talk) 15:39, 4 September 2022 (UTC)Reply[reply]
Just to be clear: I agreed with you that we should remove this map and I undid my edit on Levantine Arabic. On my side, the debate regarding this map is over.
Then, I mentioned, more broadly, the issue of circularity with maps. I disagree with you (Pathawi) on this point; but it's life, let's move on :)
So let's focus on building a better map. Ethnologue has the following maps:
Did I forget one variety/dialect or one region? (Maltese may be added, but that's a detail we can discuss later). These maps have mistakes but I don't think we can find better ones. And Ethnologue is the reference anyway. If we patch these different maps together, we would have all the dialects on one single map. What do you think? A455bcd9 (talk) 15:42, 4 September 2022 (UTC)Reply[reply]
I checked the list of all recognized varieties (other than Judeo-Arabic), and Ethnologue has another map for Uzbeki Arabic (not in free access) and a not very precise map for North Mesopotamian in Turkey. So I think we have a map for all of them on Ethnologue, and now we "just have" to merge them into one big map. A455bcd9 (talk) 15:53, 4 September 2022 (UTC)Reply[reply]
I don't know about the other maps, but looking at the first one, I see a number of issues and factual inaccuracies, for example: nobody speaks Amazigh in Algeria. M.Bitton (talk) 16:12, 4 September 2022 (UTC)Reply[reply]
Amazigh on the map refers to the Kabyle language, spoken east of Algiers. A455bcd9 (talk) 16:17, 4 September 2022 (UTC)Reply[reply]
Whomever made the map doesn't obviously doesn't know the difference between Kabyle and Amazigh (a language that nobody speaks in Algeria). M.Bitton (talk) 16:21, 4 September 2022 (UTC)Reply[reply]
Amazigh is a synonym of "Berber", as explained here. See also: "At the core of their struggle in recent years – amid a long history of colonial suppression – is their stated desire to no longer be referred to as Berbers but as Amazigh, meaning "free people", and for their language to be known as Tamazight."
Anyway, Ethnologue is the reference in linguistics, it's a highly-reliable source. If we find any mistake, as I'm part of their contributor program and I can forward it to them and they would correct it in February next year. A455bcd9 (talk) 16:22, 4 September 2022 (UTC)Reply[reply]
Ah, and also, Amazigh is the official language of Algeria, together with Arabic: (see also Languages of Algeria) A455bcd9 (talk) 16:28, 4 September 2022 (UTC)Reply[reply]
Actually, it's Tamazight (an invented language that nobody speaks) and not Amazigh that is an official language, and Amazigh is not synonymous with the Kabyle language (there is no question that this is a mistake). I'll look into it further once the map is complete. M.Bitton (talk) 16:34, 4 September 2022 (UTC)Reply[reply]
What you say was historically true but with the Amazigh movement, Kabyle speakers more and more refer to their own language as "Amazigh". See: Berbère/Amazigh ou Kabyle ? Évolution et fluctuation d’une dénomination en contexte d’idéologies dominantes.
