Talk:Phonological history of French

Latin AU

According to the rules described on this page so far, Latin aucellus would have ended up as oseau (/o/). Something needs to explain how aucellus became OF aucel (/au/), thence oiseau. i.e. AU -> oi (/ua/) Fletpedia (talk) 21:24, 26 May 2014 (UTC).

See Wiktionary. The Old French form is oisel, not **aucel. The Latin ancestor could have been *avicellus instead, in which case the development might have been regular. I can't think of a parallel example (where -avi- stood in unstressed, or even better, immediately pretonic, first syllable) right now, though. User:Benwing is more savvy than I about Old French. --Florian Blaschke (talk) 15:44, 27 August 2014 (UTC)

Ah! There appear to be details not mentioned here or in Old French. It seems that oisel is like plaisir and intervocalic -s- is really /z/ (as still in Modern French) from /dz/, which behaves like a palatalised consonant. So, /ɔjzɛl/ < */ɔjdzɛl/ (early de-affrication of /dz/) < */aujdzɛl/ (monophthongisation) < */audzɛl/ (/dz/ ejects /j/ into a preceding open syllable) < */autsʲɛllu/ < */aukɛllu/ < */aukelluː/ ⟨aucellum⟩. So the i is a result of /j/-ejection, not from Latin -i-, which would have been syncopated anyway. --Florian Blaschke (talk) 16:20, 27 August 2014 (UTC)

Right. Single palatalized consonants in Gallo-Romance between vowels spat out a /j/ before them, and also after them if they were followed by a stressed open A or Ē, which resulted in A -> /ie/, Ē -> /i/. Thus plaisir /plaizir/ (OF) < plaidzir < plaidzʲieir < pladzʲeir < platsʲeːr < platsʲeːre < platsʲere < Latin placēre(m). I do think this is documented somewhere on this page. I'll take a look. Benwing (talk) 07:33, 29 August 2014 (UTC)

This is mentioned in "Table of vowel outcomes" second bullet point, but granted it could be clearer. Benwing (talk) 07:37, 29 August 2014 (UTC)

I was confused by the s (rather than c or z). Old French says that /dz/ is spelt as z, as in doze, which I expected to be present here too, but then I realised that /dz/ had already become /z/ at this point – onze, doze and treize could be special cases in that they had /ddz/ as the result of Gallo-Romance syncope, which was simplified to /dz/ along with the other geminates but not de-affricated (this appears to suggest that devoicing and /dz/ > /z/ actually preceded degemination). Moreover, the resulting diphthong /auj/ looked strange to me – I began to doubt whether /j/ was also ejected into a syllable ending in a closing diphthong, but since the rules only specify an open syllable for ejection to occur, which /au-/ certainly is, I figured /auj/ must be correct, seeing that the result is as expected. That was quite an aha! effect: simply following the rules faithfully and mechanically actually works out. --Florian Blaschke (talk) 23:54, 29 August 2014 (UTC)

/θ/

Second lenition: Same changes as in first lenition, applied again

And yet a couple of lines later we have "loss of θ" with no (other) origin for the segment given. So I gather it's not quite the same after all, and here /t/ lenited instead to [θ]? What about /p/ and /k/, then? Spirantization? Voicing? Nothing? --Trɔpʏliʊm • blah 22:26, 7 October 2014 (UTC)

virginis > vierge?

How did Latin virginis /ˈwɪr.ɡɪ.nɪs/ develop into French vierge /vjɛʁʒ/? By the looks of the vowel chart, it seems to me that it should have developed into verge /vɛʁʒ/. Most other words I can think of follow the rules of this chart, so I'm just curious as to why "vierge" apparently does not. Thanks! 76.18.160.47 (talk) 20:59, 2 November 2014 (UTC)