Arabic and the vernaculars, part 6
« previous post | next post »
This post grew out of a comment I was making yesterday to a previous post about a wall at INALCO (Institut National des Langues et Civilisations Orientales [National Institute for Oriental Languages and Civilizations]) (established 1669) in Paris that listed the many languages taught at that venerable institution.
As my eyes surveyed the mass of names on the wall, one thing struck me powerfully: the large number of different Arabic languages. This raised an interesting question: common "wisdom" is that there is only one Arabic language, viz., Modern Standard Arabic [MSA], so how come there are so many different Arabic languages taught at INALCO?
Since the Arab vernaculars have been one of our favorite foci here at Language Log (see "Selected readings" below), I was interested to see how many different varieties of Arabic are represented on this wall:
Judéo-Arabe, Moroccan Arabic, Algerian Arabic, Libyan Arabic (but that is MSA), Yemeni Arabic (also MSA, though it is generally considered to be a very conservative dialect cluster), Lebanese Arabic, Palestinian Arabic, Egyptian Arabic, Arabe Littéral (which I take to signify written / literary MSA) in contrast to dialectal Arabic (though I'm not sure how it differs from regular MSA; perhaps it is hyper-conservative to a degree that it it not really "sayable", i.e., "writable but not sayable", cf. "Sayable but not writable" [9/12/13]; i.e., MWA [Modern Written Arabic]?).
I do not include Maltese because of the Romance superstrata, nor do I include Sorabe because that only refers to the script used to write the Austronesian language known as Malagasy, much as the Perso-Arabic script is used to write Sinitic Hui (Muslim) Mandarin.
It would appear that, to the teachers and administrators at INALCO, there must be sufficient differences among all these Arabic languages that they merit having separate offerings.
It seems that Arabic speaking countries possess a diglossia of MSA for formal, "proper" occasions and usage and a variety of vernaculars on a dialectal continuum for daily informal usage. SFAIK, only Egyptian vernacular has developed a widely used literary / written / cinematic form, even transcending its political borders to a certain degree. To what extent the other vernaculars are popular in music / song / TV, etc., in general, the situation must vary widely by country.
Worldwide, Arabic has about 372.7 million speakers (source), yet it has all of these different Arabic languages listed on the INALCO wall, despite the fact that theoretically and doctrinally there is only one (MSA).
Contrast that with India, which has a total population of 1.417 billion people, yet only has two or three languages represented on the INALCO wall, if you count Sanskrit, which is overwhelmingly a classical language. Here at Penn, we teach virtually all of India's 22 official languages, at least 10 at any given time.
Urdu, which is mutually sayable with Hindi, but not mutually readable, is not taught at INALCO (source); with 230 million speakers, it is the 10th-most widely spoken language in the world (source).
Hindi, which is taught at INALCO, has 615 million speakers and is the 3rd-most widely spoken language in the world (source).
Bengali is also taught at INALCO, probably because it is the official language of Bangladesh, has 272.7 million speakers and is the 7th-most widely spoken language in the world (source).
China has a total population of 1.412 billion people, of whom Ethnologue claims there are roughly 900 million speakers of Mandarin. I don't believe it. These are simply political propaganda figures put out by the CCP / PRC government. At least half of those 900 million don't understand the other half. It is telling that the wall refers to "Chinese" (Mandarin?) as Chinois in French and Zhōngwén 中文 (lit., "central script / writing") in Sinographs. Even if they're referring to "Mandarin" writ exceptionally large, Pǔtōnghuà 普通話 ("Common Speech") or Guóyǔ 國語 ("National Spoken Language") would be a better designation.
The INALCO wall also lists Cantonese, which certainly is a separate language, worldwide has nearly 100 million speakers. Taiwanese (Táiyǔ 臺語) is listed on the wall and has approximately 15 million speakers. Added to their related confreres on the mainland they constitute a group of approximately 50 million Minnan ("Southern Min") speakers. I am pleased to see that the INALCO wall includes Uyghur, Mongolian, and Tibetan, all three of which non-Sinitic languages the CCP/PRC is actively trying to eliminate.
On balance, the INALCO wall reveals a great deal about the geopolitics of French language policy in the world today.
Selected readings
- "Arabic and the vernaculars, part 5" (8/20/22) — if you are interested in the problems raised in the current post, part 5 in the series is must reading.
