Mark Liberman at the Log reports on a phenomenon of spoken French I hadn’t been aware of, the addition of a final nasal syllable to words that end in a consonant — hence “bonjourin” for bonjour. I was, of course, well aware of the addition of an added [ə] (“bonjour-euh”), which was omnipresent when I was in Paris in the ’80s, but I don’t remember noticing the nasal variant. Mark links to a delightful video in which David Castello-Lopes (auteur of the “Depuis Quand” program) gives examples and interviews Anita Berit Hansen, a linguist who has studied the phenomenon (Mark links to several papers by her, e.g., “The Covariation of [ə] with Style in Parisian French: An Empirical Study of ‘E Caduc’ and Pre-Pausal [ə]“). The executive summary is that it’s first attested in 1972 (and presumably existed for some unknown prior period) and nobody knows why it arose or spread (except that presumably it sounded “cool” to a lot of people), but the video’s fun and short enough (three minutes) it might be worth watching even if you don’t know French.
The title first got me off on a totally different track. Some Lebanese friends had told me French speakers in Beyrouth had the habit of saying Bonjour-ein, the “ein” being the Arabic mark for the dual. So it basically meant “Bonjour deux fois”.
@Paul: Interesting. I’ve read that it’s a Levantine thing to respond to marḥabā, “hello”, with the dualized form marḥabtain.
Now that you say it, I dimly recall something like this. Was in the gulf and hardy speak Arabic, so it didn’t really register …
Marhabtiin sounds completely normal to me, and the only variety I know is Moroccan Arabic. It’s definitely very common in Levantine (and Gulf) Arabic; it already occurs in the 1880 Baedeker for Palestine and Syria.
Does “bonjourə” have normal final-syllable stress on “rə”? What about “bonjourɛ̃” on “rɛ̃”?
No, the stress is on -jour; the rest is unstressed, sort of a filler.
You also get the same /ə/ as LH notes in between 3 consonants, e.g. in Arc de Triomphe, after the /k/.
Also, no, in reply to the comment there, which I replied to there as well, unless I am missing something about “Sur le pont d’Avignon,” that should be “e muet” in “danse,” since it is followed by punctuation and then a vowel. It is also true of spoken French that a sound that happens to sound like /ə/ or /E/ (sorry, I know it’s not quite correct) is added, which confuses L2 learners, but it fills a gap when northerners might add a sound and southerners might say the letter.
I think the first two times anticipate the third, which is on s’y danse tout au rond.
I believed you that the video was going to be delightful, but I didn’t know I’d be so lucky that “Auteuil Neuilly Passy” would be one of its main examples!
unless I am missing something about “Sur le pont d’Avignon,” that should be “e muet” in “danse,”
I believe that “e muet” is pronounced in poetry and art song, for the sake of the meter. It certainly is pronounced in recitals of French art song by French singers, except before a vowel in the following word.
As I understand it, when “e muet” ceased to be pronounced in conversation, the poets refused to throw away all their experience and knowledge of meters and come up with new meters to fit the new pronunciation.
Gary, I fully agree. Songwriters (not just arty ones) have more leeway than regular speakers about whether to pronounce “e muets” or not. It is not just for the sake of the meter: for instance, said or read aloud (whether in verse, prose or regular speech), Je vois la vie en rose would end with [roz], but in song the final e shows up, ending the musical phrase rather than the verse, hence [ro-z∂]. This is not because the singer chooses to add it, it is indeed part of the song and leaving it out would be as bad as leaving out the final musical note.
As for bonjou-re, I noticed this in France a few years ago among young people. I thought it was a way of adding audibility to the final r, which is otherwise very weak and sometimes hardly audible.
I don’t recall bonjourin but my experience is limited as I don’t spend much time in France. But it might result from the addition of “hein” (silent h), which is indeed similar in use to Canadian “eh”.
Sur le pont d’Avignon (a children’s song and dance)
(sung while doing a round dance)
Sur le pont
d’Avignon
L’on y danse, l’on y danse, (no pause between the phrases)
Sur le pont
d’Avignon,
L’on y danse tous en rond.
(On the bridge of Avignon
There is dancing, there is dancing.
On the bridge of Avignon
There is dancing, we”re all in a round).
