Help talk:IPA/French/Archive 2

Some of these seem plain wrong
saut/haut/bureau/chose are said with a sound similar to o as in load/blow/retroactive, however story is said with o as in store/pour/or/oar etc.

pâte seems to be dependent on region but the 'common' pronunciation seems to be a as in bat rather than bra

sort/minimum/hôpital are all listed as sounding like the british off. However, I don't see that those three french words actually use the same vowel sounds. Listening to the the various pronunciations on wiktionary suggest that they are different sounds. sort seems similar to off, but the others sound like different sounds. The sound in minimum sounds more like or perhaps. The sound in hôpital sounds more like.

I'm posting here for discussion first as I imagine these are pretty subjective.

Source: English native/resident - I don't believe that any of my comments here are affected by regional variations, at least not within the UK.

Stripybadger (talk) 08:36, 18 December 2020 (UTC)
 * It's kind of difficult to find something perfect for these sounds. The sound of French haut is . We were trying to go with something monophthongal, and the sound of English load is in American English, which is at least somewhat close, and  in RP, which is pretty off. For some of these other IPA guides, we've done dialect-specific approximations and that may be called for here.
 * Do you have a source that says as much about pâte? I'm seeing the opposite claim elsewhere around the project, but I didn't notice any citations to see what experts have to say on the matter.
 * It's not necessarily a good idea to rely on our ears for this sort of thing. It may be one of those dialect-variant kind of things that we then just have to make a choice about (I notice that the wiktionary entry on hôpital lists two different pronunciations. It looks like a few words were added by in this edit. — Æµ§œš¹  [lɛts b̥iː pʰəˈlaɪˀt]  17:27, 18 December 2020 (UTC)
 * The vowel in story is often transcribed with ⟨ɔ⟩, but the actual quality in both RP and GA nowadays is between the cardinal and, if not closer to [o]. The vowel in off is /ɒ/ in RP and /ɔ/ in traditional GA, and in cot–caught merging accents it is somewhere in the vicinity of [ɑ~ɒ~ɔ]. The authoritative Trésor de la langue française informatisé transcribes both minimum and hôpital with [ɔ]. There are not many places where pâte is still pronounced differently from patte at least in France, but in those places the vowel in the former is indeed closer to that in bra than in bat. Nardog (talk) 17:42, 18 December 2020 (UTC)
 * To me it doesn't come to a surprise that the vowel in sort is further back than the last vowel in minimum or the first vowel in hôpital. Though I am more familiar with Belgian than Metropolitan French it seems clear that pre-rhotic vowels are way backer than their allophones in other environments. That has no impact on our (pan-)English equivalents, though. Love —LiliCharlie (talk) 19:08, 18 December 2020 (UTC)

I have always seen the French e written with a schwa (inverted e) as here. The entry compares the sound with the English first syllable in "about". But to my ears, the sound has always been closer to the oo in foot. Is it just me? Johnm307 (talk) 16:38, 6 April 2021 (UTC)
 * It's not just you, French /ə/ is [ø] or [œ] in realization. Rounding and backing both lower F2, so English /ʊ/ is not a bad approximation. But neither is English /ə/, which has phonological and historical reasons going for it. Nardog (talk) 00:30, 7 April 2021 (UTC)

Guttural R example is wrong
The page claims:

ʁ 	regarder, nôtre[3] 	Guttural R, Scottish English loch, but voiced

But the ch sound in loch sounds nothing like a R of any kind. — Preceding unsigned comment added by 88.97.62.77 (talk) 12:23, 8 October 2020 (UTC)
 * Voice it and pull it a bit further back in your throat. (Although I wonder if "guttural R" really does sound like an R to monolingual English-speakers.) Double sharp (talk) 09:09, 6 April 2021 (UTC)
 * French /r/ is devoiced in many positions and ranges between velar and uvular, so [x] is a fairly good approximation. You don't even need to voice it or pull it back. Nardog (talk) 00:22, 7 April 2021 (UTC)
 * Fair enough. But I did wonder if [ʀ] might sound more like an R of some kind to the OP's ears. Double sharp (talk) 06:53, 7 April 2021 (UTC)

Anomaly?
Hi.

I don't think /t͡ʃ/ in French is a "phonotactic anomaly" but rather a full-fledged phoneme since:

1) it appears in NO native French words allophonically (except for some dialects such as Acadian, where it is an allophone of /k/, but not in Standard (European) French), only in more recent loanwords;

2) native speakers intuitively feel that it's a separate phoneme so they use a special trigraph "tch" to write it;

3) if it were just an "anomaly" (whatever that's supposed to mean), it would arise NATURALLY in every single environment in which there were /t/ and /ʃ/ one after the other, which does not happen at all from what I know;

4) if it's not a phoneme, why did you not remove /ŋ/ as well? The status of this sound as an independent phoneme is even more questionable than that of /t͡ʃ/, let's be honest.

All in all, I'd say /t͡ʃ/ is a marginal phoneme borrowed from recent loanwords, especially from English, rather than merely an allophone of /t/. If you're still not convinced, I've got another argument: /t/, /ʃ/, and /t͡ʃ/ form minimal pairs, e.g. "thèque", "chèque", and "tchèque", in which only the first consonant is distinctive. That's enough to consider the latter an independent phoneme, albeit with a limited distribution. 83.9.190.72 (talk) 06:09, 30 July 2022 (UTC)


 * We shouldn't include $⟨tʃ⟩$ separately in the key unless there's consensus among linguists that it is a distinct phoneme. A quick look at Fougeron & Smith (1999), Walker (2001), Gess, Lyche & Meisenburg (2012), and Battye, Hintze & Rowlett (2000)—all of whom do include /ŋ/—indicates not.
 * We don't list $⟨tɬ⟩$ in the Spanish key either even though /tl/ occurs tautosyllabically in Nahuatl loans in certain varieties of Spanish, as this can be explained as a difference in phonotactic constraints (much like rhoticity in English) and there's evidence that it's not articulated differently from other stop–lateral clusters. I don't see how this is different from that.
 * I don't think anyone would dispute that [tʃ] occurs in French, but you need to show us reliable sources if you insist it's a distinct phoneme and not a cluster comprising /t/ + /ʃ/. Nardog (talk) 06:49, 30 July 2022 (UTC)