- "Arabic and the vernaculars" (3/6/22)
- "Arabic and the vernaculars, part 2" (3/8/22)
- "Arabic and the vernaculars, part 3" (3/9/22)
- "Arabic and the vernaculars, part 4 — the case of Bible translations" (3/10/22)
- "Arabic as a macrolanguage" (10/21/18)
- "Arablish" (9/23/18)
- "Mutual intelligibility" (5/28/14)
- "Sayable but not writable" (9/12/13)
- "Sinitic is a group of languages, not a single language" (10/12/17)
- "Cantonese and Mandarin are two different languages" (9/25/15)
- "Taiwanese in France" (5/11/24)
- "Sanskrit is far from extinct" (11/29/23)
- "Sanskrit resurgent" (8/13/14)
- "Spoken Sanskrit" (1/9/16)
- "'In Pāṇini We Trust'" (12/15/22)
Peter B. Golden said,
May 12, 2024 @ 5:45 pm
When I was still teaching (until 2012), I had students from the Maghreb (Morocco, N. African Arabic) whose spoken language was completely unintelligible to Arab students from the Levant…or so they told me. Egyptian Arabic retains some old/archaic pronunciations, e.g. the letter "jim" ( ج) is pronounced "g"/, not "j," "ž" etc. as in the Levant. Because of its population size and developed motion picture and music industries, Egyptian (Cairene) Arabic is widely understood. In addition to numerous local dialects, there were also religion-specific dialects, e.g. Judeo-Arabic written in Hebrew letters, and Christian Arabic written in a Syriac-based alphabet, Karshuni.
GeorgeW said,
May 12, 2024 @ 6:04 pm
@Peter B. Golden: Yes, some of the Arabic dialects are not mutually intelligible. And, as noted, Egyptian Arabic is widely understood because of Egypt's political importance, film and music.
This is a problem with Arabic second-language learning. MSA is necessary for reading and writing, but doesn't work well for verbal communications, particularly listening.
Noam said,
May 12, 2024 @ 6:33 pm
> Urdu, which is mutually sayable with Hindi, but not mutually readable, is not taught at INALCO (source);
Although it is listed on the wall (Ourdou, two or three lines below the n of inalco)
Benjamin Geer said,
May 13, 2024 @ 1:19 am
Nobody anywhere speaks MSA in everyday life. It’s used only for written communication and formal, prepared speech, which includes TV news and (in some contexts) academic lectures. Nobody speaks it with their children or spouse, or uses it to order in a restaurant. MSA *vocabulary*, and to a much lesser extent structure, is used in all dialects, depending on the subject of conversation and the educational background of the speakers.
There are groups of dialects that are geographically near to one another and relatively mutually intelligible. Geographically distant dialect groups aren’t mutually intelligible.
I once interviewed the Tunisian filmmaker Moufida Tlatli. Ahead of me in the queue to interview her was a Palestinian journalist. I watched as the two of them tried without success to communicate in Arabic. They ended up doing their interview through an interpreter, with Tlatli speaking French and the journalist speaking English.
Fred said,
May 13, 2024 @ 2:18 am
‘Sorabe’ refers to Sorbian, not Sorabe.
anonymous said,
May 13, 2024 @ 6:08 am
India has Telugu (Telougou), Nepali, Bengali, Sanskrit, Hindi, and Tamil (Tamoul), that is 6 of its 22 "Scheduled" languages, represented on that wall.
It would be interesting to find the origin of the use of "ou" in the French name of Tamil. My theory is that it's because the "i" in the word தமிழ் is often pronounced somewhat like IPA ɨ by native speakers, especially colloquially.
anonymous said,
May 13, 2024 @ 6:09 am
… make that 7, as Urdu is also one of India's official languages.
Victor Mair said,
May 13, 2024 @ 6:19 am
@Benjamin Geer:
Thank you for your mind-blowing, yet wonderfully clarifying, comment. It confirms what I wrote about in part 5 of this series — which I could scarcely believe when I experienced it
Victor Mair said,
May 13, 2024 @ 6:22 am
More and more, the parallels between "Arabic" and "Chinese" emerge, and not just linguistically, but also politically, socially, culturally, and ideologically.