(spoken after stopping)
Les beaux messieurs font comm’ ça
Et puis encor’ comm’ ça
(the “fine gentlemen” each time take off their hats and wave them around near the ground in a sweeping motion before putting the hats back – 3 Musketeers style)
(sung resuming the dance)
Sur le pont (etc)
(spoken, as above)
Les bell’s dam’s font comm’ ça
Et puis encor’ comm’ ça
(the “fine ladies” are curtseying, holding their wide skirts)
(sung, etc)
Sur le pont (etc)
(Other people doing typical things can be added ad libitum)
marie-lucie, the version at the linked article uses “on,” and now, especially with “l’on,” it makes more sense. (Sorry, I was referring to the first instance of “danse”).
But no, generally, in traditional poetry, “e muet” is pronounced only between two consonants, and never at the end of a line, in which case it is not counted for purposes of syllabification. I believe that the omission also occurs when it is followed by a vowel. I need to check on whether punctuation affects this, e.g. a comma in the midst of the line, when the next word still begins with a consonant. Contemporary poetry sometimes follows these rules, but often it is simply silent. I was listening to “La Vièrge à midi” by Paul Claudel, and the reader didn’t always pronounce it, nor did he always pronounce “e caduc” mid-word.
Matthew,
The point is that the rules are different for poetry (never a schwa at the end of a line, never counts for syllabification) and song (schwa often pronounced at the end although may or may not be pronounced within the line). So the rules are actually more relaxed for song. Modern poetry as you say is a little more like song with respect to schwa inside a line.
When writing the song above I first wrote on y danse, on y danse, which indeed would still have schwa in both verbs since the repeated sequences are separate sentences (punctuation reflects this), then I remembered it was l’on, which sounds smoother but otherwise there is no difference for the final of danse.
While I am at it: I could not think of a less awkward way of describing the fine gentlemen’s sweeping hat gesture. I asked my daughter and son-in-law who are English speakers and they could not either! Perhaps you have a suggestion?
Bonjour, hein? sounds a bit rude to me.
What is this supposed to mean?
Hello, huh?
Nice day, eh?
> final r, which is otherwise very weak and sometimes hardly audible
I swear I’ve heard native French speakers pronounce final (and possibly postvocalic) /r/ as something akin to [ɐ], especially after /i/ and /ɛ/. Maybe /r/ in that position is somehow phonotactically unstable (postvocalic /r/ Sprachbund phenomenon with German/Danish/RP?) and some people overcompensate as m-l suggests?
Compare Pardon, quoi. “Uh, excuse me, please let me get past you; your conversation is blocking my access to the table.”
Fascinating – I’ve never noticed.
However, it’s precisely along the French border that German keeps [ʁ ~ χ] behind short vowels.
Bonjour, hein? sounds a bit rude to me.
It’s definitely not hein — it’s just a nasalized version of the “standard” form with [ə] (“euh”),
I don’t think French de-rhotification is an area effect from German, since it happens in Quebec too. It seems like it’s more common in certain words/expressions, so I’m not sure how much of it is an active phonological thing, but it’s certainly audible in “parce que” being pronounced as “paskə”. I’m curious if this would make “par” and “pas” homophonous in certain contexts or if the /r/ is there enough to make them come out [pa] and [pɑ] (or however far down and round the speaker has that vowel), respectively—but I keep forgetting to trick my francophone friends into ad hoc phonology experiments.
Cluster simplification of /rskj/ in parce qu’il y a, which I’ve seen spelled paskya in a comicbook, may be a separate phenomenon.
dainichi: “parce que” being pronounced as “paskə”.
This is the way I pronounce it in casual conversation.
For me this is not homophonous with pas ce que because I have the traditional two a ‘s, but someone who only has the front one would probably pronounce both sequences “paskə”.
David: Cluster simplification of /rskj/ in parce qu’il y a, which I’ve seen spelled paskya in a comicbook, may be a separate phenomenon.
I think it is just the same phenomenon. I also pronounce Il y a as [ija] or [ja] in casual speech.
I’m not talking about “parce que”, which I see as an isolated case. I’m talking about wholesale vocalization of postvocalic /r/, possibly only after certain vowels.
https://learnfrenchbypodcast.com/mp3/004.mp3
At 1:43, the native speaker says “voyage d’affaires”, ending in what to me sounds like [ɛɐ̯] or thereabouts.
Marie-lucie, right, I got it. All I’m saying, again, is that the original post had a response which was not accurate.
Something like they “doff their hat like a Musekteer” would be the best way.
When I was learning, I always heard /r/, so an elided “pas ce que” or “parce que” are not the same. But that also could be in error…being that my mum’s family is very non-rhotic, from MA, I am sensitive to dropping /r/ and people adding it, whether at the end, as people sometimes do in New England, or in the middle of a word, as Southerners sometimes do (“warsh” for “wash”). This carries over to French a bit.
