Help talk:IPA/Lebanese Arabic

Too many footnotes

The footnotes should be designed to help editors transcribe and readers understand. As such, there are way too many footnotes. This should not be a content fork for Lebanese Arabic. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 20:01, 8 January 2020 (UTC)

@Aeusoes1: is it the amount of footnotes or the amount of text in the footnotes that is the problem? Nehme₁₄₉₉ (talk) 00:14, 9 January 2020 (UTC)

@Aeusoes1: You're completely right, I overshot. When I get home tonight I'll rework what I've written. M. I. Wright (talk) 00:35, 9 January 2020 (UTC)

@Aeusoes1 and Nehme1499: So I'm with Nehme in that I don't feel like having a large number of footnotes is necessarily an issue on its own, as they don't impede reading of the table (the main content of the page) and they may indeed be necessary to "help editors transcribe and readers understand". Instead, the problem here was my missing the point of a Help:IPA page, as you've both mentioned, by writing too much and inadvertently creating an alternative Lebanese Arabic article in the wrong place where there was no use for it. I've tried to clean up after myself by shortening things up, using clearer wording, and focusing more on editing guidelines in the footnotes — but it still needs a once-over, and I'd like it if someone who isn't me could take care of the last four (I genuinely can't tell if the information is both useful and readable or not). But, again, I think it's OK not to reduce the overall number of footnotes any further, since it seems that can't be done without omitting important info. M. I. Wright (talk) 07:52, 9 January 2020 (UTC)

Let me go through each one.

a) Arabic emphatic consonants are conventionally transcribed as if pharyngealized, even though they are in fact uvularized in speech. Certain Lebanese accents today may further advance this articulation to velarization or even lose it altogether, but in general, IPA transcriptions should prefer the ˤ symbol in keeping with convention.

The particularities of emphatic consonants can be covered with a simple link to emphatic consonant somewhere in the table. Information about Lebanese accents can go elsewhere, since this information does not guide our transcription conventions. So this note can be removed.

b) The original Arabic interdentals may be preserved today in individual words whose pronunciation is a Modern Standard Arabic affectation. In all other cases, though, the interdental letters represent dental fricatives in Lebanese Arabic, where the spelling indicates that these words are Modern Standard Arabic borrowings rather than native vocabulary. (In native vocabulary, the interdentals were merged into the dental stops, but these will tend to be spelled as such rather than with the interdental letters.)

This seems like important information, but it's unnecessarily wordy and it's not clear how interdentals play into transcriptions if they're only used for "Modern Standard Arabic affectation." I would recommend something like ⟨ث⟩, ⟨ذ⟩, and ⟨ظ⟩ represent [t] [d], and [dˤ], respectively, except in certain words borrowed from Modern Standard Arabic, which are pronounced with [s] [z], and [zˤ].

c) There is a noticeable quantity of words which exhibit /zˤ/ as a reflex of original Arabic /sˤ/. Given the lack of a completely adequate letter to represent the new pronunciation, speakers may ignore it and spell such words with etymological ص regardless.

This is a factoid about etymology and should be removed.

d) Original */q/ ق continues to be pronounced as /q/ by the Lebanese Druze, but speakers in the rest of the country exhibit this pronunciation only marginally and mostly in higher-register words like /qur.ʔaːn/ “Qur'an”. In all other words, despite occasional regional variation, its most-standard pronunciation is /ʔ/.

How are we transcribing this sound? Are we using ⟨q⟩ or are we using ⟨ʔ⟩? Are we doing both and if so, what are the conditions in which we use one or another? If not, we don't need this note at all.

e) The */ʔ/ sound that was historically represented by the letters ء أ إ ئ ؤ has disappeared in native Lebanese Arabic vocabulary (as in all other contemporary Arabic varieties), and the letters are only used today in Lebanese Arabic to spell sounds that are still pronounced /ʔ/.