Victor Mair said,
May 13, 2024 @ 6:42 am
CRUCIAL QUESTIONS
When "Arabic" is taught in schools, colleges, universities, etc. outside of Arabic-speaking countries, are they being taught MSA? Is it only for writing? Do they also learn to speak it? How useful is MSA for getting around in the Arabic-speaking world?
When Arabic-speaking politicians and scholars and business persons get together for pan-Arab meetings, conferences, workshops, etc., what type of Arabic do they use?
What type of Arabic is used in the United Nations and other international organizations?
Coby said,
May 13, 2024 @ 8:42 am
What I find striking about the listing is that, as far as I can see, it contains only one "berbère", while (according to Wikipedia) "[t]he Berber languages have a similar level of variety to the Romance languages", something that the French — having long ruled over the territory where Berber is spoken — ought to know.
GeorgeW said,
May 13, 2024 @ 8:48 am
@Victor Mair: MSA is taught in schools. However, some of the pronunciation might not be standard like the 'jeem' [j] would be pronounced 'geem' [g] in Egypt.
I don't know for certain, but confidant that formal addresses and written material in pan-Arab conferences would be MSA (with some pronunciation differences). How they would speak informally, I don't know.
FWIW, there is yet another variety of Arabic in use – Classical Arabic in religious contexts. In the written form, diacritics are mandatory where they are not so much in MSA, except to disambiguate.
Jonathan Wright said,
May 13, 2024 @ 11:32 am
@Victor Mair: As several people have noted above, no one speaks at MSA at home and very few people can speak it competently off the cuff (the exceptions would include TV and radio reporters, some academics, and people with a religious training, especially preachers). In the eastern Arabic world, many people can communicate with strangers within the region by elevating their register, approaching MSA as far as they can. MSA really is an artificial language that arose in the 19th century from the interaction between the written literary Arabic of the pre-modern period (i.e. before 1800) and the impact of European dominance. New words and calque expressions were invented to meet the needs of modernity and European languages heavily influenced literary style. But MSA admittted very very few European words.
I'm a literary translator from MSA and some colloquials and I believe that the use of MSA is a serious obstacle to self-expression and creativity. Some Arabs think likewise but so far there are an eccentric minority. It's a fascinating subject.
Jonathan Wright said,
May 13, 2024 @ 11:39 am
I might add that neither Libyan or Yemeni Arabic are any more MSA than any other version. I've never found any claim that one dialect is closer than others to MSA or classical Arabic is credible. They have all evolved in different and interesting ways. Their rules on word stress and consonant clusters, for example, are very diverse. It's important to remember all Arabs these days are exposed to a wide range of registers from an early age, especially through media, schooling and religion. But many of them feel they fall short of MSA competence.
David Marjanović said,
May 13, 2024 @ 12:52 pm
I'd be quite surprised if it revealed anything about current French language policy directly. Rather, it reveals what is available at the INALCO, which in turn remains strongly influenced by French history (unusually good representation of North Africa, for example).
I concur that sorabe refers to Sorbian – I wonder if Upper, Lower or both; they're not particularly similar to each other apart from shared archaisms. (In terms of innovations, Upper is more similar to Czech, Lower to Polish, as you'd expect from geography.)
Mark Liberman said,
May 13, 2024 @ 12:57 pm
See "Maltese Arabic: Correction?"…
David Marjanović said,
May 13, 2024 @ 1:15 pm
More striking yet: "chleuh / tacelḥit" is listed separately.
Frédéric Grosshans said,
May 13, 2024 @ 3:08 pm
In addition to Berbère and Chleuh, there is also Kabyle, which is also a Berbère language
Francis Deblauwe said,
May 13, 2024 @ 4:46 pm
When I studied Arabic at the Centrum voor Levende Talen in Leuven (Belgium) in the mid '80s, it was MSA. It was helpful in my archaeological fieldwork in Syria and Yemen around the same time but unfortunately I don't recall to what extent, too long ago.
Benjamin Geer said,
May 13, 2024 @ 6:51 pm
‘When "Arabic" is taught in schools, colleges, universities, etc. outside of Arabic-speaking countries, are they being taught MSA? Is it only for writing? Do they also learn to speak it? How useful is MSA for getting around in the Arabic-speaking world?’