Also, yes, teaching final “ll” followed by a vowel or final “–re” is difficult. You have to know to listen for it occasionally, and you have to know how to do it for emphasis or reading poetry, but otherwise the sound is subtle or nonexistent.
Right, but the Quebec accent usually neutralizes the a-â distinction to â in word-final open syllables, which is why I’m curious about how far the r-weakening goes and how it interacts with normal open-closed syllable behaviour in Quebec French.
No, the /r/ is there, as some kind of voiceless uvular approximant or something, much like in “Berlin” at 1:39 – clearly different from the [ɛɐ̯] that my native German phonologies would predict. But between these two, in “dernier”, it’s completely absent; I think it’s just [dɛnje] without even a diphthong.
…And now that I hear it again, I remember that I’ve noticed people pronouncing “dernier” like this before. Perhaps it’s the front vowel in maximally unstressed position that kills the /r/.
> the /r/ is there, as some kind of voiceless uvular approximant or something
I say it’s there in the form of [ɐ̯] :P. I do hear some kind of glottal stop after that, not sure if that’s the voiceless uvular approximant you hear. I guess it’s possible that’s actually the /r/, and I’m reanalyzing it as hiatus-avoidance, but it’s obvious to me that what comes before it is a diphthong. When I listen to the syllable in isolation, it sounds exactly like Danish “færre”, which is something like [fɛ(:)ɐ̯]. I definitely also hear a diphthong in the first syllable of “dernier”.
To me, it looks like a case of the /r/ conditioning a diphthongal allophone in the preceding vowel, after which the /r/ itself (almost) falls away, which I assume was the course of events in RP/German/Danish.
@ə de vivre: I have the impression that in Quebec French /ɑ/ is often used instead of /a/ before “r” even when the word is not spelled with an accent; e.g. “tard” and “barreau” apparently have /ɑ/, not /a/. Given this, I wonder if “par” and “parce” even have [a] in the first place in Quebec French, or if they might have /ɑ/ all the time, even with a pronounced /r/—unfortunately, I’m not sure if I have any good way to check this at the moment; do you know?
You’re right about dernier after all; I seem to have interpreted the [ɐ̯] as part of the [n]! In Berlin it’s not there, however; all there is is a voiceless lenis fricative [ʁ̟̊]. In affaires there’s an unreleased version of the same thing, possibly preceded by a very, very reduced [ɐ̯].
OK, so I got myself together and booted up Praat.
For the “er” in “dernier” (at 100.88 to 10.94 sec. in) I get approximately an F1 rising from 410 to 750 Hz and an F2 falling from 2020 to 1660 Hz.
For the ehm… “air” in “voyage d’affaires” (at 103.16 to 103.23 sec. in) I get approximately an F1 rising from 560 to 670 Hz and an F2 falling from 1960 to 1760 Hz.
So there seems to be some opening/centering movement in both cases (more in the former, as @DM correctly points out), but whether it’s enough to call them diphthongs is beyond my knowledge. Also, the endpoints of the diphthongs (if that’s what they are) are probably fronter than I realized, so maybe I should have said something like [ɛa̯] or [ɛæ̯].
(I’m a complete amateur at Praat – and phonetics for that matter – so feel free to take my analysis apart.)
Eli: in Quebec French /ɑ/ is often used instead of /a/ before “r” even when the word is not spelled with an accent; e.g. “tard” and “barreau” apparently have /ɑ/, not /a/
In my own speech (conservative European French), le barreau, like la barre and other derivatives, has /ɑ/, not /a/. Similarly for le carreau though I think I hesitate about le carré.
That explains a few things.
This could be helpful in your ecommerce business;
compare ecommerce platforms
compare ecommerce prices
top ecommerce vendors
top ecommerce website builders
top rated ecommerce platforms
best ecommerce platform
Anything Cyrillic gets flagged, but that spam gets through?
Unrelated fact, I’m sure: Many e-commerce platforms are built on WordPress.
Remarkable. I removed the URL and am looking askance in the direction of Akismet.
(postvocalic /r/ Sprachbund phenomenon with German/Danish/RP?)
There’s a big fat hole in the middle of that theory, and its name is Dutch. 🙂 In that language, /r/ can be [r], [ɾ], [ʀ], or even [ɹ] depending on dialect, pre- or post-vocalic position, and so on. But one thing it never is, is zero.