"L’usage généralisé du mot amazigh (pl. imazighen) comme ethnonyme et glottonyme unificateur qui se réfère à toutes les zones berbérophones, la Kabylie inclue, où le terme n’était pas connu, est davantage un phénomène qui a son origine dans l’activisme et le renforcement du Mouvement berbère ou amazigh. Ce terme a été utilisé dans sa nouvelle (ou ancienne) acception la première fois en Kabylie dans les années 1945-50, par les nationalistes berbères Kabyles qui chantaient la liberté dans leur langue maternelle : le kabyle. [...] De fait, de nos jours, le terme amazigh est en train de remplacer de plus en plus fortement, aussi bien au Maroc qu’en Algérie, le terme arabo-européen ‘berbère’ dans presque tous les médias, quotidiens, revues, télévision, radios, sur Internet ou même dans la littérature. [...] L’usage en français du terme tamazight, donne l’impression d’un corps étranger, car c’est le nom de cette langue mais en berbère, comme l’est, par exemple Deutsch en allemand pour désigner la langue allemande, ou English an anglais ou encore al-’Arabiyya en arabe. Mais dans l’usage français, ces mots deviennent respectivement ‘allemand’, ‘anglais’ et ‘arabe’. Il est clair que ces mots pour s’adapter à la langue française ont dû prendre les formes phonétiques et morphologiques de la langue d’accueil. Ce qui n’est pas le cas de tamazight. Il serait plus correct d’adapter cet usage et en faire par exemple ‘l’amazigh’ pour ainsi avoir des possibilités de dérivation comme ‘la langue amazighe, un amazigh, une femme amazighe, l’amazighité’ etc… Chose qui ne serait pas possible avec la structure morphologique berbère du terme tamazight. A455bcd9 (talk) 16:44, 4 September 2022 (UTC)Reply[reply]
It's still as true today as it was before. What you're citing is someone's opinion/wish with regard to all the Berbers and not specifically the Kabyles or their language (that they wouldn't change for the world). Besides, even if his wish became true, why should the Kabyle language be singled out? M.Bitton (talk) 16:56, 4 September 2022 (UTC)Reply[reply]
I won't spend more time on this issue which is irrelevant to the Arabic dialect map discussion. (it's a good example of Brandolini's law though...) @Pathawi, what do you think about my proposal? A455bcd9 (talk) 17:04, 4 September 2022 (UTC)Reply[reply]
I won't spend more time on it either, but the factual inaccuracy in one of the map (I haven't checked the others) casts a shadow over the so-called reliability of ethnologue. M.Bitton (talk) 17:10, 4 September 2022 (UTC)Reply[reply]
Sorry, A455bcd9, for the delayed reply: I was dealing with a family issue. I have the same critiques of Ethnologue that some other academics do, but I think it's pretty clear that Wikipedia has a more or less solid history of treating it as a reliable source (there are exceptions, but they don't result in general deprecations of the whole). My fantasy map would draw directly on Arabic dialectological work, but that would take a lot longer. I think this is a viable path forward that relies exclusively on a reliable source. There will still be errors because reliable sources can contain errors! But it will avoid the biggest of those that we've recently had. Thanks for proposing a workable path forward. Pathawi (talk) 17:49, 4 September 2022 (UTC)Reply[reply]
Once a starting large map is made, a good path to improving it might be to work on each of the individual dialect maps and change the larger map to reflect that information. High surv (talk) 21:31, 4 September 2022 (UTC)Reply[reply]

