zompist bboard

Are there any words with /yj/, which would presumably be spelled <uille>?

drime.a3nm.net says none, and I even tried making the final /i/ syllabic. Perhaps there was once a sound change that el iminated preexisting/ yi/, since I'd think it'd be common in diminutuves.

You can see a list of words with <uille> here.

Most can be eliminated by eye, but I had to check a few. There are a few words like cuiller and juilllet where <uille> becomes [ɥijɛ], and I'm guessing that's what happened to any possible /yj/.

I can confirm that there are no words with the sequence /yj/ word-finally. The words in Zompist's list are either written -ouill, or have another vowel afterwards so the sequence is /ɥijV/.

I thought about it some time ago: I was trying to explain why French maille can be analyzed as three phonemes /m a j/ while English my is better analyzed as two phonemes /m aɪ/.

Among the vowels of French:
- /a i ɛ ɑ ɔ œ u/ can readily appear before a final /j/ (maille, fille, paye, bâille, momoï, feuille, bouille).
- /o e ə ø/ can't, but that's not very surprising since they rarely appear before a final consonant.
- Nasal vowels can't appear before a final /j/ either. Since they rarely if ever appear directly before a vowel, this isn't very surprising.

The lack of a /yj/ sequence is indeed a weird gap. I guess that's because /yj/ shifted to /ɥi/ instead. But there's a solid contrast between /uj/ and /wi/, so I don't know if it's a very good theory.

Ryusenshi wrote: Among the vowels of French:
- /a i ɛ ɑ ɔ œ u/ can readily appear before a final /j/ (maille, fille, paye, bâille, momoï, feuille, bouille).
- /o e ə ø/ can't, but that's not very surprising since they rarely appear before a final consonant.

Are there any cases at all where /e/ or /ə/ occurs before a word-final consonant in standard French?

My understanding is that /y/ in French usually comes from Latin /uː/. And final /j/ in nouns usually comes from Latin /lia/ or /lea/, which developed to /ʎə/. In some verbs, like "paye", final /j/ comes from words that lenited Latin intervocalic "c" or "g". But some people say "paie" instead, with no /j/. As far as I know, the same applies to other -ayer verbs.

I haven't found any Latin word that ends in /uːlea/ (Perseus words spelled with "-ulea").

"Julia" seems an obvious example of a Latin word ending in /uːlia/, but for some reason the French equivalent seems to be "Julie" instead of something like "Juille". However, as zompist mentioned, modern French does have "juillet". Wiktionary says it is pronounced /ʒy.ljɛt/ in Louisiana French. The CNRTL entry says

Prononc. et Orth. : [ʒ ɥijε]. Prononc. mod. fin xviies. sous l'influence de la graph. (Buben 1935 § 62). Vieilli : [ʒyjε] ds Fér. Crit. t. 2 1787 et Gattel 1841 mais non ds Fér. 1768 ni Land. 1834 ([ʒ ɥi-]); [ʒy-] ou [ʒ ɥi-] ds Littré, DG, Passy 1914, Barbeau-Rodhe 1930; [-εt] ds Land. 1834 et Passy 1914, rejeté comme vieilli par Mart. Comment prononce 1913, p. 326. Att. ds Ac. dep. 1694.

Looking up Latin words ending in "ulia" on Perseus, I found the place name Āpūlia/Appūlia which does seem to have a variant "Appuille" that was used at some point (at least in English!) although I don't know how it was pronounced. In contemporary French, "appuille" seems to occur as a misspelling of "appuie", which is pronounced with /ɥi/.

"Amuille" seems to occur in some French name or title "Mareschal d'Amuille", but I don't know how it is/was pronounced.

There seem to be "communes" in France named "Juillé" and "Écuillé", but I don't know how they are pronounced.

Ryusenshi wrote:I can confirm that there are no words with the sequence /yj/ word-finally. The words in Zompist's list are either written -ouill, or have another vowel afterwards so the sequence is /ɥijV/.

Actually, the sequence (well, its analogue/-ɥij/ anyway) occurs word-finally in aiguille.

You'll also hear the same sounds when people pronounce verbs in -uyer in a nonstandard way with -j instead of -i, again with /-ɥij/. That alternative pronunciation is probably the source of the misspelling "appuille" that Sumelic notes (i.e. eye dialect). There would be pretty much no reason for the -lle to show up if the speaker's pronunciation was /ɥi/ (i.e. as in the standard appuie).

Ryusenshi wrote:I thought about it some time ago: I was trying to explain why French maille can be analyzed as three phonemes /m a j/ while English my is better analyzed as two phonemes /m aɪ/.

Wouldn't that just be because (standard) French isn't generally considered to have diphthongs?

Sumelic wrote:Are there any cases at all where /e/ or /ə/ occurs before a word-final consonant in standard French?

Nope. There's a clear phonological constraint against that, reflected in morphology:
- geler /ʒəle/ -> il gèle /ʒɛl/
- céder /sede/ -> il cède /sɛd/

I didn't want to bring up the nuance that /o/ and /ø/ rarely appear before a final consonant (saute, meute) while /e/ and /ə/ never do. But since you've asked...

Sumelic wrote:My understanding is that /y/ in French usually comes from Latin /uː/. And final /j/ in nouns usually comes from Latin /lia/ or /lea/, which developed to /ʎə/.

Interesting.

Sumelic wrote:In some verbs, like "paye", final /j/ comes from words that lenited Latin intervocalic "c" or "g". But some people say "paie" instead, with no /j/. As far as I know, the same applies to other -ayer verbs.

Yep. I should have taken another example: oreille unambiguously has /ɛj/.

As for your other examples: I'm not sure about juillet, but I've certainly heard cuillère pronounced /ky.jɛr/ (though I have /kɥi.jɛr/). It's possible that some French dialects keep a clear distinction between /yj/ and /ɥi/; it just doesn't happen IMD, and apart from the proper names you mention, there are no words where the spelling clearly indicates /yj/. The funny thing is that /yj/ doesn't seem at all weird or difficult to pronounce (while */ej/ or */øj/ certainly would); it just happens, uh, to not happen.

Circeus wrote:Actually, the sequence (well, its analogue/-ɥij/ anyway) occurs word-finally in aiguille.

True, but it's still not /yj/.

Circeus wrote:You'll also hear the same sounds when people pronounce verbs in -uyer in a nonstandard way with -j instead of -i, again with /-ɥij/. That alternative pronunciation is probably the source of the misspelling "appuille" that Sumelic notes (i.e. eye dialect). There would be pretty much no reason for the -lle to show up if the speaker's pronunciation was /ɥi/ (i.e. as in the standard appuie).

True. As I said, that depends on the dialect.

Circeus wrote:Wouldn't that just be because (standard) French isn't generally considered to have diphthongs?

Well, that's precisely the question: why do we consider that English has diphthongs but French doesn't? As far as pronunciation goes, English my and French maille aren't that different. The reason has to do with phonology, not phonetics.

In English, the start of the vowel in my is neither the vowel of cat nor that of father (though that depends on dialects). Similarly, the start of boy, bee, bay don't exactly match any other vowel. Plus there are no sequences such as to *[ʊɪ] or *[ʌɪ]. So, in the end, it makes more sense to describe my as having an atomic vowel that happens to be a diphthong.

In French, the beginning of maille is exactly the same as that of mate. Each vowel either:
- can appear before a final [j];
- or, if it can't, has other reasons for not doing so.
So, it makes more sense to describe maille as having a vowel followed by a semi-vowel consonant.