From: Miguel Carrasquer
Message: 32596
Date: 2004-05-13
> For a better picture, I'll resume it as follows:Thanks. That (apart from the analogical stuff) can go
>
>I. Latin -cl-, -gl- > Common Romanian -kl'-, -gl'-, conserved
>as such in AR, MglR & IR;
>
>II. In DR, early enough we have -kl'-, -gl'- > k,g + palatal
>glide [j]
>
>III. Before (stressed) vowel, this palatal glide...
>
>1. is preserved as non-syllabical element of diphthong with
> /a/ and /o/ (e.g. "scloppus" > /s^kjop/, "clarus" > /kjar/,
>"clagum" > /kjag/, "(per) vigila(re)" > /privegja/, etc.);
>
>2. is absorbed by front vowels /e/, /i/ ("inclino" > /1nkin/,
>"*glemus" [for cl. "glomus"] > /gem/);
>
>3. induces /u/, /&/ (/1/) > /i/ and is subsequently absorbed by
>the vowel ("includo" > /1nkid/, "glandem" > /gind&/, etc.);
>
>IV. Further evolution of diphthong [ja] resulted from above
>described process is conditionned by phonetical environment
>(including umlaut) and follows mostly similar patterns with
>/ea/ (in alternance) since they merge naturally:
>
>1. preserved as such before an /a/ or /&/ in the next syllable:
>/kjam&/ (written "cheamã"), /gjac&/ (written "gheaTã"), /gjaca/
>(written "gheaTa"), etc.
>
>2. turned into /e/ if the next syllable contains an /e/, or
>in alternance with /e/ in the next syllable (if any), or if
>dipthong is final; it also shifts into /e/ when becoming
>unstressed:
>/ke(j)e/ (written "cheie"), /keme/ (written "cheme"), /kem'/
>(written "chemi"), /kemàre/ (written "chemare"), etc.
>
> Since a word like "chiar" is not involved in any alternance,
>the diphthong [ja] is preserved even if the syllable is final;
>the same holds for "cheag" whose form does not change at the
>plural ("cheaguri").