Yes Ethnologue is far from perfect, but again, that's the best we have. If you want to improve it I suggest you join their Contributor Program:

It's really nice because it gives you a free access to the whole website. A455bcd9 (talk) 19:06, 4 September 2022 (UTC)Reply[reply]

New map
Thanks to @Goran tek-en's wonderful work we now have a great map, based on reliable sources (mostly Ethnologue). I updated the article accordingly. Are there other articles where this map could be added @Pathawi? A455bcd9 (talk) 22:25, 7 November 2022 (UTC)Reply[reply]
I greatly appreciate your & @Goran tek-en:'s work in making this happen, A455bcd9 (talk · contribs). This was a ton of work, & I'm thrilled that we have a map that better reflects reliable sources. I'm sorry I haven't been more involved: It's a crazy moment in the academic term for me. As for your question: There are other locations where this map should be added. I systematically removed the old map from other pages. I'll go thru my edit history & figure out what those are, but… heh-hehm… Levantine Arabic was one of them. Pathawi (talk) 23:35, 7 November 2022 (UTC)Reply[reply]
Added back to Levantine Arabic, thanks! A455bcd9 (talk) 08:38, 8 November 2022 (UTC)Reply[reply]
Pathawi, A455bcd9 thanks for this job but it based in new borders but the dialect not limited by brders cordially Nehaoua (talk) 15:06, 25 November 2022 (UTC)Reply[reply]
Nehaoua, I agree that dialects are not limited by borders. A map that shows dialects as territories like modern states isn't my preferred form of dialect mapping: If I were to make a dialect map, it would not be of this form. That said, all maps are simplifications that gesture toward (hopefully) useful information. Dialect maps of this form are very common in both Wikipedia & scholarly literature. It's possible that they mislead people (I suspect that they do), but they also seem to provide a useful overview. I think more importantly, reliable sources use such maps, which I think makes it difficult to object to their use on Wikipedia without a higher level conversation.
So while this isn't the map I would choose to create, I do support it for the following reasons:
  1. I think it's better to have a sourced map than an unsourced map, which was the prior situation & which I think would happen again if a sourced map didn't fill the vacuum. The sources in question use maps of this form.
  2. I think that dialect maps of this form are pretty clearly accepted in practice by the majority of the Wikipedia community, & are clearly used by sources I consider reliable, so I feel like we ought to accept them unless we're going to make a principled argument for Wikipedia as a whole, which would have very far-reaching consequences.
  3. While the way that A455bcd9 & Goran tek-en have chosen to do this is not the way that I would choose to do it, I think that the way they choose to do it is one reasonable way to handle the situation, & I respect their decisions & the work they've done.
Am I understanding your concern right? Does my perspective on this make sense to you? (A few other examples of similar dialect maps on Wikipedia: Middle English, Basque dialects, Languages of China.) Pathawi (talk) 15:27, 25 November 2022 (UTC)Reply[reply]
Hi @Nehaoua, I wholeheartedly agree with @Pathawi:
  • Maps following nation states' borders are not realistic, for sure. But "the map is not the territory" => all maps are abstractions of the reality.
  • A sourced map using a reliable source is better than an unsourced map (Wikipedia:No original research)
  • Absurdities (such as "Somali Arabic" or the Nubi language in Egypt, as in the previous map) are more serious mistakes than border issues.
  • This map is a general overview of all dialects. We can then have more detailed maps dialect by dialect (including sub-dialect of each dialect).
What do you think? A455bcd9 (talk) 16:14, 25 November 2022 (UTC)Reply[reply]
@Pathawi Hello, thanks, reasonable explanation.  we have two Maps one without sources but very realistic and another with Sources but not realistic, I prefer not to change without modification and approbation. Nehaoua (talk) 16:58, 25 November 2022 (UTC)Reply[reply]

Nehaoua, could you clarify a little what you mean by realistic? The previous map also showed dialects in this manner—I'm not catching the difference in realism. Pathawi (talk) 17:08, 25 November 2022 (UTC)Reply[reply]

hello @Pathawi A clear separation between the dialects' borders coincides exactly with the political borders!  is it real? cordially Nehaoua (talk) 17:46, 25 November 2022 (UTC)Reply[reply]
@Nehaoua: it is not indeed, but it's a simplification of the reality. And it is backed by a reliable source. (Also, please note that the border problem on the new map only affects a few borders (mostly in Tunisia and Algeria): the rest of the map doesn't follow political borders at all.)
On the other hand, what do you think about Nubi language in Egypt, Somali language as an Arabic dialect or Judeo-Arabic dialects spoken only in Jerusalem? Aren't these mistakes bigger than a few borders? A455bcd9 (talk) 17:53, 25 November 2022 (UTC)Reply[reply]
Hello @A455bcd9 and it is my remark (Algerian borders) because I know well this repartition and the map is not perfectly, good to correct this map thank you for this cordially Nehaoua (talk) 18:08, 25 November 2022 (UTC)Reply[reply]

Nehaoua: Hmm… Are we looking at the same map? The map under Spoken Varieties looks to me like it only coincides with national borders in two locations where the the colours appear as diagonal stripes. According to the key to the map, these are regions where there are indeterminate scattered populations, so, for example, there are speakers of Shuwa Arabic thruout Chad, but reliable sources don't indicate where within Chad, specifically. That's what the hatched areas over those two countries mean. Otherwise, it doesn't look to me like the dialect areas match up with national borders. Is it possible that you're looking at the top map on the page, which shows countries where Arabic has official status? Pathawi (talk) 18:03, 25 November 2022 (UTC)Reply[reply]