Until about ten years ago, in most university courses, they were nearly always being taught mainly or exclusively MSA or even Classical Arabic. Some universities also halfheartedly offered a bit of instruction in a dialect as an afterthought. Many students learned to speak MSA as well as read and write it. This often led to disappointment and frustration when students went to Arab countries and found that, although most people could understand them, they couldn’t understand what people were saying and couldn’t participate in ordinary conversations. You can certainly survive with MSA in Arabic-speaking countries, but if you want to do more than just survive, you’re going to feel very left out. In some Arabic-speaking countries, a few good dialect courses for foreigners could be found.
Like students of other languages, students of Arabic want to communicate with people. Therefore, in recent years, the field of teaching Arabic as a second language has slowly been undergoing a sort of revolution, as a few renegade academics have started publishing textbooks and teaching methods for dialects. For example, Munther Younes at Cornell University has developed an approach that combines MSA and dialect in a way that reflects how they’re typically used in practice. The fact that INALCO now teaches all those dialects is a sign of the times.
Chas Belov said,
May 14, 2024 @ 12:17 am
According to Wikipedia, the rock group Meen records and performs in Lebanese Arabic.
Benjamin Geer said,
May 14, 2024 @ 12:49 am
@Chas: most Arabic popular songs have always been in dialect, there’s nothing unusual about that.
Chas Belov said,
May 14, 2024 @ 7:59 pm
@Benjamin Geer: ¡Thank you! Then if I want to collect songs in a bunch of Arabic languages I just have to get them from different countries. ¡Nice!
Victor Mair said,
May 15, 2024 @ 9:56 pm
See "Aspects of Maltese linguistics" (5/14/24)
IA said,
May 24, 2024 @ 2:42 am
I could not find part 3 of 'Arabic and the vernaculars', but having just read the other parts, and even though there may be partial overlap with previous information, below I present a series of succinct quotes from over the years from linguist Steve Hewitt, someone well qualified to speak of Breton, Arabic, Welsh, and Georgian — among other things. (Here I can't reproduce any italics he used.)
1) Question: 'Can I travel all over the Arab world only speaking Modern Standard Arabic or Fusha?'
SH: 'You can, but people will find it pretty strange. No Arabs do that. Modern Standard Arabic is reserved for formal use, based on scripted texts. One Lebanese journalist tried for a day. The universal reaction was to assume that she was a foreigner, and everyone tried to reply in either English or French.'
2) Question: 'If someone wants to learn many or all of the major Arabic dialects, would learning Modern Standard Arabic first make learning the dialects easier? Or would it be better to go straight into learning them without learning MSA?'
'There are two quite different answers to this question. X
SH: 'The linguistic answer is “no, Modern Standard Arabic (MSA) is not the direct ancestor or common denominator of the various Arabic colloquials.” Charles Ferguson, in “The Arabic Koiné” [https://www.jstor.org/stable/410601] found that all Arabic colloquials share 14 separate basic traits which are absent from fuṣħà (classical / literary / standard Arabic), strongly suggesting that the various colloquials are descended from a variety which was not identical with fuṣħà, which Ferguson calls the “Arabic koiné”.
However, the practical answer is “there is little alternative to starting with MSA, since apart from Egyptian and Levantine, most Arabic colloquials are not well standardized, and really not much used in writing”. The only other possibility is to start with Egyptian and Levantine, and to branch out from there to other colloquials.
There are two further considerations which those interested should be aware of.
(a) The overwhelmingly dominant ideology in the Arab world is that fuṣħà (MSA, etc.) is the ancestor and common denominator of all Arabic colloquials, which are thought to be debased varieties of fuṣħà. Even though this is not warranted by linguistic facts, few Arabs are willing to disagree with this position.
(b) It is possible to produce fuṣħà according to two opposing poles: “full MSA”, faithfully reproducing all the various case and mood endings, most of which do not actually appear in unvowelled script, with the exception of sound plural -ūna (nom) / -īna (obl), dual -āni (nom) / -aini (obl), and indefinite accusative -an; or “streamlined MSA”, dispensing with all case and mood endings not graphically signalled by unvowelled script, thus bringing MSA much closer to the colloquials, which all lack these distinctions. Even the most formal Arabic has a rule of jazm ‘elision’ which warrants the dropping of these endings in sentence-final position. The “streamlined” production merely extends this rule to more and more elements within the sentence. A common saying about these unscripted case and mood endings is ijzim, taslam ‘elide [case and mood endings], you’ll be safe!’, i.e. you won’t make any grievous mistakes. It is this “bivalence” of the Arabic script (the possibility of either producing or dropping unscripted endings) which has strongly promoted the use of fuṣħà as a general written form. '
IA said,
May 24, 2024 @ 2:44 am
3) Question: 'What is the best way to learn Modern Standard Arabic and Lebanese dialects?'