This is a factoid about historical sound change and should be removed.

f) In some accents, approached phonetically by American English /k/ before a back vowel or velar /l/: condo, core, class

We can either echo the English equivalent at Help:IPA/Arabic with somewhat like cut or just use condo as the English equivalent. No need for a note.

g) The velar stop /g/ occurs in native Lebanese Arabic words but is generally restricted to loanwords. It is realized as /k/ by some speakers.

This is probably fine, though we want phonetic brackets, not phonemic slashes.

h) The phonemes /p, v/ are not native to Lebanese Arabic and are only found in loanwords. They are sometimes realized as /b/ and /f/ respectively.

Same as g).

i) The suffix ة indicates that a word is singular and feminine. It is pronounced /-(i)t, -(e)t/ when the word is possessed by something. Otherwise, the letter only represents a vowel that varies between two distinct sets of pronunciations: the "lowered" [a ɑ~ʌ], and the "raised" [e~i]. In broad transcription, use /a/ following an emphatic consonant or any of /ʔ h ħ ʕ x ɣ r/, except following /iːr/, in which case use /e/. Also use /e/ after all other consonants. In narrow transcription, follow the same rules, except use [ɑ] after an emphatic consonant. (The pronunciations [ʌ i] are, respectively, regional variants of [ɑ e].)

This seems like important information, but this whole broad/narrow transcription deal and the squiggly lines indicating variable pronunciation doesn't provide enough clarity. Presumably we are transcribing one single Lebanese variety at one level of precision.

j) Broad transcription should use /a/ exclusively. Narrow transcription should instead use [æ] by default, but [a] when immediately adjacent to any of [r ħ ʕ], and [ɑ] when either immediately after an emphatic consonant or anywhere before one in the same word.

Seems fine, other than the "broad" transcription issue I mentioned already.

k) Except word-finally, there is no functional phonemic distinction between "tense" [i u] and their "lax" counterparts [e o], and for some speakers they are in free variation. Broad transcription should use /i u/ exclusively. Narrow transcription can use [i u] invariably before the semivowels [j w] and in light syllables, but [e o] in stressed and heavy/superheavy syllables. (This is very reductionist, as there is no actually-simple way to list an accurate ruleset. Broader transcription should thus be preferred.)

This seems important, though it could probably do with a reword to make it simpler. Something like, Short high vowels are [i u] before the semivowels [j w] and in light syllables, and [e o] in stressed and heavy or superheavy syllables

l) (The pronunciation of "long a" in a given context varies greatly from region to region, so if a word can be demonstrated to have a common pronunciation in violation of the following guidelines, record it as well.) The default pronunciation is /eː/ [eː]. In broad transcription, use /aː/ when directly adjacent to an emphatic consonant or /r/, as well as when anywhere before an emphatic consonant in the same word, and lastly when immediately following any of /x ɣ ħ ʕ ʔ/. In narrow transcription, follow the same rules, except use [ɑː] in the same proximities of an emphatic consonant and [æː] after [ʔ]. (The pronunciation /ɒː~ɔː/ is a regional variant of /ɑː/).

This can be reformatted to focus on just one transcription style. I also dislike the suggestion that editors can go against this if they feel like it.

m) In Lebanese Arabic, the original Arabic diphthongs */aj aw/ are typically preserved in all syllables that are not word-final. In word-final closed syllables, they are only rarely (and/or regionally) conserved, instead resulting in the monophthongs /eː oː/ in "standard" speech. In both cases, certain speakers may substitute /ej ow/, but this should not be favored in transcription.

This seems important but, as with b), we should be focusing the wording on letter-to-sound correspondence, rather than etymology

n) More-conservative speakers exhibit a full range of distinction between all of /i e u o/ word-finally, regardless of their allophony or free-variation status anywhere else within a word.