Helle @Pathawi see Algerian borders Nehaoua (talk) 18:11, 25 November 2022 (UTC)Reply[reply]
@Nehaoua, regarding the Algerian border:
  1. Can you provide reliable sources showing that the map is wrong?
  2. If there's a mistake, is it a big mistake? Or only a small issue at the border?
  3. Is this issue more serious than "Somali Arabic" or "Ki-Nubi" in Egypt?
A455bcd9 (talk) 18:51, 25 November 2022 (UTC)Reply[reply]
@A455bcd9 yes, the distribution of Arabic dialects in the Maghreb is totally wrong. I am An Algerian and I know the difference between these dialects, for example, the east-south (Tébessa province) of Algeria is very close to the Tunisian dialect and the north-west of Tunisia is the same dialect as the Algerian dialect, and can't change mistake with other, we can rectify Nehaoua (talk) 19:28, 25 November 2022 (UTC)Reply[reply]
@Nehaoua, you haven't answered my questions. Could you please answer them? (1. Reliable sources. / 2. Importance. / 3. Importance compared to other "Somali Arabic" and "Nubi in Egypt".) A455bcd9 (talk) 19:30, 25 November 2022 (UTC)Reply[reply]
@A455bcd9 1. this reliable source [1] is very different for your unique reliable source Nehaoua (talk) 19:45, 25 November 2022 (UTC)Reply[reply]
@Nehaoua: Pardon me if I'm wrong, but I think this source doesn't show "Algerian Arabic" and "Tunisian Arabic". It only shows the different varieties (sub-dialects?) spoken in Algeria. Did I misunderstand something? A455bcd9 (talk) 19:49, 25 November 2022 (UTC)Reply[reply]
I think the conversation could be clarified a little. First: Nehaoua's map from Kheireddine Abainia looks great to me, & I'd probably be happy to see it serve for modifications to the dialect map currently in use. My preference will always be to see the work of linguists working on dialectology favored over Ethnologue. (The only qualification in my 'probably' is that I'd want to quickly check the works cited in Glottolog to see whether or no it was in competition with another Algerian Arabic dialect map.)
That said, I'm less clear about the topics of the remainder of the conversation. Is anyone proposing anything beyond that? There's a bit of a philosophical issue about what it means to describe some range of variation as a dialect, but honestly if there's not a concrete proposal that's a path I'd rather not wander down. Pathawi (talk) 05:08, 26 November 2022 (UTC)Reply[reply]
@Pathawi: for the context, @Nehaoua came here after they reverted my edits on the Arabic Wikipedia. They seem to prefer the old map to the new one, based on the distribution of Arabic dialects in the new one. But, like you, I don't know what they want: revert here as well? modify the old map? modify the new one? What do you suggest @Nehaoua?
Regarding Abainia's map, I wonder:
  • Does this map come from a reliable source?
    • The author: Kheireddine ABAINIA is a computer scientist (PhD in Telecommunication & Information Processing and Assistant Professor teaching "Introduction to Arduino and communication protocols") who applies his knowledge to the Algerian dialects. He may be a reliable source on computational linguistics but I don't think he's a reliable source on Arabic dialectology.
    • The publication: Language Resources and Evaluation: impact factor of 1.8 (that's "average") and focus on "acquisition, creation, annotation, and use of language resources, together with methods for evaluation of resources, technologies, and applications" => Again, I don't think this review is reliable in dialectology (and it's not even an excellent source in computational linguistics)
    • The article: "This paper aims to provide a new multi-purpose parallelcorpus (i.e., DZDC12 corpus), which will serve as a testbed for various naturallanguage processing and information retrieval applications." => the illustration of dialects is secondary to the main purpose of the article. Again, the article isn't focused on Arabic dialectology.
  • Assuming for the sake of the argument that if this map came from an acceptable source. Could we even use it in a larger map of Arabic dialects? How?
    • Is "Sahel Algerian" the same thing as Algerian Saharan Arabic? What about "Saharan Ksouri"? What about "Saharan"? What do the stripes (Berber areas) mean? How should we color the Berber areas? Is the "Extreme Eastern-Southern" dialect part of Algerian Arabic or Tunisian Arabic? Are the western dialects part of Algerian Arabic or Moroccan Arabic? The article and the map say nothing about these questions.
    • Glottolog and Ethnologue only have 3 sub-dialects in Algerian Arabic (in addition to Hassaniya and Algerian Saharan Arabic): Algiers, Constantine, and Oran.
    • How do we match them with the 13 dialects on that map? We would have to guess and use our own knowledge (WP:OR) or combine this map with other sources (WP:SYNTHESIS?).
A455bcd9 (talk) 10:59, 26 November 2022 (UTC)Reply[reply]
I think that one side effect of trying to do so much with this map at once is that there end up being three (probably more) different discussions in different venues at the same time. If there's not a proposal that's specific to this page, then I think this conversation should move to the Wikimedia Commons page for the map (or a Wikipedia page for the map itself). Pathawi (talk) 13:31, 26 November 2022 (UTC)Reply[reply]
& there we go: [2]. Since this issue would affect all uses of this map, I have launched the conversation at the map's home, rather than under this somewhat old & effectively already settled conversation about a different map on a specific Wikipedia page. Pathawi (talk) 13:53, 26 November 2022 (UTC)Reply[reply]
Thanks a lot @Pathawi. A455bcd9 (talk) 14:00, 26 November 2022 (UTC)Reply[reply]
@A455bcd9 thank you for your contributions, with this map, I show you the difference between the new map that divides all of Algeria into two dialects but there is an interactive dialect with near countries (Tunisia, Libya, Mauritania and Morroco)
  • Orani, Saharan, Tell, Sahel Algerian (Sahal=plains), = Algerian sub-dialect
  • Saharan Ksouri = Algerian Saharan Arabic
  • the extreme western dialects part of Moroccan Arabic
  • Extreme east = Tunisian dialect
cordially Nehaoua (talk) 16:15, 26 November 2022 (UTC)Reply[reply]
 Moved to Commons: Commons:File_talk:Arabic_Varieties_Map.svg#Algeria. A455bcd9 (talk) 17:12, 26 November 2022 (UTC)Reply[reply]