SH: 'An excellent book to start learning Arabic with is John Mace, Arabic Today, Edinburgh University Press. It takes a very useful twin-variety approach, starting off with general Levantine Colloquial (Syria, Lebanon, Palestine, Jordan) in Latin transcription to give you a natural spoken variety for conversation. In the second half, it moves to Modern Standard Arabic (MSA) in Arabic script. Levantine and Egyptian Arabic are the two most widely understood colloquials. No Arabs speak MSA among themselves, and because of its grammatical complexity, few are capable of doing so without making basic mistakes; the idea of ordering a coffee or taking a taxi in MSA is ludicrous – no Arabs ever do that. On the other hand, nearly all written Arabic is in MSA; colloquial is very rarely written for publication. This approach follows the same order of acquisition as for Arabic-speakers: first, colloquial; then MSA.'
4) Question: 'What dialect of Arabic do people use in interviews on BBC Arabic or any other interview situation? I guess many people that are interviewed aren't proficient in MSA right?'
SH: 'Good question. It varies quite a bit, partly as a function of the seriousness of the topic and also personal preferences. Going from most formal to most informal, the following styles may be distinguished:
Full MSA with all MSA case and mood endings – extremely few Arabs (mainly religious leaders, literary academics, some broadcasters) can produce this spontaneously without making numerous basic mistakes.
Streamlined MSA, dispensing with most case and mood endings (the Arabic saying is ijzim, taslam ‘elide, you’ll be safe’!), thereby avoiding the danger of making mistakes. Many more Arabs are capable of producing this style spontaneously.
Educated Spoken Arabic (ESA), which is basically colloquial (often Egyptian or Levantine) morphology with a strong admixture of MSA lexical items. The effect of this is not so far removed from “streamlined MSA”.
Full local colloquial (usually middle-class, based on the capital city), especially Egyptian or Levantine, which are both quite acceptable in an international context – widely understood thanks to the media.
It happens quite often that the interviewer will use streamlined MSA, but the interviewee will prefer ESA or full colloquial, giving a “diglossic” conversation!'
IA said,
May 24, 2024 @ 2:46 am
5) Question: 'What kind of Arabic dialect is used in Al Jazeera?'
SH: 'In decreasing order of formality, these are the basic various levels of Arabic:
'Qur’ānic Arabic
Classical Arabic (fuṣħà t-turāθ “Heritage Classical”)
Full MSA (Modern Standard Arabic) – with all case-and-mood endings
Streamlined MSA – with case-and-mood endings only where apparent in script
ESA (Educated Spoken Arabic) – colloquial, often inspired by Egyptian or Levantine, with liberal admixture of formal lexical items from MSA, but basically colloquial morphology (simpler than MSA)
National colloquial – middle-class, based on usage of national capital
Local colloquial – lower-classes, local dialect
Usage on Al-Jazeera (and other pan-Arab stations, such as Al-Arabiya, BBC Arabic, Deutsche Welle Arabic, Radio Monte Carlo Doualiya – France, etc.) tends to be as follows:
Scripted news broadcasts, documentaries: Full MSA
Introduction to discussion programmes, spontaneous commentary: Streamlined MSA
Discussion programmes – ranges up and down between Streamlined MSA, ESA, and National colloquial when people get very excited.'
IA said,
May 24, 2024 @ 3:01 am
By the way, from The Wikipedia article about Levantine Arabic:
'"The Little Prince" was also translated into Palestinian and published in two biscriptal editions (one Arabic/Hebrew script, one Arabic/Latin script).'
Here's a sample of the first one. (The Hebrew-script looks quite modified. Can someone explain the modifications?)
https://shorturl.at/nZwjw
And here's the Latin-script one:
https://minerva-books.com/wp-content/uploads/2010/11/il-amir-le-zghir.pdf