This isn't very helpful for transcriptions.

o) Only appears in monosyllables like شِي /ʃi(ː)/ “something” as an alternative realization of /-iː/, as well as traditionally in various suffixes spelled with ـي: the verbal second-person-feminine conjugational suffix seen in تَعِي /taʕi/ “come! (feminine)”, the first-person possessive enclitic seen in كتَابِي /kteːbi/ “my book”, the nisba suffix as in لِبنانِي /libneːni/ “Lebanese”, and others. The latter set of examples is merged by many speakers into -e as seen below, but in the reverse direction, original -e is also merged by a noticeable number of speakers into -i: شَغلِة /ʃaɣli/ “thing”, originally /ʃaɣle/.

This seems fine.

p) Traditionally and for certain speakers, only appears in the "raised" pronunciation of the singular-feminine suffix: شَغلِة /ʃaɣle/ “thing”. However, many speakers merge the various suffixes more-conservatively pronounced -i into -e as well, and there are also speakers who merge in the other direction.

This seems important, but it isn't providing enough clarity. How are we transcribing with this vowel?

q) Only appears in monosyllables like شُو /ʃu(ː)/ “what” as an alternative realization of /-uː/, as well as traditionally in the verbal third-person-plural conjugational suffix: إِجُوا /ʔiʒu/ “they came”. The latter is merged by many speakers into /-o/ as seen below, however, and there is also a smaller number of speakers who merge in the other direction.

This seems important, but it isn't providing enough clarity. How are we transcribing with this vowel?

r) Traditionally and for certain speakers, only appears in loanwords like French-derived /gatˈto/ and /majˈjo/, as well as in the third-person-masculine enclitic: كتَابُو /kteːbo/ “his book”. Many speakers merge traditional -u into this, however, and there is also a smaller number of speakers who merge in the other direction.

This seems important, but it isn't providing enough clarity. How are we transcribing with this vowel?

I hope my feedback helps. :) — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:09, 9 January 2020 (UTC)

Thank you very, very much -- this is invaluable. I'll get to work again. Regarding allowing editors to go against certain guidelines, my idea was that it would be for words strongly tied to a particular region that exhibits an alternative pronunciation. For example, عَكَّار Akkar would be transcribed /ʕakkaːr/ according to the rules that footnote sets out, and this is indeed its pronunciation in most of Lebanon, but its inhabitants tend to pronounce it /ʕakkɔːr/ (or /ʕekkɔːr/), which seems like something worth recording. I'll clarify that intent. M. I. Wright (talk) 03:23, 10 January 2020 (UTC)

The problem with this page

I was tapped from Wiktionary to create this last year, and I accepted headlong without really understanding what Wikipedia wants out of these kinds of pages. That wasn't good judgment on my part, sorry.

I think the biggest issue is that Wikipedia basically expects a Help:IPA-ified language to have an official standard, like how all the foreign-language IPA templates use [square brackets] and don't allow for phonemic transcription in /slashes/. It does make sense — non-speaking readers just want to know exactly how to pronounce something, not get mired in the abstract nitty-gritty of another language's phonology — but what kills it is that there are widely-spoken languages without a standard, and it can be impossible to objectively pick a specific variety of theirs whose phonetics to use. Lebanese and other vernacular Arabics fit that bill pretty well.

So, understanding none of that, I attempted a general description of Arabic throughout Lebanon, but then tried to graft on some semblance of a standard (that's what the whole "primary vowel" business is) out of a poor sorta-comprehension of what Wikipedia wants out of this kind of page. It's obvious now that this was a bad move, because it's not some random Wiki page's job to standardize a language on its speakers' behalf, but it's nonetheless the reason for the general weirdness of the footnotes & vowel table.