Scope of this article[edit]

The "Phonology" and "Grammar" sections currently extensively describe MSA's phonology and grammar and briefly describe colloquial varieties. We have articles about Modern Standard Arabic, Classical Arabic, and all other Varieties of Arabic (e.g. Levantine Arabic). So extended description of these varieties' phonology and grammar should be discussed there only.

So I suggest to simply delete the "Literary Arabic" and "Colloquial varieties" subsections of "Phonology" and "Grammar". These sections are largely unsourced anyway.

Instead, we should have a few paragraphs about the general common features of all Arabic varieties, based on reliable sources.

What do you think? A455bcd9 (talk) 14:33, 6 September 2022 (UTC)Reply[reply]

Voiced or Voiceless Glottal Fricative[edit]

Under Arabic#Phonology, it is claimed that Arabic has a Voiced Glottal Fricative (ɦ). Is this meant to be ه (hāʾ)? I believe hāʾ is realized as a Voiceless Glottal Fricative (h) in almost all varieties of Arabic. The independent Arabic phonology page also describes this sound as voiceless.

--human1011 (talk) 18:04, 21 September 2022 (UTC)Reply[reply]

All the Arabic 'h' analogues seem to be consistently described as 'voiceless' - here's one source going through all the fricatives: [1] Iskandar323 (talk) 09:19, 22 September 2022 (UTC)Reply[reply]
I've taken a deep dive into the page history. The table was messed up here without explanation. I will change it back to correct [h]. –Austronesier (talk) 21:13, 22 September 2022 (UTC)Reply[reply]


  1. ^ "Arabic Fricative Consonants" (PDF).

Minority languages by country[edit]

These need serious sources. I doubt Arabic is recognized in most of those countries. Beshogur (talk) 22:10, 18 March 2023 (UTC)Reply[reply]

I agree but at least the Israel and South Africa appear to be actually recognized no clue on the other countries Qwv (talk) 23:24, 23 March 2023 (UTC)Reply[reply]