I dunno what to do with this page. It's neither 100% correct nor 100% appropriate for Wikipedia's needs. Trimming it down further would mean not representing "Lebanese Arabic" at all but maybe "Wikipedian Lebanese" (similar to the state of pages like Help:IPA/Egyptian Arabic, actually), but leaving it as is makes it kinda useless for the wiki... in any case, if anyone comes along and wonders why the page looks like it does, this is why. M. I. Wright (talk) 07:09, 17 April 2021 (UTC)

@Nehme1499: Thoughts? Is it possible to delete the page? (Deleting it wouldn't mean getting rid of any of the Lebanese transcriptions we did last year, it would just mean not linking them to a page that's more likely to confuse than clarify.) Thanks for adding framing like "Wikipedia transcriptions for Lebanese Arabic vowels may be either more general and abstract ... or more detailed and precise" to try to compensate for my fumbling this, by the way. M. I. Wright (talk) 07:20, 17 April 2021 (UTC)

@M. I. Wright: So, to be clear, you believe that the information being shown here is misleading/incorrect, and you want to delete it? If so, all instances of pronunciation shown in biographies should also be deleted as they are also "incorrect", if I understand correctly. Nehme₁₄₉₉ 12:31, 17 April 2021 (UTC)

@Nehme1499: None of the pronunciations are incorrect because they're valid pronunciations that would be found in Lebanon. This page is incorrect because its purpose is to make it seem like there's some standard Lebanese accent we're transcribing (when no such thing exists). And it bungles things by simultaneously trying to explain that there's actually a lot of diversity and no real standard, which just makes it confusing to someone if all they want is to figure out how to pronounce a word they've seen transcribed. (It would help a little bit if we could transcribe /phonemically/ and a lot if it were possible to hedge every link to this page with something silly like "A possible Lebanese Arabic pronunciation...", but neither of those are on the same page as what Wikipedia expects out of foreign languages.)

Does that make more sense? I guess I was too rambly at first. I don't see anything wrong with linking Lebanese transcriptions to the generic Help:IPA page, for an example of where to go from here. M. I. Wright (talk) 19:11, 17 April 2021 (UTC)

Yeah that makes sense. I'm pinging @Aeusoes1 who opened the last discussion to see what he thinks. Nehme₁₄₉₉ 19:16, 17 April 2021 (UTC)

I don't have an intimate understanding of the topic, it's my understanding that prestige can still exist within the body of non-standard Arabic varieties. Lebanon is an area where an urban population has risen from regional migration and Damascus is one of a number of Arabic-speaking capital cities with well-defined, stable prestigious varieties.

So if we are to pick one Lebanese Arabic variety, narrowing it down to the speech of Damascus and even to a particular variety within the city might be a good idea. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 22:07, 17 April 2021 (UTC)

@Aeusoes1: I assume you mean Beirut? Pinging @M. I. Wright as well. Nehme₁₄₉₉ 23:04, 17 April 2021 (UTC)

Oh, shoot. Like I said, I don't have a good understanding of the topic, and I'm apparently getting my geography mixed up. I stand by what I said about Damascus, but I'm not really sure if that same idea applies to Beirut. A source from 2003 that I read said that then-recent upheavals (e.g. war) have led to migration so that the status or direction of dialectal change in Beirut was uncertain. We'd have to check more recent sources to see if/how things have changed. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 00:39, 18 April 2021 (UTC)

@M. I. Wright and Aeusoes1: Would renaming this page "North Levantine Arabic" solve the situation in any way? Nehme₁₄₉₉ 12:31, 15 May 2021 (UTC)

Something like that would involve a larger discussion about how we divide and represent Arabic varieties. Do we have just five vernacular varieties? Do we do it by nation-state? Do we have just one with lots of columns? Something like that would involve a lot of knowledge and a lot of research. Perhaps the thing to do first is beef up Lebanese Arabic with well-researched content, and then return here to see how we want to reflect that content in our IPA guide. This key and that article should be consistent with each other anyway. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 14:04, 15 May 2021 (UTC)

Template:IPA-apc

@Nardog: I still don't understand why you would rather rename this page into Help:IPA/Levantine Arabic, rather than moving {{IPA-apc}} to {{IPA-apc-LB}}. The IPA symbols used here (and in the articles with the template) refer specifically to the Lebanese Arabic dialect, not more generally to the Levantine Arabic dialect. Nehme₁₄₉₉ 09:59, 4 November 2021 (UTC)